Hi Josh , Apple does stream @ high resolution above 48 kHz via a USB connection & airplay 2 @ 24 bit depth 48khz brick wall. I do agree with you but only for non studio playback. I’m subscribing keep up the great work 👍.
Modern vinyl & cassettes are even worse. And the horrible loudness wars. The whole entertainment industry has become one big scam, they don't care about quality, just making fast money.
Hi-res audio is not a scam.This video focuses on frequency rate and while I generally agree high frequency rates can be wasteful, the audible benefit of "hi res audio" is playback at the file's original high bit depth. Recordings almost invariably start at 24bit (or 32bit) - and there are subtle differences between the original and the dithered 16bit file, most noticeably when comparing in low level signals such as less detail in reverb trails and reduced spacial effects in stereo when comparing the original 24bit recording to a dithered 16bit file. Good equipment in a good room or good headphones are needed to hear the difference, but it is there.
Completely agree with you. It's a big topic and almost deserves its own discussion separately but you're completely right when we talk about recording, mixing, mastering, etc. After music has been mastered is a different story, since most modern music is SO heavily compressed at master that even 12 bits is more than enough to represent such limited dynamic range. Dithering definitely does affect sound and quality, and one could make a debate for (and against) that as well. I've known some who didn't want to dither, so they just recorded in 16 bit from the beginning to avoid it entirely and believed they were getting better masters as a result.
There ARE slight audible differences, even if you can't hear past 15kHz because of time domain and phase distortions created by steep anti-aliasing filters. Agree that other factors are way more important though
Yes, I have heard that argument, and it may very well be true. But as a general rule the quality of the converters will still matter drastically more. I personally have never been able to tell the difference, and haven't seen any studies that show people are able to either. Maybe a few individuals just have golden ears, who knows.
@nuznikas Maybe the reason you're hearing a difference is because Apple Music uses AAC? They may have intentionally made other formats sound not quite as good. That would definitely be incentive to get people to use Apple Music.
@@JasonWW2000maybe not making it sound worse but wouldn’t be surprised if they applied some sort of DSP effect on their own music so it tricks you into thinking it sounds better.
Designing CD players in the 1990s the digital brick-wall filter was a separate chip to the oversampled DAC. I was surprised how different different filters sounded in listening tests and part of the design process was to make up a few PCBs with different filters to find which one sounded the best. I don't think the sample rate itself is the issue, but increasing it past 44.1KHz does open the possibility for less of a "brick wall" filter and a more gradual transition between the pass band and the stop band, if that is the issue. A perfect brick wall filter, i.e. no transistion band width, will ring indefintely, The question, "does a higher sample rate sound better", maybe the wrong question if it is the digitial filter that is potentially affecting the perceived sound quality. Simply doubling the sample rate and doubling the filter clock won't fix that and will probably sound just as good/bad. The filter has to take advantage of the extra sample rate and extend the transition band. Not sure if any oversampling filter/DAC chips today even do that.
Terrific video on this subject. One of the very best I've seen so far. Would love to have you make a video on your thoughts on the SACD/DSD audio format.
Thanks for the compliment, it made my day! :) I will make more A/V videos in the future, especially focusing on things like converters, formats, etc. I'll try to squeeze that in there somewhere because it's a really interesting part of technology history. DSD vs PCM is really interesting too, that might be a good subject for a video.
@@michaelbeckerman7532 DSD was another breed , technicly it's a Delta sigma brut force RAW digital recording that not using sampling rate but a 1 bit PWM digital flux with 2 megaherz or 5 megaherz bandwith to sound almost like a 96 kHz 24 bit, but technicly it's not PCM data , this one use another Delta sigma convertion with sampling, all of the high res Wav and FLAC or the lossy format AAC, MP3 or surround DTS Dolby AC3 and atmos are derivative or based on PCM for recording and reproduction. With DSD you need that Big bandwith because it's a simplified digitized process, there is a big of digital noise that was filtered beyond 50 khz so you can have minimum true 30 kHz or almost 50 kHz or frequency response the sound is also more linear and smooth it can reproduce almost 120 db of Dynamic range but depend of your converter, there is only a few pro device that can record this in true DSD the old Sadie Sonoma system, the Pyramix and also some RME and Prism converter with a Tascam DA 3000 master recorder but it's rare to have true DSD recording you can't apply plug in or DSP on a bit stream like that, so they convert DSD to PCM and also PCM to DSD is possible but it kill the process of listening a true analog tape transfert to a DSD, it's very smooth and without any harshness if you have that it's probably a PCM to DSD convert file, so you have no benefit from against a FLAC or a Wav in 24 bit.
It makes sense for recording wildlife or using for industrial settings such as recording machinery. Pre-recorded files, in hi-res, are also fine for people that sample. The only problem with any higher def, audio or video, is what resolution the original was done in.
I'm curious, why is this only about hi res delivering a broader analogue frequency range? How does a higher sample rate impact detail by filling in more of the analogue shape and reducing averaging? Or how it can work with dac variables?
While it's not completely technically accurate I like to explain it in a visual way that makes sense to people. Think about sample resolution as the horizontal resolution, and bit depth as the vertical resolution. Both matter, but in different ways. One is "snapshots" or "steps" of the frequency shape, and the other is snapshots and steps tracking amplitude. When you think about it visually like this it suddenly becomes a lot easier to make sense of all of it.
@JoshChristiane that's where I personally don't understand. Because if there are more bits in a slice, the horizontal view becomes smoother and less jagged so I'd expect the detail to be more prevalent, no? With less bits per slice and more jaggedness, I'd expect the conversion to analog would fill less accurately with the less info given.
@@baaltroth1957 It does exactly what you're saying but human ears do not detect it. It's like looking at a monitor that doesn't flicker in our eyes because we can't pick it up but it's still happening. Same as the 4k to 12k argument.
@MrBigspendn cool. I got it now. Now then for dacs, if the bit rate is higher aka sending more detail, then the dac's performance could be better so some of the detail like piano frame resonance after releasing dampers could be fuller and more lifelike in its behavior?
I uploaded a video to my channel explaining exactly why hi-res is no better than CD quality. When you subtract the CD-quality signal from the hi-res version, it cancels out completely, leaving only some inaudible noise. The only thing hi-res can offer is frequencies above 20 kHz, which you can't hear anyway.
I have read on chose 44.1 due to limitation in the recording equipment that converted beta cam. Philips also wanted 14 Bit instead of 16 bit. but where was the review of the CD sound when it arrived. It was that the sound was flat and sharp treble. Then CD players got over-collection and dacs with more bits, there was also a time with One bit DAC technology
I recently significantly upgraded my whole hi-fi setup. It supports up to 192khz through the whole chain and more than that at the DAC. I had Qobuz for a while, sounded nice, BUT my best sounding music is still on physical CDs (Pat Metheny, Peter Gabriel, James Taylor, Toto, Doobie Brothers, Steely Dan etc.) - I have ripped my entire 10,000+ disc collection to flac with eac and play through plex to my wiim pro (bit perfect transfer to the amp via coax) and it's all up to the amp and speakers - they were perfectly paired with my room and budget (not for everyone) - Arcam SA20 and Dali Oberon 5s - to my 45 year old ears they sound amazing. (yes I could have tripled the cost and got better I'm sure) One thing of note: I also have a number of recordings in DSD (Super-audio CD) and they blow my mind - BUT again - this is to do with the mastering (NOT necessarily the format). Some standard CDs were badly compressed but the SACDs were mastered properly - if the same mix had gone on standard CD I'm sure it would have sounded amazing. This is exactly the same reason why some of my vinyl sounds better than CD - nothing to do with the format and all to do with the mastering.
Absolutely agree with you. 99% of the time people think higher rates are better is just because the master was done better on those, and has very little to do with the actual resolution itself. In the same way that photographers with more expensive cameras tend to take better pictures (because they're professionals) so it's a form of survivorship bias. Amateurs think the expensive gear is better since it's used for all of the better photos, but when a good professional uses the cheaper cameras they often get equally as great professional photos. Audio can be the same way, for sure. Guitars, microphones, preamps, etc. You can still do a million dollar recording on a Squier, but professionals just want the best. Sounds to me like you have a really good audio setup though, a lot of your gear is dream-level gear, which is awesome! Thanks so much for watching and your great comment!
I have used the digital photography analogy myself, but I add an extra bit... When you are upscaling music, you are starting with desired result. When you make a jigsaw puzzle, you start with the desired result. Chopping the jigsaw into twice as many pieces doesn't improve the the quality of the resulting image. To do that, you need to start with a higher quality image. In music, I have had mixed results listening to "Hi Res". Some tracks sound like they started with an mp3, some sound like real life, but I've not heard many that sound better than cd.
I totally agree. But there are many people who will argue upscaling improves music magically somehow, lol. If it improves it to their ears that just means they don't like the original, because all it can do it make it different, not better.
I think the better analogy is to add IR and UV to digital photography/video (and in turn, TVs and monitors) to capture a higher range of frequencies than a human can perceive as 'better quality'. You didn't even mention the bit depth and the loudness war
I agree... I'm older but still with pretty good hearing, and I have a CD player with SACD capabilities. Since these discs tend to have regular CD information as well as SACD, I figured out how to toggle between the two. I tested a few SACDs and some of them did sound a bit better set to SACD, but the differences were pretty tiny and on others I couldn't hear much of a differnce at all, which leads me to believe that the differences were likely more in the mastering than the format. Other digital audio standards that are higher bit depth don't make sense to me either as though they improve dynamic range, the dynamic range of a regular CD quality digital file is so great already that the differences aren't likely to be audible, particularly with 98% of music which doesn't even contain nearly that much dynamics. The CD Redbook standard, though very old at this point, is well enough thought out, with enough real understanding of the limits of human hearing that it still holds up. All of this is not to say that improvements can't be made on the hardware end as different DACs do sound different, and some are more resolving and smoother sounding than others. The basic format, though, at least as far as CD quality, is still about as good as it gets, though, in my opinion, at least as far as what we're able to discern...
Now a rational argumentation about if Hires worth while or not needs some problem definition: 1) sound isn't quantized nor in time nor in amplitude, both are continuous. 2) the often quoted by everybody Nyquist Shannon theory tell us the minimum requirements you need (note: minimum or enough it will depend) to retain all the signal's information. 3) every practical device has design constraint, you can't have more of one without pay on something else ( difference between engineers and stylists ) and more, practical implementation of a certain theory quite differ from theory . From 1) one can safely assume (ideally) that the more is the better both in terms of sampling rate and bit rate. Especially thinking on AD conversion. 2)&3) Theory was demonstrated mathematically but to take it as an absolute true, technology must have developed something working like this: you need all the sample (all! i.e. if you have a piece of music with timing of 5 minutes all sample from 00:00 to 05:00 min) to DA conversion taking in account and process all of them at the same time. It isn't the way present technology is working indeed, DA conversion act of a stream of samples processing them say one by one (this isn't quite true but the idea is more o less that, if you are thinking about nowadays digital filters you are pretty much on track and yes they need to be quite long, a lot of number to deal with). So, you just noted that there is gap from what it should be in theory and practical implementation, so till the gap isn't close ... in a way or in an other (going to high rates is a sort of brute force if you will). Some back thinking: - CD standard 16b/44.1kHz was adopted accounting for support's space limitation (it was the best was possible to do at that time). Means that is THE STANDARD? - Take an old CD from '80 play it from a matching class old and new players, why new player sounds better? Cd and it resolution are the same aren't? - from another piece of industry (multichannel oscilloscope dealing with hi freq. ,some GHz on every channel, set in the range 3X to 4X the theoretical sampling frequency to avoid artifact (signal shape pre and post alteration). But consider most are intended to deal with square waves signals, not audio. - one point which can mark a no go towards high resolution (actually I mean over 96-192 kHz or 88.2 - 174.6) could be a degradation due to the management of all this data not about the computation limits but the meaning itself. Control theory suggest to identify the dominant time (or equivalently frequency) behavior and limit the control bandwidth to that in order to avoid to follow ghost. Stated in alternate way sure that due to the need to have very long digital filter (isn't the filter long obviously, is the polynomial implementation which is long) we run in a data obsolescence. The risk is to better follow the high spectrum's and loose control at low end. Remember 2) take all samples at the same time is very good (100% on track with theory) stay in the middle there may be problem to deal with. - last point is something you can judge only based on personal preference and taste. Favorite music is yours favorite, no comment about. But there are some way to listen and some kind of music for which every discussion on Hi or Low resolution does not make any difference. Head phones vs room+loudspeakers have peculiar pro/cons (stage spatial reconstruction is quite limited with headphones, at most a little bit grater than you skull, and everybody can understand a classical orchestra does not fit in you skull), stated that 1kHz tone is exactly the same (just to recall my previous post). Actual played instrument, you have a reference to compare (go to concert hall) form most of the rest can be everything. Same for bit depth, a piece of music with no dynamic, more or less at same level for entire duration recorded just at -6 dB from the maximum 0dB can count on all 16 bit (taking the CD case), different is if you need to accommodate a full orchestral peek and in same piece you have a pianissimo, signal in last one can move around first 7 bits just flickering 0 1 on a couple of them and you filter can't prevent you to have something like a stair signal (good quantity of distortion).
That's because Tidal streams at much higher bitrates. Amazon music is really low bitrate, I think it's below MP3 320kbps, Tidal is significantly higher than that and offers lossless streaming. But that has nothing to do with resolution itself, that's the actual file format they're using for storage. Think like WAV or AIFF vs a Codec format such as low quality Mp3.
What i really like is music in surround (either quad, 5.1, 7.1 or any iteration of Dolby Atmos) i have an argument that clarity in recordings (or resolution) is better achieved by spreading a recording through several "pipes" instead of just two (or even one when in mono), what do you think about this argument? I would really like to know your thoughts.......
You make some excellent points and I agree with you. I also love binaural mixes, so anything specifically mixed for headphone, those mixes just have so much dimension and space. I always wish more albums had multiple versions. If it were up to me I'd produce a radio album edit (mastered), then an unmastered edit (only mixed, high dynamic range), then a binaural or surround sound mix. That'd be awesome!
The highest frequency a musical instrument can reach is generally considered to be around 5,000 Hertz (Hz), with some instruments producing even higher harmonics that may be audible to certain individuals, though most of these high frequencies fall outside the typical human hearing range. Audiologists test a person's hearing at frequencies ranging from 125 Hz to 8000 Hz, with the most common frequencies being 250, 500, 1000, 2000, 3000, 4000, 6000, and 8000 Hz. These frequencies are slightly broader than the range of human speech, which is primarily between 500 and 4000 Hz.
Totally correct. There are harmonic frequencies we can hear in the 15k range, and when I put a high cut filter at 15kHz I can definitely hear a little sparkly high-end missing, so obviously we are hearing something up to that range, but it's not what I'd call the "vital" life of the music. With that said, after about 15kHz-17kHz I hear basically nothing, and our ears are most sensitive to high frequencies between 5k and 8kHz as you mentioned, so that's where a lot of the energy goes. MP3 basically works this way, by cleverly cutting out frequencies our ears are less sensitive to in order to save space.
Nothing beyond 8kHz really matters for generic hearing. There's nothing important beyond that which will effect communication and essential living. Many elderly people can't even hear to that, but it is true that if you lose too much high-end it will affect your life negatively in other ways.
Great explanation Josh! Definitely opened my eyes to audio quality (this is new to me). I'd love to hear your take on audio in cars. What's the best way to consume music in a vehicle? I assume most people stream from Spotify or some other app. Does Aux vs Bluetooth vs USB matter?
My understanding is that Bluetooth reduces quality significantly because the stream is generally in SBC, so bluetooth (might) not be the best quality, but even then it's likely unnoticeable considering most car sound systems are very poor. I imagine you couldn't hear the difference between 128kbps and lossless on even a very nice car stereo. Spotify streams at 96kbps on "normal" settings. When you choose "High Quality" in the app it switches to 160kbps. In my testing 160kbps is quite hard to tell from lossless, you may lose some harmonics but the difference is so small it doesn't practically matter on most systems. In a nutshell Spotify is definitely good enough, especially if set to high quality in the app. I wouldn't have any complaints with it at least. Audiophiles tend to obsess over that extra .01% of quality that's impossible to notice unless you have golden ears and $2000 headphones. AUX is the best quality if you're connecting your phone directly to your car sound system, I personally just stick with that and have no issues.
That was a fascinating discussion to listen to. I had no idea how many brands, in this category, are taking advantage of the consumer by capitalizing on their ignorance in this field...Thank you for enlightening us!
This is like video games where the hardware is advertised as 4k but the game is rendered at a lower resolution and then upscaled. However, there are many instances where the upscaled 4k game looks better than playing the game natively at 4k.
That's true, and I can't debate that. Same thing with photos, sometimes upscaling does look better... But often it looks worse (due to denoising). Kind of depends on how well it was done to some degree. It can help with aliasing and comb filtering problems though. It's still important for consumers to understand exactly what they're getting, same problems with vinyl right now.
Thanks. I have an engineering background too. I still listen to vinyl as my main music enjoyment. Digital is for background music for me or in the car traveling.
Agree on most parts especially the marketing ploy, but i beg to differ in my current setup i did an abx and blind tests between 16/44.1 and 24/44.1. I've noticed a slight difference with both bit depth in terms of clarity on the instruments, synths and vocal tones much clearer. (I'm a metal fan) I do not wanna try again anything higher than 44.1khz sample rate it triggers my tinnitus and short sharp headaches on longer playback times
Very interesting. I have not heard somebody claim higher sample rates triggering their tinnitus. We definitely need more research on tinnitus specifically. Thanks for watching!
@@JoshChristiane thanks for responding. Forums like hydrogen audio and reddit as well termed it as ear fatigue iirc. Havent looked in-depth of it but their experiences makes sense especially the headache part. Agreed we need more scientific data on this matter. Take care man!
Intermodulation artifact products (either mechanically in tweeter and vinyl, or electronically in the chain) in the ultrasonic Freq's, can also produce lower audible products in the audio range, we probably don't want that
Thanks for bravely laying out the facts: they aren’t in dispute. The record business is just that: those that are willing (and hopefully able) to pay overs for a digital recording can conveniently do just that with this branding.
Thank you for the nice comment. No industry is perfect and deception exists everywhere in basically every asset class, but it's good to have information helping consumers know what to avoid.
My experience has been a little different in that I have multiple versions of some of the same recordings. While frequency response is a factor in playback and listening, there is quite a bit that happens between the high and low ends of the audible spectrum. For me the difference could be vocal clarity, the decay of a cymbal or the rolloff of a guitar riff - noticeable between pressings/versions. You could probably argue the difference is between mixes or masterings, but those differences remain. Some of my SACDs make the CD versions unlistenable by comparison, but I may have missed any references to physical media. At the end of the day, though, we all do what makes the most sense to us.
Absolutely there will be differences. But the differences you're hearing will have nothing to do with the resolution itself, and more to do with the quality of the conversion. The same exact converter at 192kHz will sound different at 96kHz, and also different at 44.1kHz. So a lot of people do these tests at home unscientifically and do notice a difference. Your DAC or AD/DA may perform much better at 192kHz and thus you're correct in saying it sounds better. But I've heard converters that sound distinctly better at 44.1. It's not the resolution but rather how the DAC handles filtering, aliasing, conversion and clocking, etc. It's all in the IC. Some of the highest quality converters ever made (and still used in Hollywood) are 44.1kHz Hollywood custom converters, as well as some older Burl gear. In scientific studies that are done with absolute equal quality conversion at 44.1 and 192kHz, no professional could tell the difference. It is will within the margin of error as well, so we know if there is a difference it's so small it doesn't matter. Ted Jensen, arguably the most famous mastering engineer to ever live said he can't hear a difference, even on his $300,000 setup that is as close to perfect as possible. With all of that said though, test your converters. Run them at 192, 96, 48, 44, and find out what sounds best on your system, then record at that. For my particular converters they sound best at 384kHz, I'm not sure of all of the internal reasons why, they just do, so I record at that in 32 bit then down-sample and dither at the end. I get an extremely clear, high-quality sound that way and will continue to do so. But I also recognize that's just my converters.
I think, it's not about Hz, but about bit in digital perspective, because CD recording is base on binnary digit. You can get same analogy in digitaly process of image. In digitaly definition of 16bit image and 32bit image, you will get much rich color in 1 squere inch of image at 32bit image. So, in 32bit sound, you can get more rich digitaly collect same Hz data to covert to digital code/definition in one package digital data sound. Sorry about my English. 🙏🏻
Oh you're absolutely correct. Nobody is debating that 192kHz is significantly higher quality than 44.1kHz. 32bit is better than 16bit, no question. The question that we must ask is if our ears can tell the difference, as well as if our speakers and AD/DA can reproduce the difference, and generally the answer to that is no. Blind tests consistently show that humans simply cannot tell the difference past the point of diminishing returns.
Came to debunk a new pair of headphones, came away with that AND valuable information regarding recording audio, which is very much appreciated for self taught music producers like me. Great video!!
In Greece also some companies put in digital platforms fake digitized files. They add strange fake frequencies and they close the sound with muddy bass. Then they increase the loudness extremely even -7LUFS integrated with True Peak over +2. They don't have the original files and they take songs from other sources. Even songs released on compact disc many companies don't own now
Another issue is simply that most hifi speakers cannot do 20hz to 20Khz. Most of the recordings out there especially early digital, were capped at 20Khz. Even much of the SACD catalogues were actually recorded from material including DAT tape, and contained no information above 20Khz. Another interesting audiophile concept is that treble beyond human hearing actually improves transients in percussion and bass. Personally not many hifi speakers even at £80'000 are doing decent transients, percusion, or bass anyway. Regardless, I love old DAC's and see no point these days in getting some of the higher end ones any more, most audiophile gear has coloured output stages to create a feeling of difference, a house sound, so people pick items they like, when in reality DAC's from Topping and SMSL as expamples are more accurate than most of the audiophile dacs, and having a neutral sound is not that appeasing. I mean there are nice DAC's, I like the Mola Mola Tambaqui and if I had the disposable income I would buy it, but the reality is it is not 10x better than a Topping D90 discreet, and none of them are significantly better than many of the late 90's audiophile dacs, which is why some manufacturers have been revisiting those chips.
You are absolutely right on! I'm glad you understand it because it seems like not a lot of audiophile people get that. Marketing is a powerful tool at getting people to spend more money on stuff they don't need. A good pair of HiFi speakers is totally worth it for me, but there is definitely a point of diminishing returns, like with anything else I suppose. I'll stick to my Focal Twin6 Be's.
However, they say that the sharp filter in CD players is negative for sound quality. It is also said that the sampling rate also affects lower frequencies
The same company pushed minidisc technology for decades on the argument that it truncated inaudible frequencies below 20 Hz and above 20 kHz. It seemed awfully hilarious that Sony was now promoting HiRes technology a few decades later.
Very true. Sony does a really good job in terms of engineering for many of their products, especially their higher end stuff, but for some of the cheaper things they sell they like to lean into gimmicky marketing.
I was choosing between an iPod Classic 7th gen, and a FiiO M6. Chose the last one because the price was insanely more convenient (€50) which I bought obviously used. I now read some things about Hi Res audio… If I paired tha iPod with some Audio Technica headphones, would I get the same quality than on the FiiO?
With something like an iPod or any other small listening device the weak point won't be file format. You will almost certainly hear no difference between the formats and resolutions the devices use, the big difference will come from the discrete headphone amp built into the devices. Some units have really high quality headphone amps and sound amazing as a result, and others sound awful. I can't comment on the FiiO because I've never used one, but iPods were known for having decent quality for their size, especially with low Ohm headphones. My advice would be headphones that are low Ohm (resistance) enough not to stress the amp on such a small music player though.
As I have listened with headphones of 30 euros (around 30-40 dollars), SENNHEISER HD-250, all Hi-Res sound to me so fakely digitized. They have fake frequencies. They have hard kick drums that distort, dull muddy bass and Extreme loudness. Many older are -8LUFS and True Peak over+2, even original cds are mastered at -13LUFS with True Peak below 0. Also at original compact discs they have balanced bass and treble and higher frequencies are clear, generally in original they have crystal clear sound even I rip in Mp3 or flac.
I have noticed some of this as well. Obviously it's a complex topic, but if you have a trained ear you can hear things like processing, low quality EQ, fake soundstaging, etc.
I totally agree with CD standards. Me thinks there's more life in Red Playback than many would have us think. Only way to go higher res is that hi res equipment must be used during recording sessions, so how many recordings have been done this way ????? Even so how much better is it compared to 16/44 ?
It's a good point. It's not that you can't do better, of course you can do better, but after modern mastering smashes all of that dynamic range anyways you're looking at less than 12 bits of necessary range, at most.
How could a consumer find out what the sample rate of the original master recording so that when buying music, we know whether we are being sold upsampled or original content?
Great question. Unfortunately, there is no easy answer. You only know what information they make available. But I can say as having been a tracking engineer in Hollywood for years that no studio I ever went to recorded above 48kHz. I'm sure some exist, but all of the studio work I did was 44.1 and 48kHz 24 or 32 bit. At my own home studio I record 384kHz 32 bit, but that's unusual. Many studios have older high-end converters as well, and not all of them support 192kHz. Some of the older converters sound better at 44 anyways. I only knew what Coldplay's album was recorded in because I read an interview in SoS magazine about the gear they used, and the mixing engineer specified the recording details, otherwise I'd never have known. I've also worked on albums where I know for a fact they were recorded at 48kHz, and now suddenly I see 192kHz versions of those albums on Qobuz.
You've saved me from falling for the scam ever again. Didn't realize Hi Res is just another marketing ploy. This really helped me understand. Super informative!
@@JasonWW2000kind of, but the hi-res sticker doesn't make whatever headphone capable. The recording/mastering makes a muuuuuuuuuuuch bigger difference.
100%. Pitch correction is on every single vocalist now, it's so frustrating and awful sounding. Billie Eilish doesn't use pitch correction, so I can appreciate her and her producers for being a modern artist that doesn't cave in to the pressure. Modern mastering is another can of worms, absolutely awful quality and destroys everything it touches.
These things confuse me a little. For instance, my very nice Topping USB DAC has "hi-res" labelling, but but will only stream via my computer at 44.1kh and appears to not be something I can change in the DAC settings (not that I need to, or care). For fun, I change the audio property settings on my PC to 48kh and it literally sped up the music. Also, interestingly, I have a rather decent amplifier with a "higher end" DAC chip which streams at 48kh, but the 44.1kh Topping DAC sounds better and more detailed. Like they say, I guess, it's usually not about the DAC chip, but the way it's implemented.
This is a classic case of the disconnect between marketing and reality. Many DAC's sound better at 44.1kHz. If you change a sample rate of a song for playback without doing an actual conversion then it will just speed it up or slow it down, you need to do an actual conversion before you can playback, so that's why you had that issue. Some converters can be a pain to switch around for different music listening as well. In the end the quality of the conversion itself matters WAY more than any fixed resolution.
@@JoshChristiane Thank you for the reply and detailed answer. I'm glad I don't have to worry about bit rates and sampling frequencies so much, just so long as the music sounds good! :)
Well CD quality is good . Well even somewhat esoteric, I believe it was said that 96 khz and more were not necessarily audible, but sometimes responsible for a certain ambience etc . Subconsciously. One would have to compare a to b.
I read many years ago that 44.1 means that a 20 Hz audio signal is sampled 2000 times per second while a 20 khz audio signal is sampled twice per second
The best sound test comparison could be made in the mountains in the big picture in nature. The best ears are definitely the people who live in the mountains far from civilization. Not audiophiles. Noise is poison to our senses. Audiophiles who start reviewing audio equipment in such conditions are probably the closest to the truth. This is a necessary post to people would get closer to the truth. Happy end of the new year to everyone
Totally agree. Technical arguments are right, yet most of the streaming services follow the path of offering more and more Hi-res content. I personaly only want the 44KHz versions? but can’t find them anymore on streaming services. I need to play music with digital altered files that have probably artifacts in the frequency and time domain.
The reasons, why SACD and all other HiRes formats haven't make a big breakthrough: 1. decent price players and sound systems capable to take full advantage of HiRes sound quality potential are not widely available. 2. Very few titles produced in TRUE HiRes are available, and popular hit music nearly at all. 3. The sound quality of CD 16/44.1 is good enough. 4. Most of commercial music even nowadays, the sound quality is not even good as CD makes possible. So, if the sound quality on the studio master is not even good enough to take advantage the full potential of CD quality, SACD and all other HiRes formats just simply does not make any sense...
You are absolutely completely dead on correct. This was the perfect summarisation too. The great majority of music comes nowhere near maxing out the headroom and quality representable by 44.1/16, so we need to focus more on the mastering wars first. Mastering is basically destroying music, and the industry should put its focus on solving that problem before anything else.
@@aquaevitae I agree... I've heard some SACDs that seem to have a bit more detail and sound more open than their CD counterparts, but I'm not convinced the SACD versions weren't mastered a but differently and if we were able to co pare the same master on the two different formats, we likely wouldn't hear a differnce..
Totally agree but it's so hard telling people. They're convinced they can hear the difference between a 24-bit 96 kHz (or whatever) & a CD quality file. The only exception I'd make is for multi-channel music mixes eg Dolby TrueHD Atmos where the extra bandwidth actually serves an obvious purpose.
Absolutely agreed. 100% right and good point about multi-channel audio as well, since obviously that's split up. People really think they have golden ears, it's just an arrogance thing. When I bring people into my studio and compare the difference literally no one has ever been able to hear a difference. Ted Jensen (the most iconic mastering engineer in history) claims he cannot hear a difference as well. If he can't, nobody can.
I always assumed (and it's just an assumption) that "Hi-Res" just means higher bit rates than, say, the old iPod .mp3 and AAC standards (128kbps, 256kbps, etc.). But you're right -- if you think you're hearing a "better quality" sound from 24/96 or 24/192 files than from 16/44.1, it probably has to do with the recording, mixing, mastering and playback of the track than it does with the data rates.
Hi-Res is only a scam if you ignore the other variables that usually go with "Hi-Res" recordings. What's most important is how something is recorded and mastered. A lot of today's music is recorded and mastered for sale to people wearing earbuds or listening with car audio. "Hi-Res" audio is recorded for people listening with high-resolution playback equipment. Mundane content and the music industry not rewarding artists, recordings with compression, equalization, noise reduction, etc. all contribute to what Steve Wilson in "The Sound of Muzak" calls "squeezing the soul out" in much of modern music. Many CD-resolution recordings sound great when these other variables are considered. But it takes a lot of listening and research to find them. It's much easier to find high quality recordings in "Hi-Res". Beware of "Hi-Res" tracks from poor masters. But if modern pop music is your thing, ear buds, subwoofers, and standard def are great. Streaming vs physical media is an entirely different topic, but can also make a big difference, albeit in sometimes unexpected ways.
However, a hi res edition may have another better Master with better dynamics less dynamic compression. even a Vinyl release today can have better dynamics than the CD version, less dynamic compression
This is a big part of the equation. A LOT of times the masters for the hi-res counterparts are different and can deceive people into thinking the technology itself is better, when in reality it's just a higher dynamic range master or something of the sort. It's not unusual to have mastering engineers produce 3 or 4 versions of every song. One super compressed for radio, one slightly better dynamically for the CD, one for digital release, etc.
Who has the perfect listening room, without reflections? Most recordings are mixed using a conbination if high end speakers and cheap in car setups to get the best compromise. What I would live to see is a different mixes/masters for different setups. I.e. a high end mix for speaker setups that can produce the full audible dynamic range which will be very different from the average consumer based mixed which sounds reasonable in cheap speakers/headphones.
Analogue audio is sampled an infinite number of times i.e. it is continuous, CD are 44.1 KHz so it not continuous there are small pauses of nothing we perceive it as continuous the same way a TV appears to show motion but what is happening is the picture is changed 50 times a second....
Exactly. That's why vinyl is so interesting to me, and making a big comeback. I love old analogue equipment, and especially old vinyl records from when everything was recorded on tape. Same is true for film cameras.
Um, no, that's not how it works. Imagine a plunger in a pool: the plunger taps the surface of the pool at specific intervals. This creates waves which are seen by our eyes as waves. A cd provides signals like the plunger. We hear the "waves" in the air created by the "pings" of the digital signal. We absolutely do not "hear" the spaces between the digital "pings." You never actually hear the "pings at all. You interpret the waves created by the pings as sound. (in this example, higher bit depth simply means the plunger can go deeper into the pond--which leads to a difference in dynamic range, not a difference in frequency.) As Josh says, all you really need to cover the range of human hearing is a digital signal at double the speed of the highest frequency we can hear. This isn't just a guess, it's relatively easy to prove mathematically. This is completely different than the persistence of vision effect that allows us to see a procession of still images as motion.
The fact that dither is added to 16-bit is proof that 16-bits is not enough. Now 32-bit float is often used for production, so there's no real reason not to release modern recordings in 24-bit. One way to use the 8 additional bits might be to use 4 of them at the top so less clipping is necessary and 4 of them at the bottom so dither is unnecessary.
So dithering is only added to 16 bit if you're down-stepping from 24 or 32 bit. If the music was recorded in 16 bit then dithering is not necessary and not used. So it has nothing to do with whether 16 bit is good or bad. I do agree that recording in 24b and 32b brings advantages that are worth it, but in the actual end product our ears can only perceive so much dynamic range and 16 bit already exceeds that. The problem is even more obvious when we consider almost all music today is mastered, and mastering destroys most of music's dynamic range anyways. The vast majority of modern music could easily be played back at 12 bit and sound exactly the same, that's how little dynamic range is retained after mastering. Still, for my own productions I record in 32 bit for mixing purposes.
In consumer audio there is no need to have more than 70 dB (about 13 bits) of dynamic range. This limit comes from the properties of human hearing, the practical limits of the acoustic environments people listen to music and what makes a pleasant listening experience in the first place (there are no symphonies written for farting ants and rocket engines, because nobody would enjoy such insane variation in loudness level. Instead the masses listen to DR6 pop! ). 16 bit digital audio doesn't really "need" dithering from perceptual point of view, but dither is added, because it is a very easy way to avoid quantisation distortion. It is at least technically the right thing to do. Distortion free sound with a little higher uncorrelated noise is better than correlating distortion due to truncation. As a matter of fact 24 bit version can contain so much noise that it acts as self-dither when truncating into 16 bit. Since dither allows sound to "fade into" noise the same way it does in analog audio, it is about how much the dither noise masks the signal. By using aggressive noise shaped dither, it is possible to reach perceptual dynamic range of 110-120 dB in 16 bit digital audio! From practical point of view this is insane overkill, but it is possible. In music production 24 bit is very useful, because it provides a lot of freedom as to how the dynamic range is used and also all the manipulation of the sound doesn't lead to harmful cumulation of noise. However, when a piece of music has been mixed and mastered, the consumers don't need more than 13 bits, so exporting into 16 bit makes a lot of sense.
Unfortunately the ones that can actually hear hi resolution audio will never notice or care about it. By the time you can you are to old and cannot hear those frequencies. As a child I could walk into a home and hear the CRT tvs squealing even if the volume was down to nothing. That lasted until I was around 16-18.... After years of factory work, Gunshots at the range, club music turned up way to loud, now i can not hear the beep of the thermometer when i take my temperature. I have to have my kids or wife tell me when its done. Im only 46 years old. When i bought my home i installed speakers into the walls and ceilings of the living room to get perfect surround sound. I can still hear the surround sound but the high frequencies are gone. Not even the beep from your thermometer.. it happens faster than you think. Time is unstoppable and moves faster as it passes. Think of a roller coaster hill. The longer your on the hill the faster it goes. Slowly at first and faster and faster and faster... I hate my ears have become crap. Dont get me started on my eyes. THE YOUTH IS WASTED ON THE YOUTH. Belive that
This is a good point too. You become wiser and more knowledgeable with age, but by then you're maxed out at 15kHz anyways, haha. Thanks for watching and the great comment!
Hi Josh, Really appreciate this video. It helped explain a lot of things to me. I had a couple of follow up questions for you. If I'm looking at purchasing an album on Qobuz (for this example its the Album "Tell me I'm Alive" by All time Low) it will give me the option to purchase it in both 16bit/44kHz and 24bit/48kHz. Is there a way to find out whether the original recording was recorded in 44kHz vs the 48kHz? How does the bit rate of 16bit vs 24 bits affect the sound (assuming the sampling rate stays the same)? Thanks!
Great question, and therein lies a major issue in the industry. It's pretty much impossible to know what it was originally recorded in unless the creators specifically state it on its wikipedia page, or you happen to know the exact gear and settings they use some other way. SoundonSound magazine had info on the Coldplay album, so that's how I found out what it was originally recorded in, but otherwise I'd never have known outside of just using common sense. I will say this though, as an ex-producer of many albums for many artists, 90%+ of music recorded digitally is recorded 24-bit - 44.1kHz. I can only recall one single time sitting in a session where 48kHz was used, and that was a Switchfoot album (vice verses), specifically because they intended the songs to be used in movie soundtracks potentially. Final delivery of that album was 44.1kHz though, so in the end it was pointless. In terms of the difference between 44.1kHz and 48kHz, you simply cannot tell a difference unless one of them was a bad convert, so I wouldn't worry about that. Bit depth matters more, but only during recording. When you're tracking you get better headroom, lower noise-floor, and more detailed recordings by recording > 24bit, and some albums are recording at 32bit now even which helps prevent clipping issues later on. Bit depth just gives you more vertical resolution (so each snapshot/stem itself has a great range of velocity storage). But once the album is bounced down and mastered there won't be an audible difference between 16bit and 24 or 32bit, because the headroom is the same for everything (0dB) in the digital space. In a nutshell: Just buy the cheapest one you can get. Nothing will be better than 44.1kHz - 16bit with very very few exceptions.
@@JoshChristiane Hi Josh, thanks for the reply. I appreciate your honesty with how audio tracks are recorded. After watching your video it sparked my curiosity and I did further research and yes I agree with you the bit depth gives you a larger noise floor when you are at 24 bits, but as you mentioned in your video even if you had perfect ears (I certainly don't) and the top of the line equipment you wouldn't be able to hear the difference unless you blasted the sound at a high enough level that it would cause hearing damage. at this point it seems like the only value of the "hi-res" tracks are that they show up as HQ on my DAP. I have a follow up question. If CD quality is the best that you are going to be able to get. Do you recommend ripping music from a physical CD or is downloading from a site like HD tracks or Qobuz just as good? Reason I ask is I didn't know if they did any signal processing or funky things like that. I saw an article where they compared an analog signal, vs the CD, vs a newer CD, vs a download magicvinyldigital.net/2022/06/18/dire-straits-money-for-nothing-review-lp-cd-qobuz-1988-remastered-2022/ One interesting article I found last night was this NPR article where you can play a lossless .wav track against a compressed 320kbps and a compressed 128 kbps file of the same track. and you had to pick which one is the lossless file. What was interesting was that majority of the files I selected the lossless version of the track, but for the 2 of the 6 that I selected the compressed file, I selected the 128kbps track. I would have thought I would have selected the 320kbps. its an interesting experiment. www.npr.org/sections/therecord/2015/06/02/411473508/how-well-can-you-hear-audio-quality
Totally agree with you on all that. In terms of CD quality (digital) vs CD RIP, there shouldn't be any difference assuming you have a good quality ripper. All things being equal they should basically be the same. I personally just buy digitally on 7Digital, Qobuz, and very often on iTunes (for their mastered for iTunes albums). iTunes is VBR 256 minimum in the highly efficient AAC Codec (which is better than MP3), the lowest part of any song is 256kbps due to the VBR, but in more complex parts of the song it can exceed 400kbps. In my opinion iTunes rips are actually very good. They're not a good archival format in case you needed to convert later multiple times, but for straight listening 256kbps (VBR) AAC is way more than good enough. I stop being able to hear the difference at a constant bit rate around 200kbps, often much lower. For a good VBR I could do with an even lower minimum closer to 160. So don't be afraid of iTunes in case you're on the Mac ecosystem and were wondering if they're any good or not, their mastered for iTunes albums are second to none. Outside of iTunes I really like 7Digital because their selection is fairly wide, prices are good, and they offer 320kbps AAC with the base purchase (usually $9 for an album). Qobuz is great as well, top notch quality, but prices are a little bit high. I've done those experiments on NPR as well as many others online, and it's amazing how minor the differences are. The site I recommend to test your hearing and knowledge is: abx.digitalfeed.net/ It takes a minute to understand how it works, start with the lowest quality shootout first (96kbps) to help gauge things. And tell me what you come up with as I'm curious! Sometimes lower bitrates actually sound better, which is a bit of a weird phenom.
I don't think albums sold at higher sample rates are upsampled. They are probably recorded, mixed and mastered at 24/96 or 24/192 and then downsampled to "consumer-grade quality" 16/44.1 for CD and streaming. Does not mean it sounds worse, I'm with you with this. I can't hear any difference on streaming services that offer "lossless, better than CD quality".
Some are, some aren't. An example is Coldplay's album X&Y. I read a technical document years ago from the lead engineer, and he said it was all recorded at 44.1khz on an RME interface (which is not abnormal in studios, especially back then). But yet I see the album commonly available in 192khz. So 100% that album was just upscaled. This is only one example, and there are many more. I also personally worked on an album for a Christian band years ago where we did all of it at 44.1, and now the album is magically available in "hi-res 192khz". One could debate whether this should be considered consumer fraud or not. Another example is 4k UHD Bluray copies of movies like "Hugo", but Hugo was filmed entirely in 1080p on the Arri Alexa at the time. So obviously it's just a fake cheap upscale. The difference between lossless and lossy is a different argument entirely though, because that's less to do with resolution and bit depth and more to do with the actual casing compression for delivery format. When we are talking about a low quality lossy MP3 vs WAV or FLAC then of course there may be some audible differences, though the differences become basically negligible past 256kbps AAC or 320kbps MP3. Everybody in audio has opinions on these things, but the actual testing is pretty definitive that there is a point of diminishing returns on perceptive quality.
Most older catalog albums were upscaled or digitized from analog recordings. Considering the Universal fire destroyed a surprisingly large amount of analog masters, new issues maybe coming from a digital source at 44.1 kHz. New recordings made today are made at various sample rates from 44 to 192. Plenty of music originally recorded at 44 are being sold today at higher sample rates.
My understanding is that the quality of the recording itself trumps other factors. I like wide soundstage and good bass response among other things... to me it seems that IF it was not in the original recording it is not going to be in my system regardless of any other factors. I have hearing loss and tinnitus on top of it, but I am grateful for still being able to play my guitar and listen to wonderful music. I have to admit I am always searching for the next best thing... but right now I do have some decent equipment to play through and hear through. What I really need is a brain upgrade. ;-)
Totally agreed. Recording quality is by far the most important factor, it's a real shame that modern mastering ruins so many wonderful mixes. But since we can't do much to fix the loudness wars of today the next best thing is to try to appreciate older music with better gear, haha :). Thanks for watching and your nice comment, happy listening!
@@JoshChristiane My pleasure Josh. I am looking forward to more content from you and I really enjoyed your presentation and demeanor. God bless you and all you love.
Hello Brian ... Something to keep in mind. The "soundstage" and all the attributes of openness and detail that we all love are actually burned into the recordings at the studio by Panning, Phasing and Delay techniques. It does not come from our home equipment. That said... while our stereos cannot create or enhance "Soundstage"... poor speaker placement or room modes can and do tend to sabotage it.
@@Douglas_Blake_579 Hi Douglas and thanks for that great insight! I think sometimes that I am guilty of trying to get something out of the music that was never in there in the first place! Even the best live performances are not always perfect but they sure are fun!
@@BrianVallotton There's no such thing as a perfect recording. What varies the most is their distance from perfection... some get close, most don't. That search for "something extra" is this hobby's Achiles heel. As a service tech I saw plenty of really bizarre stuff... everything from $5,000 power cords to interconnects so heavy they broke the connectors on the back of some truly expensive gear... all in the name of better sound.
So we talked about those "snapshots" which are the captures of velocity that map out a waveform to represent audio. How fast those snapshots happen is the sample rate, but how many bits are used to capture each individual snapshot is the bit depth. The more bits you have, the greater the resolute range of data on that vertical axis. Basically think about it as the sample rate being the horizontal resolution, and the bit depth being the vertical resolution. Hopefully that makes sense, thanks for the great question!
@@JoshChristiane Right, but even if the frequenzy does not benefit from being beyond 44khz, surely a higher bit depth will more accurately reproduce the true analog waveform?
Yes, more accurately. But again we run into the same situation where we have to ask "how much is enough"? Can a human ear tell the difference between 16bit and 32bit depth? The answer of course is no, outside of extremely unusual circumstances. Higher bit depths should be used for recording because it allows effectively more dynamic range in post production, but after mixing and mastering the dynamic range is so decremented that difference is completely negligible. CD quality is still the best the human ear can hear, 16 bit 44.1kHz. While it can easily get better than that, we simply cannot hear the difference.
@@JoshChristiane Yes, technology develops for the reason that these creators of technology also want bread on the table. Unfortunately, technology develops faster than human evolution allows. Obviously, we need to start producing products that allow humans to hear as well as dogs.lol
Ya know, we kind of settled this argument back when CDs came out and we endured the CD vs vinyl wars. The specification for music CDs was actually simple and brilliant. Today I'm a huge fan of the FLAC format. A significant reduction from .wav file sizes and in my most discerning listening using IEMs, I just can't imagine I'm missing any detail. Now if we could get lossless compression for video. Movies are the storage gluttons on the server.
There are a lot bigger grounds to be made up in video for sure. If there are any audible differences between 44.1 to 192 for example, it's so small it doesn't matter. When talking about streaming video, compression types, and even just editing or refining techniques (denoising, glossing, etc) these differences matter significantly. Look at how bad 4k streaming is, but yet how far it's come in 10 years. HEVC has been a pretty big help for me storing all of my videos, but even then I feel like it has a long way to go. Thanks for watching and commenting :)
That's an easy one to answer: Vinyl. Vinyl is not a scam assuming that the album was recorded on real tape (like Fleetwood Mac, Lez Zepp, etc.), and then printed from that tape onto vinyl records. The quality of of a real vinyl print is vastly superior to anything digital could ever reproduce. Way higher dynamic range, lovely sounding smooth analog distortion when you exceed the red line, and true-to-life representation of what was recorded. However, with that said, vinyl offers absolutely no benefits over regular CD's if the album was not recorded on tape. Meaning that any album recorded after year 2000 is basically digital anyways, and going from digital to analog serves no benefits. So yes to vinyl, but only for old albums that were on tape originally. Most of which are better than modern music anyways.
@@JoshChristianeThat’s how I draw the line. I love all types of media: vinyl, CDs and streaming but I don’t go newer than 1985 for vinyl, I don’t see the point.
Im a bit confused by this. When you reference our ears not hearing past 20khz versus sample rates and bits, aren't you talking about 2 different things? Hearing high-pitched frequencies (20 khz) isn't the same thing as a digital recreation of a sine wave.
Sorry about any confusion, I'll try to clear it up here. The two are interlinked. If a sinusoidal wave pattern transitions between polarities faster than a computer is digitally taking "snapshots" then it can't resolve that information. It literally can't capture a wave that's moving (transitioning) faster than the digital captures. Think about it like a movie camera. If you're shooting 24fps and you try to film a bullet exiting a gun, you're very unlikely to capture the bullet. You need more frames per second to capture that fast moving object. But if you record at 3000 frames per second you could capture the bullet, and even slow that footage down to see the bullet path. This is the same analogy with audio. A 20kHz wave is a very fast wave, the only way to capture that information is to have enough resolution (snapshots of data) per second to actually store and replicate that audio. Hopefully that clarifies the link between recording rate and audio frequencies.
The quality really depends on the recording, from tracking to mastering. The delivery in 16/44 is more than enough! So get the CD, make a FLAC and be happy...
You're exactly right. The source material is so much more important. The technological debate does matter because it's how we progress and build new tech in the future, but if you're a musician or audiophile recording music then by FAR the most important thing is tracking, mixing, and mastering. And I could argue the musical performance itself trumps all else.
Many years ago I read about a study where they added synthetic harmonics I think it was 50hz to regular AM radio and the result was apparently people could hear the sound more clearly. then Pioneer brought out CD players that created high sound frequencies artificially
Live Engineer here with a background in recording. I am currently running my Yamaha Rivage PM3 at 24 bit 96kHz and I can hear a huge difference over my old M7CL that ran in 16 bit 48kHz. That being said, I think it's laughable when someone spends $1000 on a USB cable and swears they can hear a difference in their 1s and 0s being transferred to their DAC.
Hi, I uploaded a video to my channel explaining exactly why there's no audible difference between hi-res and CD quality. In a nutshell, when you subtract the CD-quality signal from the hi-res version, it cancels out completely. There's nothing left in the hi-res signal that you can hear - CD quality represents all the musical detail just fine.
Frequencies you can't hear still shape the frequencies you can hear. This is how we use new tech like phase steering to manipulate the direction of sound and create dead zones in isles while giving every seat in the house a perfect stereo mix. Anyway, not going to argue that you need anything higher than CD quality while listening to recordings... but it makes a massive difference live when you are working on top end gear.
@@BionicBurke No, those ultrasonics really have no effect on what you hear, unless there is intermodulation distortion. If your speakers are rated that high, it can be avoided, but if not it's best to just filter them out (which a lot of speakers do).
That's really cool! I was also an FOH engineer for festivals back in the day, who knows maybe we even met each other. The difference you are hearing could be other factors though, right? The new Yamaha Rivage PM3 could just be a lot better converter than the previous. Even the same converter will use different IC's at different resolution sample rates, so just using one converter is also not a scientific test as many think it is. A good converter will use entirely separate IC chains, separate word clocks, different aliasing and filters. So just because a converter sounds better at 96kHz does not necessarily mean it's the sample resolution itself. I mentioned this in other comments but I have a custom Burl converter that sounds better at 96kHz than it does at 192kHz, and I have older Hollywood Converters (the best in the world) that sound better at 44.1kHz than they do at any other rate. So there are many factors. When the testing is done purely scientifically using a singular conversion chain the studies I saw showed that professionals and audiophiles can't tell a difference. They could however tell a difference on many high-end brand converters in unscientific settings. I'm just saying it's possible the difference you're hearing is the converter, not the resolution.
Mostly marketing, but not entirely. Super Disks and DSD's were awesome for a lot of reasons, as they could store higher quality music. The ONLY time those disks are legitimately better though is when you are printing unmastered recorded performances. So for example if you're recording an extremely dynamic instrument like piano that has a performance that goes very very very quiet, then becomes very very loud. That type of performance is notoriously difficult to capture all of its detail. But almost all music today is mastered and its dynamic range is intentionally reduced for easier listening, because people don't want to jump for the volume knob constantly. A little compression is good, a lot of compression is bad and most modern masters use far too much to compete in the "loudness wars" which in turn destroys music. So I like the idea of a higher quality disk for archival reasons, but you won't hear a difference outside of extremely unique situations that are rarely ever present.
How about instead of the endless argument about formats audio engineers and audiophiles and the community in general push for a simple designation standard of recorded music. Like SCA the coffee brewing association has. A brewed cup of coffee has to have the following elements, temp, grind size etc. So if an album is recorded at say the proper bit rate and depth and filtered properly or whatever the criteria should be for high quality sound. ***** star it and let the buying public know how it was recorded and mastered rather than if its Hi-Res ,etc etc etc. When I stream an album that sounds great no matter the resolution I have taken to make h it as I love symbol most services have. And only use that designation for that purpose? Just a thought?
Totally agree. Mastering is the greatest weakness in the industry right now as well, the quality ranges from "okay" to incredibly bad. We need to produce an environment where music isn't being destroyed to compress the heck out of everything just to make it louder.
Its not a scam in the sense of buying high end headphones and pairing it with a high end setup and then using that to play a loseless audio file for its higher quality as well. Its not all about what spectrum of frequencies we can hear. We just want something that sounds more clear and detailed to our ears than the norm. Is it overkill? Possibly. Does it sound better? Absolutely.
Absolutely. Buying a good pair of high quality headphones, a good headphone amp, and playing back lossless files is definitely a great way to enjoy music. That aspect is definitely not a scam, I am more just talking about 44.1khz lossless vs 96khz and 192khz lossless here. You have to pay like $10 more for the higher sample rate on a lot of albums, and you gain no quality.
@@JoshChristiane just watched the full video 😂 I have to agree there. When my dac switches between the different khz I can't tell the difference in quality.
Yeah same here, nobody can tell the difference unless there are some really specific conditions. If you slowed the music down 300 or 400% then maybe differences would start to become apparent in the high-end, but at that point the audio quality would be so messed up it wouldn't matter anymore. Some converters sound better at 192khz which deceives a lot of people, but that's just the quality of the conversion itself not the frequency range difference.
Yes it is. external Headphone amps/dacs are generally nonsense as well. Most mobile devices has sufficient audio processing unless the drivers specifically need amplification which most consumer grade units don't. Lossless is a subjective basket being that music production sometimes derives sounds from less than stellar sources. I.E Where was the instrument from a rompler sampled from? at what rate? Was a sample pulled from Vinyl, or an obscure, out of production album that someone uploaded to youtube? are the headphones neutral monitors, or capable of being adjusted to be so? The Audiophile experience is often placebo, namely because most audiophiles don't understand the music production process in the first place.
I'm old, so I remember when TVs were nothing but CRT screens, and they had a flyback oscillator that me, as a child, could hear running at 15.7 KHz. If I wanted to know if my folks were still awake I would open the door the the TV room and hear the whistle. Now well into my 50s I can hear maybe 10 KHz tops, and the funny thing is it seems mostly old guys my age (but I suspect with more money) want to extend the response out to over 20 KHz. Forget about it.
That's a great example of how hearing changes over time and can be influenced by environmental factors, and it's a real bummer how some of the frequencies we could hear as children fade away as we age! The youth truly is wasted on the young.
It's a bigger topic than I can cover in one video, but you're definitely making a great suggestion! I'll make sure to do a video on SACD sometime in the future, thanks for watching!
Actually CD quality has greater bandwidth than human hearing. Pat Brown did an experiment ("Is CD Quality Good Enough?") where he tested dozens of audio engineers and nobody - not ONE subject - was able to hear a difference between standard CD quality and hi-res. I think that says something!
@@nicholassheffo5723 YT comment filtering won't let me post the link, but it's at ProSoundTraining, May 31, 2014 by Pat Brown. Title is "Is 'CD Quality' Good Enough?"
FLAC and ALAC (Apple's version of it) are both compressed lossless codecs. A codec means "compression-decompression". So both are file formats that compress and then decompress the audio on the fly. The reason they're popular is because you can keep the lossless quality of WAV or AIFF, but in a smaller file format with files that are often half the size. FLAC (and ALAC) are fantastic, but they have nothing to do with the quality of the audio itself, they just store the audio. Think about FLAC, ALAC, WAV, AIFF, and any other lossless format as all containers that hold the thing itself, but the qualities of that thing are not necessarily entirely determined by the container. MP3, AAC, and OGG are lossy formats so they will reduce the quality of the sound they store. FLAC supports 16bit, 24bit, and 32bit audio (for bit depth), and it supports 44.1khz all the way up to 600,000khz+, so the range of what it can store is basically infinite. I'm actually surprised the format doesn't get used more often than it does, as it's my favorite. Assuming you're building a music catalogue or collection, I highly recommend storing all of your files in 44.1khz 16-bit, FLAC or ALAC with full compression (8) enabled.
Question: it’s it worthwhile recording from yt music for playback in public space for background music vs just recording the yt audio/video version of it?
I think that answer depends on exactly what you're doing with it. If you were a DJ just downloading songs from TH-cam Music for playback then I don't see an issue with it, the quality isn't great but it's not horrible either. But when you're streaming it from TH-cam the quality is actually significantly lower. You have to google what quality they stream at, but I remember it being low. Depends on the event, but for most situations it probably won't be a big deal.
Great video, thank you. Although your last statement about collecting music in CD Flac file format, I disagree for old music the best is to put together a vinyl collection of original first release records. 😊
Thank you so much for watching and commenting. I do agree that vinyl is awesome for old music, especially music recorded on tape. Nothing better than the real thing :)
And then Philips and sony brought out SACD. which was not PCM based at all but DSD although many albums apparently converted from PCM. others opted for DVD audio instead. but it went to the grave. although you can find some discs still
DSD was really cool, especially because the masters for it were particularly well done. A big reason for this was they knew the target market were a more caring and sophisticated listener. Wish they'd bring it back to popularization!
I don't know about your ears but I listen to all my CD music as 128kbpsMP3's on a Windows PC connected to an AV Receiver at 24bit88.2 Studio Quality. Compression or weak MP3 format or not, the music is light years sounding better than the same CD I ripped from into the PC DVD-ROM drive or through a direct 4kBD player. Infact it sounds too rich that if I play it even higher than 24bit96 the audio feed from the PC will brick the high end AVR. I can agree that HighRes Audio was developed for hardware. Even the HEVC DD48k 5.1 is much richer than through a BD player just because Windows is configured backend at 88.2. In saying that Cold Play knows more about sound recording quality is like saying Doctors know about medical technologies and chemistry, they don't... tech companies and pharmaceuticals create medicines and MRI tech.
what about dsd? i have a lot of thoughts on dsd i have no idea if it sounds better or more faithfully recreates live music in large part because files recorded in dsd often are kinda niche auidiophile type music and that type of music is loaded to sound good coming out of a stereo. look for a well known track and it would be a master tape transfer to dsd which is unlikely to yield benefit over cd which is the best format imo. the way dsd is implemented in theory more accurately recreates a sine wave as well, as opposed to the 'little steps' approach of more traditional digital formats. also i wish he would tell people that vinyl is not a thing that would make a great video. everything here is more true of vinyl
Vinyl is awesome when it's real. The problem is most vinyl today comes from digital prints, so it's basically pointless lol. The entire point of vinyl originally was to convert recordings from tape into an actual analog product without ever having to digitize it. This is technically better and worse than digital in a myriad of ways, but mostly it sounds better and warmer to our ears. As far as DSD goes it's not actually better in terms of technology, but the masters were better leading everybody to think the format was better. There is a good article on the technology that I recommend reading here: www.mojo-audio.com/blog/dsd-vs-pcm-myth-vs-truth/
I agree with every word you've said. It's a huge marketing ploy but unfortunately the general public has fallen for it hook line and sinker. It's become an urban myth and trying to convince people otherwise is like whistling in the wind. Keep up the good work. 👍😢
There are those who are critical of sound quality when it comes to streaming. these people think, for example, that streaming in their own network or playback from physical media sounds better
Hopefully, everyone is aware that the term "scam" can apply to anything. It's basically a person's opinion of something. Taxes can be a scam. Airplanes can be a scam. Greek literature can be a scam.
This is an excellent point. But then almost everything on the internet is an opinion anyways, outside of scientific research. But even research can be heavily biased by opinions, nothing is for certain but death.
I got so aggravated with your video I only made it about half way through. Lol Hopefully you explained that Hi Res is a standard and that 99.9% of people shouldn't care about it. It's really only for people who are building a full Hi Res system. Personally, I'm more than happy with a non Hi Res system.
Thank you for this video. I totally understand the math and logic behind these numbers - yet the reality looks a bit different. Probably for completely other reasons, as there are many more factors involved than just these numbers. At least in the few examples and tests I have done myself, HiRes can _definitely_ sound much better than a CD. Even though technically it should not. I own a few albums both on CD as well as in a 96kHz/24bit FLAC version from an online HiRes audio file seller. With some of these albums I cannot make out any difference between the CD version and the "HiRes" audio version. In those cases they clearly sold merely an upsampled version of the CD. And I agree, a total scam. However, I _do_ have a few albums where the HiRes version _definitely_ has a lot more fine detail, that is nearly muffled or sounds totally flat, like the instruments being recorded under a felt hat or something, on the CD version in comparison. Maybe the CD version was poorly engineered? Or despite being brand-new and having no scratches the error correction destroyed the sound due to a poor CD pressing? I doubt my CD player is to blame as other CDs do sound exactly like their HiRes audio file version. So I don't know the exact reasons, but I _can_ hear a clear difference. If as a consumer my only choices for a certain album are either a CD version or a HiRes FLAC audio file at 96kHz/24bit - with no 44.1kHz FLAC version sold - then the HiRes version _can_ sound much better than the CD version, and hence would still be worth buying. Despite the pure numbers not making any sense. Perhaps the difference then is not the kHz but the bit resolution. 44.1kHz recordings are often done at 16 bit, while 48kHz or 96kHz typically use 24 bit. If a 44.1kHz/16bit recording is upsampled to 96kHz/24bit it will not sound any better, and likely even worse due to the resampling. However, if the original source was a 48kHz/24bit recording, the CD at 44.1kHz/16bit might indeed sound a bit more muffled and flatter than a HiRes 48kHz/24bit FLAC file - even if that one was upsampled to 96kHz/24bit. That is the only explanation I can think of, for the clear difference I hear in some HiRes audio files, compared to "CD quality".
Great comment. So if you're hearing a difference between those sampling rates it's not because of the rates themselves, as you stated, it's from other factors. Some converters just operate better at 192kHz than they do at 44.1kHz. Some word clocks operate better at higher rates, some operate better at lower rates. There are least a dozen factors that can make a higher (or lower) sampling rate sound better than another, and it's notoriously difficult to get files for blind testing on this topic because of those differences in filtering and aliasing. However, when the tests are done purely scientifically with extreme care to get exactly matching files, people cannot tell the difference audibly between 44.1 and 192. So the key here is making sure your recordings are tracked well, mixed well, and mastered well, because the sample rate won't matter much in the end if at all. Thanks for your great comment, I read everything you wrote and thought it was very well written.
@@JoshChristiane To make matters worse, HiRes audio file vendors never disclose any technical details. What was their original source? A 192kHz/24bit file? Or merely a 44.1kHz/16bit file? What was the hardware and software used to upsample or downsample these files? At the end of the day they make you buy the proverbial cat in the bag. It's a lottery game. As we do not have enough data to make a truly informed decision, we are left with having to buy HiRes files hoping for the best. That for whatever reason they may - or may not - sound better than the CD. And like with the lottery, sometimes we win, and indeed the HiRes files sound crisper with more detail than the (now comparatively) crummy CD. And sometimes we lose and both the HiRes and CD version sound exactly the same. And it was all a waste of money. Lastly I would also like to bring this item into the discussion: As a fan of Howard Shore's soundtrack for The Lord of the Rings, I have three versions of it: the CD version, the Audio-DVD and the Audio-Blu-ray version. The dynamic range and clearness of the Audio-DVD version compared to the CD version is truly stunning! It literally sounds 10x better than the CD. With the CD it sounds like you are sitting in the auditorium of a concert hall, listening to the orchestra in the distance. Yet with the Audio-DVD version you are literally sitting _inside_ the orchestra, amongst the musicians! And surprisingly, the Audio-Blu-ray version yet again sounds noticeably better than the Audio-DVD version! Perhaps not another 10x better, but still noticeably better again! To continue the earlier analogy, now it seems you are actually sitting on an orchestra player's lap - even hearing him breathing. The clarity of the sound is just amazing! So why is that? Is it merely a better remastering? Yet they are all ultimately from the very same original source. Or is it better mastering equipment? Possible, as these three were released many years apart from each other. Or is it perhaps indeed somehow related to the higher bitrates and sampling frequencies, with the CD at 44.1kHz/16bit, this Audio-DVD at 48kHz/24bit, and this Audio-Blu-ray at 96kHz/24bit? Frequency-wise according to the math all three versions should sound exactly identical. Yet they are miles apart from each other! So in the end, in reality, it is still _very_ _well_ worth it to buy HiRes audio versions. In some cases they really _do_ sound amazingly better - for whatever reason... Yet every time we buy them, it is a lottery game. It may sound amazingly better - or not at all better than the CD.
People get so transfixed with Hi Res, this is not the most important thing. The one thing imho that makes a track sound great is the quality of the recording you are listening to. For years i was a subscriber to Qobuz listening to most tracks at 24 bit 192 khz through my HD600s and Chord Mojo and yes most tracks sounded great, but a lot did not and some music videos on TH-cam sounded better than those badly recorded tracks. For the last 6 months I have been listening to Spotify through my Sony XM5s through Bluetooth LDAC. I have Spotify set to Normal best streaming quality with Dolby Atmos turned off. Through my Samsung smartphone i have the HD audio turned on in the Developer Options playing through 16 Bit, 92khz at 660 kbps. In the Sony Sound Connect App I set my favourite EQ settings to Custom 2. I am telling you honestly that listening to Spotify this way the sound is just as good as when listening to Qobuz through more expensive equipment 🤷♂️
Absolutely 100% correct. The quality of the mixing and mastering is so much more important that the entire topic is bordering on pointless. Even if you're listening to super low quality MP3's at 128kbps, a good mix will still sound vastly better than a bad mix, lol. So by far the most important thing is the mixing and mastering.
Hi-Res allows for less steep filters, therefore less damage to the audio signal. The filters are not only applied in DA conversion, but also in the studio‘s AD conversion. Gentler filters = better audio. Studies have shown that the human hearing resolution is ca. 7 microseconds. You claim that more than 44.1 kHz sampling frequency is not necessary. However, CD quality only means a time resolution of 20 µs. Only at 192 KHz sampling do we get down to 5 µs. From this I conclude that Hi-Res can make sense over this point of view. Granted, not every listener will have that ability, but prof. musicians and conductors with trained listening may have.
I hear what you're saying, but plenty of studies and blind tests have been conducted during research on this topic. Every single one I've ever read concluded that professionals and amateurs alike could not tell a difference. I'm not saying there are not differences, of course there are... But they're just no longer noticeable. People mention filtering and such, but mostly that's based on old information. Modern converters work quite differently to produce extremely high quality filtering even at low rates. It's so good today that I doubt any living person could tell a difference in a blind test even between low-quality modern converters.
@ Sadly, you offer a totally uninformed opinion. Rather than try to learn about and understand what may be different in High Res, you wildly speculate about something with insufficient understanding. But you are young and still have time to catch up.
@ actually the only digital of the whole thing is the signal, which is decoded by tuner. Same thing with color antennas, no such thing, the signal is using the same spectrum bands.
yeah but i still have noise wenn i rip a cd to alac or wav. I looked at spectrograms of my alacs and they all had noise. i didnt convert anything i just converted it from one format to another. i converted one cd to flac. still noise in the spectrogram
That's often from the CD press itself. I used to make my own CD's (from actual bounced prints) and depending on how you do it, it can cause noise. Dithering adds noise intentionally to help reduce bit error noise when going from 24bit to 16bit for CD standards, so it's possible that's also what you're hearing. This is why I buy digitally (7Digitial, iTunes, Qobuz) then download 44.1 16 bit CD quality FLAC or ALAC for storage. Nothing wrong with physical CD's though, which ripper are you using? I just use Foobar2000 and have had good results from it.
My biggest concern and complaint is mastering itself, not any file format. Nothing destroys dynamic range today more than the loudness wars, we need to find a solution to that.
@@JoshChristianeI went back to CDs for this reason as I got better equipment, a lot of the masters on streaming are using the brickwalled remasters. An 80s CD with proper dynamic range sounds better to me than a poorly remastered 96khz-192khz lossless file.
I don't know much about the current high-res players but I want to make the sure the audio file is at least the same SCD quality and I want to make sure the player and whatever I'm listening on can reproduce The full quality of the recording and I'm listening to you I mean that's what's important here. I would hope that a music player that's listed as high res can play back a audio file and it's full quality now as far as headphones being high res I don't know about that, I can understand a music file and an audio player being high res but headphones I mean you got awesome headphones decent headphones and sucky headphones so if you get a set of really good headphones it should be order do the job, I don't know if I've ever seen a set of headphones labeled high res but any truly good to the headphones should be able to play back a high-res file and make it sound awesome. But 44.1 versus 48 You're probably not going to hear much of a difference, 192 versus 320, you're probably not going to hear much of a difference so I get what this video is about I think he's putting out a good message here
You are right, it is all about the chain and it doesn't really matter if a device says "Hi-Res" if it can't actually reproduce that quality. In the end the source material and how it's processed (mixed, mastered) will matter by far the most.
"Hi-Res" is mostly Scam... What ? As a speaker builder, musician, mixing and mastering artist i can say. That most modern music is only 12 Bit, after the mastering stage... Cause 1Bit is 6dB in dynamics... So 16 Bit is 96 of Dynamics. The lowest ( level ) tone in a song is around -35 dBFS and fades are down to -60 / -70 dBFS... The highest ( level ) tones are going up to -5 dBFS... So you have a dynamic range of 12 Bits = 72 dB... Classic music, real classic music without compression on.. so with the brutal original dynamics... can benefit from High Resolution Media... Okay kHz... 44,1 vs 192... My experiences are... the higher the kHz, the more unnatural the sound is... it gets more and more clinical sterile... So 16/44 is good enough for me... WHEN the mastering was a good one... AND THAT is the biggest problem today... over 80 to 90% of music out there... today... is sonic garbage ! I know, when i hear a modern produced song, what went wrong in point of physics, psychological and technical points... beside the stupid loudness of today... Cause of this mess... Vinyl is at moment, although it is the worse format... the better choice over Hi-Res... Why ? Cause the cutting engineer need tracks which are less loud than the CD / Streaming versions... Cause of that... dynamics came back and more of the original mix sound... Cause for CD and Streaming, the life is squeezed out of the mix, to BE THE LOUDEST ON THE STREET !!! A complete stupid mind set in the biz... Why ? Cause near everywhere audio is normalized... Even on CD... cause you are the normalization algorithm with you hands... Cause you dial the Volume knob of your amp... 16/44 on CD, when the Mastering was a good one, is fine and can sound warm like Vinyl... without the technical restrictions of Vinyl... I hope this helps a bit :-)
Totally agreed with most of this. This was a good analysis with some excellent overall points made. I know it's a nuanced topic, and many confuse converter quality with format quality, so that alone can be deceptive. But in the end quality can only get so good before something sounds real, of course there will be a diminish of returns.
There are formats that have even higher sampling rates. and there the idea is that it is precisely the number of all of them that gives higher sound quality. and also maybe the idea is that it should sound more natural or analog two ex DXD 352.8 kHz DSD DSD 64. 2.822 MHz
FYI In respect of pixels on a camera phone they just double up and multiply the resolution as you realistically are probably getting half meg on a so called typical 40 meg camera. It's impossible to obtain these resolutions from a small lense regardless of the sensor, it's just marketing hype as usual 🙃 As for the Hires yes and no. The main reason for up sampling is to add more ingredients in the music designed to then be digitally enhanced in the detailing specific to the quality of the speaker or headphones etc. I do agree though in normal daily hearing frequencies 16 bit is ample. Thanks 👍
Thanks for the comment! Totally right on about cameras though, adding more pixels absolutely does not give you more "real resolution" unless the lens can actually resolve that information. But for example on a lot of small smartphones with cheap glass that's not even possible, so definitely some heavy marketing at play there.
I do a simple test with my focal monitoring speaker and switch to youtube from Qobuz and between the AAC from youtube and the FLAC there especialy on the same album with the same master source the FLAC win hand on whaterver resolution it is especialy with a 48 khz aac 16 bit and the FLAC 48 khz 24 bit , so ok if you don't compare it's Impossibles to listen to what you realy loose , but on cymbals , high pitch and medium frequencies you loose a lot, FLAC 16 bit or 24 is more than ok at 44,1 khz the sampling rate is not that important but the bit use to digitized is more relevant and squeeze all that to lossy codec make thing worse AAC and MP3 is like compare a PNG to GIF there is some lose of color .
Thanks so much for watching. Don't forget to subscribe and be my friend on X at x.com/Josh_Christiane
Subscribed 🙂
Thanks, Roman! :D!
Hi Josh , Apple does stream @ high resolution above 48 kHz via a USB connection & airplay 2 @ 24 bit depth 48khz brick wall. I do agree with you but only for non studio playback. I’m subscribing keep up the great work 👍.
@gg.6967 Agreed, thanks!
Modern vinyl & cassettes are even worse. And the horrible loudness wars. The whole entertainment industry has become one big scam, they don't care about quality, just making fast money.
Hi-res audio is not a scam.This video focuses on frequency rate and while I generally agree high frequency rates can be wasteful, the audible benefit of "hi res audio" is playback at the file's original high bit depth. Recordings almost invariably start at 24bit (or 32bit) - and there are subtle differences between the original and the dithered 16bit file, most noticeably when comparing in low level signals such as less detail in reverb trails and reduced spacial effects in stereo when comparing the original 24bit recording to a dithered 16bit file. Good equipment in a good room or good headphones are needed to hear the difference, but it is there.
Completely agree with you. It's a big topic and almost deserves its own discussion separately but you're completely right when we talk about recording, mixing, mastering, etc.
After music has been mastered is a different story, since most modern music is SO heavily compressed at master that even 12 bits is more than enough to represent such limited dynamic range. Dithering definitely does affect sound and quality, and one could make a debate for (and against) that as well. I've known some who didn't want to dither, so they just recorded in 16 bit from the beginning to avoid it entirely and believed they were getting better masters as a result.
I can hear a difference but can’t explain why
There ARE slight audible differences, even if you can't hear past 15kHz because of time domain and phase distortions created by steep anti-aliasing filters. Agree that other factors are way more important though
Yes, I have heard that argument, and it may very well be true. But as a general rule the quality of the converters will still matter drastically more. I personally have never been able to tell the difference, and haven't seen any studies that show people are able to either. Maybe a few individuals just have golden ears, who knows.
Some technical people said that you cant distinguish betwean aac and wav sound, but on ipod i can hear difference
@nuznikas Maybe the reason you're hearing a difference is because Apple Music uses AAC?
They may have intentionally made other formats sound not quite as good. That would definitely be incentive to get people to use Apple Music.
@@JasonWW2000maybe not making it sound worse but wouldn’t be surprised if they applied some sort of DSP effect on their own music so it tricks you into thinking it sounds better.
Designing CD players in the 1990s the digital brick-wall filter was a separate chip to the oversampled DAC. I was surprised how different different filters sounded in listening tests and part of the design process was to make up a few PCBs with different filters to find which one sounded the best.
I don't think the sample rate itself is the issue, but increasing it past 44.1KHz does open the possibility for less of a "brick wall" filter and a more gradual transition between the pass band and the stop band, if that is the issue. A perfect brick wall filter, i.e. no transistion band width, will ring indefintely,
The question, "does a higher sample rate sound better", maybe the wrong question if it is the digitial filter that is potentially affecting the perceived sound quality. Simply doubling the sample rate and doubling the filter clock won't fix that and will probably sound just as good/bad. The filter has to take advantage of the extra sample rate and extend the transition band. Not sure if any oversampling filter/DAC chips today even do that.
Terrific video on this subject. One of the very best I've seen so far. Would love to have you make a video on your thoughts on the SACD/DSD audio format.
Thanks for the compliment, it made my day! :)
I will make more A/V videos in the future, especially focusing on things like converters, formats, etc. I'll try to squeeze that in there somewhere because it's a really interesting part of technology history. DSD vs PCM is really interesting too, that might be a good subject for a video.
@@JoshChristiane Agreed, 100%. Thank you!
@@michaelbeckerman7532 DSD was another breed , technicly it's a Delta sigma brut force RAW digital recording that not using sampling rate but a 1 bit PWM digital flux with 2 megaherz or 5 megaherz bandwith to sound almost like a 96 kHz 24 bit, but technicly it's not PCM data , this one use another Delta sigma convertion with sampling, all of the high res Wav and FLAC or the lossy format AAC, MP3 or surround DTS Dolby AC3 and atmos are derivative or based on PCM for recording and reproduction. With DSD you need that Big bandwith because it's a simplified digitized process, there is a big of digital noise that was filtered beyond 50 khz so you can have minimum true 30 kHz or almost 50 kHz or frequency response the sound is also more linear and smooth it can reproduce almost 120 db of Dynamic range but depend of your converter, there is only a few pro device that can record this in true DSD the old Sadie Sonoma system, the Pyramix and also some RME and Prism converter with a Tascam DA 3000 master recorder but it's rare to have true DSD recording you can't apply plug in or DSP on a bit stream like that, so they convert DSD to PCM and also PCM to DSD is possible but it kill the process of listening a true analog tape transfert to a DSD, it's very smooth and without any harshness if you have that it's probably a PCM to DSD convert file, so you have no benefit from against a FLAC or a Wav in 24 bit.
It makes sense for recording wildlife or using for industrial settings such as recording machinery. Pre-recorded files, in hi-res, are also fine for people that sample. The only problem with any higher def, audio or video, is what resolution the original was done in.
Absolutely true. Especially wildlife, I'd want the most dynamic range and best preamps possible for something so intricate and dynamic.
I'm curious, why is this only about hi res delivering a broader analogue frequency range? How does a higher sample rate impact detail by filling in more of the analogue shape and reducing averaging? Or how it can work with dac variables?
While it's not completely technically accurate I like to explain it in a visual way that makes sense to people. Think about sample resolution as the horizontal resolution, and bit depth as the vertical resolution. Both matter, but in different ways. One is "snapshots" or "steps" of the frequency shape, and the other is snapshots and steps tracking amplitude. When you think about it visually like this it suddenly becomes a lot easier to make sense of all of it.
@JoshChristiane that's where I personally don't understand. Because if there are more bits in a slice, the horizontal view becomes smoother and less jagged so I'd expect the detail to be more prevalent, no? With less bits per slice and more jaggedness, I'd expect the conversion to analog would fill less accurately with the less info given.
@@baaltroth1957 It does exactly what you're saying but human ears do not detect it. It's like looking at a monitor that doesn't flicker in our eyes because we can't pick it up but it's still happening. Same as the 4k to 12k argument.
@MrBigspendn cool. I got it now. Now then for dacs, if the bit rate is higher aka sending more detail, then the dac's performance could be better so some of the detail like piano frame resonance after releasing dampers could be fuller and more lifelike in its behavior?
I uploaded a video to my channel explaining exactly why hi-res is no better than CD quality. When you subtract the CD-quality signal from the hi-res version, it cancels out completely, leaving only some inaudible noise. The only thing hi-res can offer is frequencies above 20 kHz, which you can't hear anyway.
I'll go check out your video!
I have read on chose 44.1 due to limitation in the recording equipment that converted beta cam. Philips also wanted 14 Bit instead of 16 bit. but where was the review of the CD sound when it arrived. It was that the sound was flat and sharp treble. Then CD players got over-collection and dacs with more bits, there was also a time with One bit DAC technology
I recently significantly upgraded my whole hi-fi setup. It supports up to 192khz through the whole chain and more than that at the DAC. I had Qobuz for a while, sounded nice, BUT my best sounding music is still on physical CDs (Pat Metheny, Peter Gabriel, James Taylor, Toto, Doobie Brothers, Steely Dan etc.) - I have ripped my entire 10,000+ disc collection to flac with eac and play through plex to my wiim pro (bit perfect transfer to the amp via coax) and it's all up to the amp and speakers - they were perfectly paired with my room and budget (not for everyone) - Arcam SA20 and Dali Oberon 5s - to my 45 year old ears they sound amazing. (yes I could have tripled the cost and got better I'm sure)
One thing of note: I also have a number of recordings in DSD (Super-audio CD) and they blow my mind - BUT again - this is to do with the mastering (NOT necessarily the format). Some standard CDs were badly compressed but the SACDs were mastered properly - if the same mix had gone on standard CD I'm sure it would have sounded amazing.
This is exactly the same reason why some of my vinyl sounds better than CD - nothing to do with the format and all to do with the mastering.
Absolutely agree with you. 99% of the time people think higher rates are better is just because the master was done better on those, and has very little to do with the actual resolution itself. In the same way that photographers with more expensive cameras tend to take better pictures (because they're professionals) so it's a form of survivorship bias. Amateurs think the expensive gear is better since it's used for all of the better photos, but when a good professional uses the cheaper cameras they often get equally as great professional photos. Audio can be the same way, for sure. Guitars, microphones, preamps, etc. You can still do a million dollar recording on a Squier, but professionals just want the best.
Sounds to me like you have a really good audio setup though, a lot of your gear is dream-level gear, which is awesome! Thanks so much for watching and your great comment!
I have used the digital photography analogy myself, but I add an extra bit... When you are upscaling music, you are starting with desired result. When you make a jigsaw puzzle, you start with the desired result. Chopping the jigsaw into twice as many pieces doesn't improve the the quality of the resulting image. To do that, you need to start with a higher quality image.
In music, I have had mixed results listening to "Hi Res". Some tracks sound like they started with an mp3, some sound like real life, but I've not heard many that sound better than cd.
I totally agree. But there are many people who will argue upscaling improves music magically somehow, lol. If it improves it to their ears that just means they don't like the original, because all it can do it make it different, not better.
I think the better analogy is to add IR and UV to digital photography/video (and in turn, TVs and monitors) to capture a higher range of frequencies than a human can perceive as 'better quality'.
You didn't even mention the bit depth and the loudness war
I agree... I'm older but still with pretty good hearing, and I have a CD player with SACD capabilities. Since these discs tend to have regular CD information as well as SACD, I figured out how to toggle between the two. I tested a few SACDs and some of them did sound a bit better set to SACD, but the differences were pretty tiny and on others I couldn't hear much of a differnce at all, which leads me to believe that the differences were likely more in the mastering than the format. Other digital audio standards that are higher bit depth don't make sense to me either as though they improve dynamic range, the dynamic range of a regular CD quality digital file is so great already that the differences aren't likely to be audible, particularly with 98% of music which doesn't even contain nearly that much dynamics. The CD Redbook standard, though very old at this point, is well enough thought out, with enough real understanding of the limits of human hearing that it still holds up. All of this is not to say that improvements can't be made on the hardware end as different DACs do sound different, and some are more resolving and smoother sounding than others. The basic format, though, at least as far as CD quality, is still about as good as it gets, though, in my opinion, at least as far as what we're able to discern...
Great comment! Thanks for sharing all that info, I loved hearing about your tests and experience.
Now a rational argumentation about if Hires worth while or not needs some problem definition: 1) sound isn't quantized nor in time nor in amplitude, both are continuous. 2) the often quoted by everybody Nyquist Shannon theory tell us the minimum requirements you need (note: minimum or enough it will depend) to retain all the signal's information. 3) every practical device has design constraint, you can't have more of one without pay on something else ( difference between engineers and stylists ) and more, practical implementation of a certain theory quite differ from theory . From 1) one can safely assume (ideally) that the more is the better both in terms of sampling rate and bit rate. Especially thinking on AD conversion. 2)&3) Theory was demonstrated mathematically but to take it as an absolute true, technology must have developed something working like this: you need all the sample (all! i.e. if you have a piece of music with timing of 5 minutes all sample from 00:00 to 05:00 min) to DA conversion taking in account and process all of them at the same time. It isn't the way present technology is working indeed, DA conversion act of a stream of samples processing them say one by one (this isn't quite true but the idea is more o less that, if you are thinking about nowadays digital filters you are pretty much on track and yes they need to be quite long, a lot of number to deal with).
So, you just noted that there is gap from what it should be in theory and practical implementation, so till the gap isn't close ... in a way or in an other (going to high rates is a sort of brute force if you will).
Some back thinking:
- CD standard 16b/44.1kHz was adopted accounting for support's space limitation (it was the best was possible to do at that time). Means that is THE STANDARD?
- Take an old CD from '80 play it from a matching class old and new players, why new player sounds better? Cd and it resolution are the same aren't?
- from another piece of industry (multichannel oscilloscope dealing with hi freq. ,some GHz on every channel, set in the range 3X to 4X the theoretical sampling frequency to avoid artifact (signal shape pre and post alteration). But consider most are intended to deal with square waves signals, not audio.
- one point which can mark a no go towards high resolution (actually I mean over 96-192 kHz or 88.2 - 174.6) could be a degradation due to the management of all this data not about the computation limits but the meaning itself. Control theory suggest to identify the dominant time (or equivalently frequency) behavior and limit the control bandwidth to that in order to avoid to follow ghost. Stated in alternate way sure that due to the need to have very long digital filter (isn't the filter long obviously, is the polynomial implementation which is long) we run in a data obsolescence. The risk is to better follow the high spectrum's and loose control at low end. Remember 2) take all samples at the same time is very good (100% on track with theory) stay in the middle there may be problem to deal with.
- last point is something you can judge only based on personal preference and taste. Favorite music is yours favorite, no comment about. But there are some way to listen and some kind of music for which every discussion on Hi or Low resolution does not make any difference. Head phones vs room+loudspeakers have peculiar pro/cons (stage spatial reconstruction is quite limited with
headphones, at most a little bit grater than you skull, and everybody can understand a classical orchestra does not fit in you skull), stated that 1kHz tone is exactly the same (just to recall my previous post). Actual played instrument, you have a reference to compare (go to concert hall) form most of the rest can be everything.
Same for bit depth, a piece of music with no dynamic, more or less at same level for entire duration recorded just at -6 dB from the maximum 0dB can count on all 16 bit (taking the CD case), different is if you need to accommodate a full orchestral peek and in same piece you have a pianissimo, signal in last one can move around first 7 bits just flickering 0 1 on a couple of them and you filter can't prevent you to have something like a stair signal (good quantity of distortion).
The audio snobs with lots of money to spend or lots of money to make wont be happy with these facts...shhhh 😂
What about TIDAL? I notice my music sounds better on there than Amazon Music.
That's because Tidal streams at much higher bitrates. Amazon music is really low bitrate, I think it's below MP3 320kbps, Tidal is significantly higher than that and offers lossless streaming. But that has nothing to do with resolution itself, that's the actual file format they're using for storage. Think like WAV or AIFF vs a Codec format such as low quality Mp3.
it does sound better --- also on qobuz
What i really like is music in surround (either quad, 5.1, 7.1 or any iteration of Dolby Atmos) i have an argument that clarity in recordings (or resolution) is better achieved by spreading a recording through several "pipes" instead of just two (or even one when in mono), what do you think about this argument? I would really like to know your thoughts.......
You make some excellent points and I agree with you. I also love binaural mixes, so anything specifically mixed for headphone, those mixes just have so much dimension and space. I always wish more albums had multiple versions. If it were up to me I'd produce a radio album edit (mastered), then an unmastered edit (only mixed, high dynamic range), then a binaural or surround sound mix. That'd be awesome!
The highest frequency a musical instrument can reach is generally considered to be around 5,000 Hertz (Hz), with some instruments producing even higher harmonics that may be audible to certain individuals, though most of these high frequencies fall outside the typical human hearing range.
Audiologists test a person's hearing at frequencies ranging from 125 Hz to 8000 Hz, with the most common frequencies being 250, 500, 1000, 2000, 3000, 4000, 6000, and 8000 Hz. These frequencies are slightly broader than the range of human speech, which is primarily between 500 and 4000 Hz.
Totally correct. There are harmonic frequencies we can hear in the 15k range, and when I put a high cut filter at 15kHz I can definitely hear a little sparkly high-end missing, so obviously we are hearing something up to that range, but it's not what I'd call the "vital" life of the music. With that said, after about 15kHz-17kHz I hear basically nothing, and our ears are most sensitive to high frequencies between 5k and 8kHz as you mentioned, so that's where a lot of the energy goes.
MP3 basically works this way, by cleverly cutting out frequencies our ears are less sensitive to in order to save space.
@JoshChristiane thanks! I quoted that from Google.
I recently had an audiologist test and was told there's no point measuring beyond 8 khz.
Nothing beyond 8kHz really matters for generic hearing. There's nothing important beyond that which will effect communication and essential living. Many elderly people can't even hear to that, but it is true that if you lose too much high-end it will affect your life negatively in other ways.
Great explanation Josh! Definitely opened my eyes to audio quality (this is new to me). I'd love to hear your take on audio in cars. What's the best way to consume music in a vehicle? I assume most people stream from Spotify or some other app. Does Aux vs Bluetooth vs USB matter?
My understanding is that Bluetooth reduces quality significantly because the stream is generally in SBC, so bluetooth (might) not be the best quality, but even then it's likely unnoticeable considering most car sound systems are very poor. I imagine you couldn't hear the difference between 128kbps and lossless on even a very nice car stereo. Spotify streams at 96kbps on "normal" settings. When you choose "High Quality" in the app it switches to 160kbps. In my testing 160kbps is quite hard to tell from lossless, you may lose some harmonics but the difference is so small it doesn't practically matter on most systems. In a nutshell Spotify is definitely good enough, especially if set to high quality in the app. I wouldn't have any complaints with it at least. Audiophiles tend to obsess over that extra .01% of quality that's impossible to notice unless you have golden ears and $2000 headphones. AUX is the best quality if you're connecting your phone directly to your car sound system, I personally just stick with that and have no issues.
@@JoshChristiane thank you for a nice breakdown!
That was a fascinating discussion to listen to. I had no idea how many brands, in this category, are taking advantage of the consumer by capitalizing on their ignorance in this field...Thank you for enlightening us!
Thanks, I agree. Thank you for watching!
If Audio Note continue to use standard cd quality thats a good a benchmark as your need.
Good point.
This is like video games where the hardware is advertised as 4k but the game is rendered at a lower resolution and then upscaled.
However, there are many instances where the upscaled 4k game looks better than playing the game natively at 4k.
That's true, and I can't debate that. Same thing with photos, sometimes upscaling does look better... But often it looks worse (due to denoising). Kind of depends on how well it was done to some degree. It can help with aliasing and comb filtering problems though. It's still important for consumers to understand exactly what they're getting, same problems with vinyl right now.
Thanks. I have an engineering background too. I still listen to vinyl as my main music enjoyment. Digital is for background music for me or in the car traveling.
Same for me. I love both and they all have their place. But I cannot deny the quick convenience of iTunes haha.
Agree on most parts especially the marketing ploy, but i beg to differ in my current setup i did an abx and blind tests between 16/44.1 and 24/44.1. I've noticed a slight difference with both bit depth in terms of clarity on the instruments, synths and vocal tones much clearer. (I'm a metal fan) I do not wanna try again anything higher than 44.1khz sample rate it triggers my tinnitus and short sharp headaches on longer playback times
Very interesting. I have not heard somebody claim higher sample rates triggering their tinnitus. We definitely need more research on tinnitus specifically. Thanks for watching!
@@JoshChristiane thanks for responding. Forums like hydrogen audio and reddit as well termed it as ear fatigue iirc. Havent looked in-depth of it but their experiences makes sense especially the headache part. Agreed we need more scientific data on this matter. Take care man!
Intermodulation
artifact products (either mechanically in tweeter and vinyl, or electronically in the chain) in the ultrasonic Freq's, can also produce lower audible products in the audio range, we probably don't want that
You're 100% correct, and also made an interesting point that many don't consider.
Thanks for bravely laying out the facts: they aren’t in dispute. The record business is just that: those that are willing (and hopefully able) to pay overs for a digital recording can conveniently do just that with this branding.
Thank you for the nice comment. No industry is perfect and deception exists everywhere in basically every asset class, but it's good to have information helping consumers know what to avoid.
My experience has been a little different in that I have multiple versions of some of the same recordings. While frequency response is a factor in playback and listening, there is quite a bit that happens between the high and low ends of the audible spectrum. For me the difference could be vocal clarity, the decay of a cymbal or the rolloff of a guitar riff - noticeable between pressings/versions. You could probably argue the difference is between mixes or masterings, but those differences remain. Some of my SACDs make the CD versions unlistenable by comparison, but I may have missed any references to physical media. At the end of the day, though, we all do what makes the most sense to us.
Absolutely there will be differences. But the differences you're hearing will have nothing to do with the resolution itself, and more to do with the quality of the conversion. The same exact converter at 192kHz will sound different at 96kHz, and also different at 44.1kHz. So a lot of people do these tests at home unscientifically and do notice a difference. Your DAC or AD/DA may perform much better at 192kHz and thus you're correct in saying it sounds better. But I've heard converters that sound distinctly better at 44.1. It's not the resolution but rather how the DAC handles filtering, aliasing, conversion and clocking, etc. It's all in the IC. Some of the highest quality converters ever made (and still used in Hollywood) are 44.1kHz Hollywood custom converters, as well as some older Burl gear.
In scientific studies that are done with absolute equal quality conversion at 44.1 and 192kHz, no professional could tell the difference. It is will within the margin of error as well, so we know if there is a difference it's so small it doesn't matter. Ted Jensen, arguably the most famous mastering engineer to ever live said he can't hear a difference, even on his $300,000 setup that is as close to perfect as possible.
With all of that said though, test your converters. Run them at 192, 96, 48, 44, and find out what sounds best on your system, then record at that. For my particular converters they sound best at 384kHz, I'm not sure of all of the internal reasons why, they just do, so I record at that in 32 bit then down-sample and dither at the end. I get an extremely clear, high-quality sound that way and will continue to do so. But I also recognize that's just my converters.
it sounds better because people think it sounds better
@@itsratso.Double blind tests have been performed. It's not placebo.
My understanding is that the improvement is due to the extended upper frequencies and not having that sharp filter above 20k.
I think, it's not about Hz, but about bit in digital perspective, because CD recording is base on binnary digit. You can get same analogy in digitaly process of image. In digitaly definition of 16bit image and 32bit image, you will get much rich color in 1 squere inch of image at 32bit image. So, in 32bit sound, you can get more rich digitaly collect same Hz data to covert to digital code/definition in one package digital data sound. Sorry about my English. 🙏🏻
Oh you're absolutely correct. Nobody is debating that 192kHz is significantly higher quality than 44.1kHz. 32bit is better than 16bit, no question. The question that we must ask is if our ears can tell the difference, as well as if our speakers and AD/DA can reproduce the difference, and generally the answer to that is no. Blind tests consistently show that humans simply cannot tell the difference past the point of diminishing returns.
Came to debunk a new pair of headphones, came away with that AND valuable information regarding recording audio, which is very much appreciated for self taught music producers like me. Great video!!
Thanks for the nice comment, :) I'm glad I could help people learn something about audio, definitely a passion of mine.
In Greece also some companies put in digital platforms fake digitized files. They add strange fake frequencies and they close the sound with muddy bass. Then they increase the loudness extremely even -7LUFS integrated with True Peak over +2. They don't have the original files and they take songs from other sources. Even songs released on compact disc many companies don't own now
Another issue is simply that most hifi speakers cannot do 20hz to 20Khz.
Most of the recordings out there especially early digital, were capped at 20Khz. Even much of the SACD catalogues were actually recorded from material including DAT tape, and contained no information above 20Khz.
Another interesting audiophile concept is that treble beyond human hearing actually improves transients in percussion and bass. Personally not many hifi speakers even at £80'000 are doing decent transients, percusion, or bass anyway.
Regardless, I love old DAC's and see no point these days in getting some of the higher end ones any more, most audiophile gear has coloured output stages to create a feeling of difference, a house sound, so people pick items they like, when in reality DAC's from Topping and SMSL as expamples are more accurate than most of the audiophile dacs, and having a neutral sound is not that appeasing. I mean there are nice DAC's, I like the Mola Mola Tambaqui and if I had the disposable income I would buy it, but the reality is it is not 10x better than a Topping D90 discreet, and none of them are significantly better than many of the late 90's audiophile dacs, which is why some manufacturers have been revisiting those chips.
You are absolutely right on! I'm glad you understand it because it seems like not a lot of audiophile people get that. Marketing is a powerful tool at getting people to spend more money on stuff they don't need. A good pair of HiFi speakers is totally worth it for me, but there is definitely a point of diminishing returns, like with anything else I suppose. I'll stick to my Focal Twin6 Be's.
However, they say that the sharp filter in CD players is negative for sound quality. It is also said that the sampling rate also affects lower frequencies
The same company pushed minidisc technology for decades on the argument that it truncated inaudible frequencies below 20 Hz and above 20 kHz. It seemed awfully hilarious that Sony was now promoting HiRes technology a few decades later.
Very true. Sony does a really good job in terms of engineering for many of their products, especially their higher end stuff, but for some of the cheaper things they sell they like to lean into gimmicky marketing.
I was choosing between an iPod Classic 7th gen, and a FiiO M6. Chose the last one because the price was insanely more convenient (€50) which I bought obviously used. I now read some things about Hi Res audio…
If I paired tha iPod with some Audio Technica headphones, would I get the same quality than on the FiiO?
With something like an iPod or any other small listening device the weak point won't be file format. You will almost certainly hear no difference between the formats and resolutions the devices use, the big difference will come from the discrete headphone amp built into the devices. Some units have really high quality headphone amps and sound amazing as a result, and others sound awful. I can't comment on the FiiO because I've never used one, but iPods were known for having decent quality for their size, especially with low Ohm headphones. My advice would be headphones that are low Ohm (resistance) enough not to stress the amp on such a small music player though.
As I have listened with headphones of 30 euros (around 30-40 dollars), SENNHEISER HD-250, all Hi-Res sound to me so fakely digitized. They have fake frequencies. They have hard kick drums that distort, dull muddy bass and Extreme loudness. Many older are -8LUFS and True Peak over+2, even original cds are mastered at -13LUFS with True Peak below 0. Also at original compact discs they have balanced bass and treble and higher frequencies are clear, generally in original they have crystal clear sound even I rip in Mp3 or flac.
I have noticed some of this as well. Obviously it's a complex topic, but if you have a trained ear you can hear things like processing, low quality EQ, fake soundstaging, etc.
I totally agree with CD standards. Me thinks there's more life in Red Playback than many would have us think. Only way to go higher res is that hi res equipment must be used during recording sessions, so how many recordings have been done this way ????? Even so how much better is it compared to 16/44 ?
It's a good point. It's not that you can't do better, of course you can do better, but after modern mastering smashes all of that dynamic range anyways you're looking at less than 12 bits of necessary range, at most.
How could a consumer find out what the sample rate of the original master recording so that when buying music, we know whether we are being sold upsampled or original content?
Great question. Unfortunately, there is no easy answer. You only know what information they make available. But I can say as having been a tracking engineer in Hollywood for years that no studio I ever went to recorded above 48kHz. I'm sure some exist, but all of the studio work I did was 44.1 and 48kHz 24 or 32 bit. At my own home studio I record 384kHz 32 bit, but that's unusual. Many studios have older high-end converters as well, and not all of them support 192kHz. Some of the older converters sound better at 44 anyways. I only knew what Coldplay's album was recorded in because I read an interview in SoS magazine about the gear they used, and the mixing engineer specified the recording details, otherwise I'd never have known.
I've also worked on albums where I know for a fact they were recorded at 48kHz, and now suddenly I see 192kHz versions of those albums on Qobuz.
You've saved me from falling for the scam ever again. Didn't realize Hi Res is just another marketing ploy. This really helped me understand. Super informative!
Glad I could help! Thank you!
It's actually legit.
@@JasonWW2000kind of, but the hi-res sticker doesn't make whatever headphone capable. The recording/mastering makes a muuuuuuuuuuuch bigger difference.
Good point. Sadly the worst issues among others to me are compression of audio and pitch manipulation.
100%. Pitch correction is on every single vocalist now, it's so frustrating and awful sounding. Billie Eilish doesn't use pitch correction, so I can appreciate her and her producers for being a modern artist that doesn't cave in to the pressure. Modern mastering is another can of worms, absolutely awful quality and destroys everything it touches.
These things confuse me a little. For instance, my very nice Topping USB DAC has "hi-res" labelling, but but will only stream via my computer at 44.1kh and appears to not be something I can change in the DAC settings (not that I need to, or care). For fun, I change the audio property settings on my PC to 48kh and it literally sped up the music. Also, interestingly, I have a rather decent amplifier with a "higher end" DAC chip which streams at 48kh, but the 44.1kh Topping DAC sounds better and more detailed. Like they say, I guess, it's usually not about the DAC chip, but the way it's implemented.
P.S., thank you for the very informative video!
This is a classic case of the disconnect between marketing and reality. Many DAC's sound better at 44.1kHz. If you change a sample rate of a song for playback without doing an actual conversion then it will just speed it up or slow it down, you need to do an actual conversion before you can playback, so that's why you had that issue. Some converters can be a pain to switch around for different music listening as well. In the end the quality of the conversion itself matters WAY more than any fixed resolution.
@@JoshChristiane Thank you for the reply and detailed answer. I'm glad I don't have to worry about bit rates and sampling frequencies so much, just so long as the music sounds good! :)
Well CD quality is good . Well even somewhat esoteric, I believe it was said that 96 khz and more were not necessarily audible, but sometimes responsible for a certain ambience etc . Subconsciously. One would have to compare a to b.
I read many years ago that 44.1 means that a 20 Hz audio signal is sampled 2000 times per second while a 20 khz audio signal is sampled twice per second
The best sound test comparison could be made in the mountains in the big picture in nature. The best ears are definitely the people who live in the mountains far from civilization. Not audiophiles. Noise is poison to our senses. Audiophiles who start reviewing audio equipment in such conditions are probably the closest to the truth. This is a necessary post to people would get closer to the truth. Happy end of the new year to everyone
Totally agree. Technical arguments are right, yet most of the streaming services follow the path of offering more and more Hi-res content.
I personaly only want the 44KHz versions? but can’t find them anymore on streaming services.
I need to play music with digital altered files that have probably artifacts in the frequency and time domain.
Very true. For me I just want FLAC or uncompressed files. I care much more about the container than I do the sampling rate.
The reasons, why SACD and all other HiRes formats haven't make a big breakthrough:
1. decent price players and sound systems capable to take full advantage of HiRes sound quality potential are not widely available.
2. Very few titles produced in TRUE HiRes are available, and popular hit music nearly at all.
3. The sound quality of CD 16/44.1 is good enough.
4. Most of commercial music even nowadays, the sound quality is not even good as CD makes possible. So, if the sound quality on the studio master is not even good enough to take advantage the full potential of CD quality, SACD and all other HiRes formats just simply does not make any sense...
You are absolutely completely dead on correct. This was the perfect summarisation too. The great majority of music comes nowhere near maxing out the headroom and quality representable by 44.1/16, so we need to focus more on the mastering wars first. Mastering is basically destroying music, and the industry should put its focus on solving that problem before anything else.
@@aquaevitae I agree... I've heard some SACDs that seem to have a bit more detail and sound more open than their CD counterparts, but I'm not convinced the SACD versions weren't mastered a but differently and if we were able to co pare the same master on the two different formats, we likely wouldn't hear a differnce..
Totally agree but it's so hard telling people. They're convinced they can hear the difference between a 24-bit 96 kHz (or whatever) & a CD quality file. The only exception I'd make is for multi-channel music mixes eg Dolby TrueHD Atmos where the extra bandwidth actually serves an obvious purpose.
Absolutely agreed. 100% right and good point about multi-channel audio as well, since obviously that's split up. People really think they have golden ears, it's just an arrogance thing. When I bring people into my studio and compare the difference literally no one has ever been able to hear a difference. Ted Jensen (the most iconic mastering engineer in history) claims he cannot hear a difference as well. If he can't, nobody can.
I always assumed (and it's just an assumption) that "Hi-Res" just means higher bit rates than, say, the old iPod .mp3 and AAC standards (128kbps, 256kbps, etc.). But you're right -- if you think you're hearing a "better quality" sound from 24/96 or 24/192 files than from 16/44.1, it probably has to do with the recording, mixing, mastering and playback of the track than it does with the data rates.
Exactly right!
Hi-Res is only a scam if you ignore the other variables that usually go with "Hi-Res" recordings. What's most important is how something is recorded and mastered. A lot of today's music is recorded and mastered for sale to people wearing earbuds or listening with car audio. "Hi-Res" audio is recorded for people listening with high-resolution playback equipment. Mundane content and the music industry not rewarding artists, recordings with compression, equalization, noise reduction, etc. all contribute to what Steve Wilson in "The Sound of Muzak" calls "squeezing the soul out" in much of modern music. Many CD-resolution recordings sound great when these other variables are considered. But it takes a lot of listening and research to find them. It's much easier to find high quality recordings in "Hi-Res". Beware of "Hi-Res" tracks from poor masters. But if modern pop music is your thing, ear buds, subwoofers, and standard def are great.
Streaming vs physical media is an entirely different topic, but can also make a big difference, albeit in sometimes unexpected ways.
You made some really great points, thanks for leaving the comment and being so thorough!
However, a hi res edition may have another better Master with better dynamics less dynamic compression. even a Vinyl release today can have better dynamics than the CD version, less dynamic compression
This is a big part of the equation. A LOT of times the masters for the hi-res counterparts are different and can deceive people into thinking the technology itself is better, when in reality it's just a higher dynamic range master or something of the sort. It's not unusual to have mastering engineers produce 3 or 4 versions of every song. One super compressed for radio, one slightly better dynamically for the CD, one for digital release, etc.
Who has the perfect listening room, without reflections? Most recordings are mixed using a conbination if high end speakers and cheap in car setups to get the best compromise. What I would live to see is a different mixes/masters for different setups. I.e. a high end mix for speaker setups that can produce the full audible dynamic range which will be very different from the average consumer based mixed which sounds reasonable in cheap speakers/headphones.
For sure. Listening environment, speakers, amps, converters, and everything else is going to make a way bigger difference than resolution sample rate.
Analogue audio is sampled an infinite number of times i.e. it is continuous, CD are 44.1 KHz so it not continuous there are small pauses of nothing we perceive it as continuous the same way a TV appears to show motion but what is happening is the picture is changed 50 times a second....
Exactly. That's why vinyl is so interesting to me, and making a big comeback. I love old analogue equipment, and especially old vinyl records from when everything was recorded on tape. Same is true for film cameras.
Tell me you don't understand how digital audio works without telling me that you don't understand how digital audio works.
Um, no, that's not how it works. Imagine a plunger in a pool: the plunger taps the surface of the pool at specific intervals. This creates waves which are seen by our eyes as waves. A cd provides signals like the plunger. We hear the "waves" in the air created by the "pings" of the digital signal. We absolutely do not "hear" the spaces between the digital "pings." You never actually hear the "pings at all. You interpret the waves created by the pings as sound. (in this example, higher bit depth simply means the plunger can go deeper into the pond--which leads to a difference in dynamic range, not a difference in frequency.) As Josh says, all you really need to cover the range of human hearing is a digital signal at double the speed of the highest frequency we can hear. This isn't just a guess, it's relatively easy to prove mathematically. This is completely different than the persistence of vision effect that allows us to see a procession of still images as motion.
With digital, I'd be more concerned as to whether jitter was bad enough to be discernable.
Very true. Good point.
Could you please describe your problems with jitter?! As a former audio engineer i am confused!
Great video and explanation. Appreciated. Thank you
Glad you liked it! Thanks for commenting.
The fact that dither is added to 16-bit is proof that 16-bits is not enough. Now 32-bit float is often used for production, so there's no real reason not to release modern recordings in 24-bit. One way to use the 8 additional bits might be to use 4 of them at the top so less clipping is necessary and 4 of them at the bottom so dither is unnecessary.
So dithering is only added to 16 bit if you're down-stepping from 24 or 32 bit. If the music was recorded in 16 bit then dithering is not necessary and not used. So it has nothing to do with whether 16 bit is good or bad. I do agree that recording in 24b and 32b brings advantages that are worth it, but in the actual end product our ears can only perceive so much dynamic range and 16 bit already exceeds that. The problem is even more obvious when we consider almost all music today is mastered, and mastering destroys most of music's dynamic range anyways. The vast majority of modern music could easily be played back at 12 bit and sound exactly the same, that's how little dynamic range is retained after mastering. Still, for my own productions I record in 32 bit for mixing purposes.
In consumer audio there is no need to have more than 70 dB (about 13 bits) of dynamic range. This limit comes from the properties of human hearing, the practical limits of the acoustic environments people listen to music and what makes a pleasant listening experience in the first place (there are no symphonies written for farting ants and rocket engines, because nobody would enjoy such insane variation in loudness level. Instead the masses listen to DR6 pop! ). 16 bit digital audio doesn't really "need" dithering from perceptual point of view, but dither is added, because it is a very easy way to avoid quantisation distortion. It is at least technically the right thing to do. Distortion free sound with a little higher uncorrelated noise is better than correlating distortion due to truncation. As a matter of fact 24 bit version can contain so much noise that it acts as self-dither when truncating into 16 bit. Since dither allows sound to "fade into" noise the same way it does in analog audio, it is about how much the dither noise masks the signal. By using aggressive noise shaped dither, it is possible to reach perceptual dynamic range of 110-120 dB in 16 bit digital audio! From practical point of view this is insane overkill, but it is possible.
In music production 24 bit is very useful, because it provides a lot of freedom as to how the dynamic range is used and also all the manipulation of the sound doesn't lead to harmful cumulation of noise. However, when a piece of music has been mixed and mastered, the consumers don't need more than 13 bits, so exporting into 16 bit makes a lot of sense.
Unfortunately the ones that can actually hear hi resolution audio will never notice or care about it. By the time you can you are to old and cannot hear those frequencies.
As a child I could walk into a home and hear the CRT tvs squealing even if the volume was down to nothing.
That lasted until I was around 16-18....
After years of factory work, Gunshots at the range, club music turned up way to loud, now i can not hear the beep of the thermometer when i take my temperature. I have to have my kids or wife tell me when its done. Im only 46 years old. When i bought my home i installed speakers into the walls and ceilings of the living room to get perfect surround sound. I can still hear the surround sound but the high frequencies are gone. Not even the beep from your thermometer.. it happens faster than you think. Time is unstoppable and moves faster as it passes. Think of a roller coaster hill. The longer your on the hill the faster it goes. Slowly at first and faster and faster and faster...
I hate my ears have become crap. Dont get me started on my eyes.
THE YOUTH IS WASTED ON THE YOUTH.
Belive that
This is a good point too. You become wiser and more knowledgeable with age, but by then you're maxed out at 15kHz anyways, haha. Thanks for watching and the great comment!
Hi Josh, Really appreciate this video. It helped explain a lot of things to me. I had a couple of follow up questions for you. If I'm looking at purchasing an album on Qobuz (for this example its the Album "Tell me I'm Alive" by All time Low) it will give me the option to purchase it in both 16bit/44kHz and 24bit/48kHz. Is there a way to find out whether the original recording was recorded in 44kHz vs the 48kHz? How does the bit rate of 16bit vs 24 bits affect the sound (assuming the sampling rate stays the same)? Thanks!
Great question, and therein lies a major issue in the industry. It's pretty much impossible to know what it was originally recorded in unless the creators specifically state it on its wikipedia page, or you happen to know the exact gear and settings they use some other way. SoundonSound magazine had info on the Coldplay album, so that's how I found out what it was originally recorded in, but otherwise I'd never have known outside of just using common sense.
I will say this though, as an ex-producer of many albums for many artists, 90%+ of music recorded digitally is recorded 24-bit - 44.1kHz. I can only recall one single time sitting in a session where 48kHz was used, and that was a Switchfoot album (vice verses), specifically because they intended the songs to be used in movie soundtracks potentially. Final delivery of that album was 44.1kHz though, so in the end it was pointless.
In terms of the difference between 44.1kHz and 48kHz, you simply cannot tell a difference unless one of them was a bad convert, so I wouldn't worry about that. Bit depth matters more, but only during recording. When you're tracking you get better headroom, lower noise-floor, and more detailed recordings by recording > 24bit, and some albums are recording at 32bit now even which helps prevent clipping issues later on. Bit depth just gives you more vertical resolution (so each snapshot/stem itself has a great range of velocity storage).
But once the album is bounced down and mastered there won't be an audible difference between 16bit and 24 or 32bit, because the headroom is the same for everything (0dB) in the digital space.
In a nutshell: Just buy the cheapest one you can get. Nothing will be better than 44.1kHz - 16bit with very very few exceptions.
@@JoshChristiane Hi Josh, thanks for the reply. I appreciate your honesty with how audio tracks are recorded. After watching your video it sparked my curiosity and I did further research and yes I agree with you the bit depth gives you a larger noise floor when you are at 24 bits, but as you mentioned in your video even if you had perfect ears (I certainly don't) and the top of the line equipment you wouldn't be able to hear the difference unless you blasted the sound at a high enough level that it would cause hearing damage. at this point it seems like the only value of the "hi-res" tracks are that they show up as HQ on my DAP.
I have a follow up question. If CD quality is the best that you are going to be able to get. Do you recommend ripping music from a physical CD or is downloading from a site like HD tracks or Qobuz just as good? Reason I ask is I didn't know if they did any signal processing or funky things like that. I saw an article where they compared an analog signal, vs the CD, vs a newer CD, vs a download
magicvinyldigital.net/2022/06/18/dire-straits-money-for-nothing-review-lp-cd-qobuz-1988-remastered-2022/
One interesting article I found last night was this NPR article where you can play a lossless .wav track against a compressed 320kbps and a compressed 128 kbps file of the same track. and you had to pick which one is the lossless file. What was interesting was that majority of the files I selected the lossless version of the track, but for the 2 of the 6 that I selected the compressed file, I selected the 128kbps track. I would have thought I would have selected the 320kbps. its an interesting experiment.
www.npr.org/sections/therecord/2015/06/02/411473508/how-well-can-you-hear-audio-quality
Totally agree with you on all that.
In terms of CD quality (digital) vs CD RIP, there shouldn't be any difference assuming you have a good quality ripper. All things being equal they should basically be the same. I personally just buy digitally on 7Digital, Qobuz, and very often on iTunes (for their mastered for iTunes albums). iTunes is VBR 256 minimum in the highly efficient AAC Codec (which is better than MP3), the lowest part of any song is 256kbps due to the VBR, but in more complex parts of the song it can exceed 400kbps. In my opinion iTunes rips are actually very good. They're not a good archival format in case you needed to convert later multiple times, but for straight listening 256kbps (VBR) AAC is way more than good enough. I stop being able to hear the difference at a constant bit rate around 200kbps, often much lower. For a good VBR I could do with an even lower minimum closer to 160. So don't be afraid of iTunes in case you're on the Mac ecosystem and were wondering if they're any good or not, their mastered for iTunes albums are second to none. Outside of iTunes I really like 7Digital because their selection is fairly wide, prices are good, and they offer 320kbps AAC with the base purchase (usually $9 for an album). Qobuz is great as well, top notch quality, but prices are a little bit high.
I've done those experiments on NPR as well as many others online, and it's amazing how minor the differences are. The site I recommend to test your hearing and knowledge is: abx.digitalfeed.net/
It takes a minute to understand how it works, start with the lowest quality shootout first (96kbps) to help gauge things. And tell me what you come up with as I'm curious!
Sometimes lower bitrates actually sound better, which is a bit of a weird phenom.
I don't think albums sold at higher sample rates are upsampled. They are probably recorded, mixed and mastered at 24/96 or 24/192 and then downsampled to "consumer-grade quality" 16/44.1 for CD and streaming. Does not mean it sounds worse, I'm with you with this. I can't hear any difference on streaming services that offer "lossless, better than CD quality".
Some are, some aren't. An example is Coldplay's album X&Y. I read a technical document years ago from the lead engineer, and he said it was all recorded at 44.1khz on an RME interface (which is not abnormal in studios, especially back then). But yet I see the album commonly available in 192khz. So 100% that album was just upscaled. This is only one example, and there are many more. I also personally worked on an album for a Christian band years ago where we did all of it at 44.1, and now the album is magically available in "hi-res 192khz". One could debate whether this should be considered consumer fraud or not. Another example is 4k UHD Bluray copies of movies like "Hugo", but Hugo was filmed entirely in 1080p on the Arri Alexa at the time. So obviously it's just a fake cheap upscale.
The difference between lossless and lossy is a different argument entirely though, because that's less to do with resolution and bit depth and more to do with the actual casing compression for delivery format. When we are talking about a low quality lossy MP3 vs WAV or FLAC then of course there may be some audible differences, though the differences become basically negligible past 256kbps AAC or 320kbps MP3. Everybody in audio has opinions on these things, but the actual testing is pretty definitive that there is a point of diminishing returns on perceptive quality.
Most older catalog albums were upscaled or digitized from analog recordings. Considering the Universal fire destroyed a surprisingly large amount of analog masters, new issues maybe coming from a digital source at 44.1 kHz. New recordings made today are made at various sample rates from 44 to 192. Plenty of music originally recorded at 44 are being sold today at higher sample rates.
it produced a standard in Japan for what would count as hi-res regarding, among other things, frequency reproduction
My understanding is that the quality of the recording itself trumps other factors. I like wide soundstage and good bass response among other things... to me it seems that IF it was not in the original recording it is not going to be in my system regardless of any other factors. I have hearing loss and tinnitus on top of it, but I am grateful for still being able to play my guitar and listen to wonderful music. I have to admit I am always searching for the next best thing... but right now I do have some decent equipment to play through and hear through. What I really need is a brain upgrade. ;-)
Totally agreed. Recording quality is by far the most important factor, it's a real shame that modern mastering ruins so many wonderful mixes. But since we can't do much to fix the loudness wars of today the next best thing is to try to appreciate older music with better gear, haha :). Thanks for watching and your nice comment, happy listening!
@@JoshChristiane My pleasure Josh. I am looking forward to more content from you and I really enjoyed your presentation and demeanor. God bless you and all you love.
Hello Brian ... Something to keep in mind. The "soundstage" and all the attributes of openness and detail that we all love are actually burned into the recordings at the studio by Panning, Phasing and Delay techniques. It does not come from our home equipment.
That said... while our stereos cannot create or enhance "Soundstage"... poor speaker placement or room modes can and do tend to sabotage it.
@@Douglas_Blake_579 Hi Douglas and thanks for that great insight! I think sometimes that I am guilty of trying to get something out of the music that was never in there in the first place! Even the best live performances are not always perfect but they sure are fun!
@@BrianVallotton
There's no such thing as a perfect recording. What varies the most is their distance from perfection... some get close, most don't.
That search for "something extra" is this hobby's Achiles heel. As a service tech I saw plenty of really bizarre stuff... everything from $5,000 power cords to interconnects so heavy they broke the connectors on the back of some truly expensive gear... all in the name of better sound.
I wholly understood aspect of the KHz but what about the bit rate
So we talked about those "snapshots" which are the captures of velocity that map out a waveform to represent audio. How fast those snapshots happen is the sample rate, but how many bits are used to capture each individual snapshot is the bit depth. The more bits you have, the greater the resolute range of data on that vertical axis. Basically think about it as the sample rate being the horizontal resolution, and the bit depth being the vertical resolution. Hopefully that makes sense, thanks for the great question!
@@JoshChristiane Right, but even if the frequenzy does not benefit from being beyond 44khz, surely a higher bit depth will more accurately reproduce the true analog waveform?
Yes, more accurately. But again we run into the same situation where we have to ask "how much is enough"? Can a human ear tell the difference between 16bit and 32bit depth? The answer of course is no, outside of extremely unusual circumstances. Higher bit depths should be used for recording because it allows effectively more dynamic range in post production, but after mixing and mastering the dynamic range is so decremented that difference is completely negligible. CD quality is still the best the human ear can hear, 16 bit 44.1kHz. While it can easily get better than that, we simply cannot hear the difference.
@@JoshChristiane
Yes, technology develops for the reason that these creators of technology also want bread on the table. Unfortunately, technology develops faster than human evolution allows. Obviously, we need to start producing products that allow humans to hear as well as dogs.lol
Ya know, we kind of settled this argument back when CDs came out and we endured the CD vs vinyl wars. The specification for music CDs was actually simple and brilliant. Today I'm a huge fan of the FLAC format. A significant reduction from .wav file sizes and in my most discerning listening using IEMs, I just can't imagine I'm missing any detail. Now if we could get lossless compression for video. Movies are the storage gluttons on the server.
There are a lot bigger grounds to be made up in video for sure. If there are any audible differences between 44.1 to 192 for example, it's so small it doesn't matter. When talking about streaming video, compression types, and even just editing or refining techniques (denoising, glossing, etc) these differences matter significantly. Look at how bad 4k streaming is, but yet how far it's come in 10 years. HEVC has been a pretty big help for me storing all of my videos, but even then I feel like it has a long way to go. Thanks for watching and commenting :)
(NO SHOUTING) *THANK YOU JOSH TO TELL THE TRUTH. NOW WHAT ABOUT VINYL VS. CD? WHAT IS THE BETTER HARDWARE FORMAT?*
That's an easy one to answer: Vinyl. Vinyl is not a scam assuming that the album was recorded on real tape (like Fleetwood Mac, Lez Zepp, etc.), and then printed from that tape onto vinyl records. The quality of of a real vinyl print is vastly superior to anything digital could ever reproduce. Way higher dynamic range, lovely sounding smooth analog distortion when you exceed the red line, and true-to-life representation of what was recorded. However, with that said, vinyl offers absolutely no benefits over regular CD's if the album was not recorded on tape. Meaning that any album recorded after year 2000 is basically digital anyways, and going from digital to analog serves no benefits.
So yes to vinyl, but only for old albums that were on tape originally. Most of which are better than modern music anyways.
@@JoshChristianeThat’s how I draw the line. I love all types of media: vinyl, CDs and streaming but I don’t go newer than 1985 for vinyl, I don’t see the point.
Im a bit confused by this. When you reference our ears not hearing past 20khz versus sample rates and bits, aren't you talking about 2 different things? Hearing high-pitched frequencies (20 khz) isn't the same thing as a digital recreation of a sine wave.
Sorry about any confusion, I'll try to clear it up here. The two are interlinked. If a sinusoidal wave pattern transitions between polarities faster than a computer is digitally taking "snapshots" then it can't resolve that information. It literally can't capture a wave that's moving (transitioning) faster than the digital captures.
Think about it like a movie camera. If you're shooting 24fps and you try to film a bullet exiting a gun, you're very unlikely to capture the bullet. You need more frames per second to capture that fast moving object. But if you record at 3000 frames per second you could capture the bullet, and even slow that footage down to see the bullet path. This is the same analogy with audio.
A 20kHz wave is a very fast wave, the only way to capture that information is to have enough resolution (snapshots of data) per second to actually store and replicate that audio.
Hopefully that clarifies the link between recording rate and audio frequencies.
@ It does, thank you.
Superb explanation, very detailed but easy to follow. Thank You. Ian (UK)
Thanks for watching and the appreciative comment, Ian. Have a great week!
The quality really depends on the recording, from tracking to mastering. The delivery in 16/44 is more than enough! So get the CD, make a FLAC and be happy...
You're exactly right. The source material is so much more important. The technological debate does matter because it's how we progress and build new tech in the future, but if you're a musician or audiophile recording music then by FAR the most important thing is tracking, mixing, and mastering.
And I could argue the musical performance itself trumps all else.
Many years ago I read about a study where they added synthetic harmonics I think it was 50hz to regular AM radio and the result was apparently people could hear the sound more clearly. then Pioneer brought out CD players that created high sound frequencies artificially
Live Engineer here with a background in recording. I am currently running my Yamaha Rivage PM3 at 24 bit 96kHz and I can hear a huge difference over my old M7CL that ran in 16 bit 48kHz. That being said, I think it's laughable when someone spends $1000 on a USB cable and swears they can hear a difference in their 1s and 0s being transferred to their DAC.
Hi, I uploaded a video to my channel explaining exactly why there's no audible difference between hi-res and CD quality. In a nutshell, when you subtract the CD-quality signal from the hi-res version, it cancels out completely. There's nothing left in the hi-res signal that you can hear - CD quality represents all the musical detail just fine.
Frequencies you can't hear still shape the frequencies you can hear. This is how we use new tech like phase steering to manipulate the direction of sound and create dead zones in isles while giving every seat in the house a perfect stereo mix. Anyway, not going to argue that you need anything higher than CD quality while listening to recordings... but it makes a massive difference live when you are working on top end gear.
@@BionicBurke No, those ultrasonics really have no effect on what you hear, unless there is intermodulation distortion. If your speakers are rated that high, it can be avoided, but if not it's best to just filter them out (which a lot of speakers do).
That's really cool! I was also an FOH engineer for festivals back in the day, who knows maybe we even met each other.
The difference you are hearing could be other factors though, right?
The new Yamaha Rivage PM3 could just be a lot better converter than the previous. Even the same converter will use different IC's at different resolution sample rates, so just using one converter is also not a scientific test as many think it is.
A good converter will use entirely separate IC chains, separate word clocks, different aliasing and filters. So just because a converter sounds better at 96kHz does not necessarily mean it's the sample resolution itself.
I mentioned this in other comments but I have a custom Burl converter that sounds better at 96kHz than it does at 192kHz, and I have older Hollywood Converters (the best in the world) that sound better at 44.1kHz than they do at any other rate. So there are many factors.
When the testing is done purely scientifically using a singular conversion chain the studies I saw showed that professionals and audiophiles can't tell a difference. They could however tell a difference on many high-end brand converters in unscientific settings.
I'm just saying it's possible the difference you're hearing is the converter, not the resolution.
What about Japanese cd’s and how the higher frequencies are different than most regular cd , are they better or is it marketing?
Mostly marketing, but not entirely. Super Disks and DSD's were awesome for a lot of reasons, as they could store higher quality music. The ONLY time those disks are legitimately better though is when you are printing unmastered recorded performances. So for example if you're recording an extremely dynamic instrument like piano that has a performance that goes very very very quiet, then becomes very very loud. That type of performance is notoriously difficult to capture all of its detail. But almost all music today is mastered and its dynamic range is intentionally reduced for easier listening, because people don't want to jump for the volume knob constantly. A little compression is good, a lot of compression is bad and most modern masters use far too much to compete in the "loudness wars" which in turn destroys music.
So I like the idea of a higher quality disk for archival reasons, but you won't hear a difference outside of extremely unique situations that are rarely ever present.
How about instead of the endless argument about formats audio engineers and audiophiles and the community in general push for a simple designation standard of recorded music. Like SCA the coffee brewing association has. A brewed cup of coffee has to have the following elements, temp, grind size etc. So if an album is recorded at say the proper bit rate and depth and filtered properly or whatever the criteria should be for high quality sound. ***** star it and let the buying public know how it was recorded and mastered rather than if its Hi-Res ,etc etc etc. When I stream an album that sounds great no matter the resolution I have taken to make h it as I love symbol most services have. And only use that designation for that purpose? Just a thought?
Totally agree. Mastering is the greatest weakness in the industry right now as well, the quality ranges from "okay" to incredibly bad. We need to produce an environment where music isn't being destroyed to compress the heck out of everything just to make it louder.
Its not a scam in the sense of buying high end headphones and pairing it with a high end setup and then using that to play a loseless audio file for its higher quality as well. Its not all about what spectrum of frequencies we can hear. We just want something that sounds more clear and detailed to our ears than the norm.
Is it overkill? Possibly.
Does it sound better? Absolutely.
Absolutely. Buying a good pair of high quality headphones, a good headphone amp, and playing back lossless files is definitely a great way to enjoy music. That aspect is definitely not a scam, I am more just talking about 44.1khz lossless vs 96khz and 192khz lossless here. You have to pay like $10 more for the higher sample rate on a lot of albums, and you gain no quality.
@@JoshChristiane just watched the full video 😂 I have to agree there. When my dac switches between the different khz I can't tell the difference in quality.
Yeah same here, nobody can tell the difference unless there are some really specific conditions. If you slowed the music down 300 or 400% then maybe differences would start to become apparent in the high-end, but at that point the audio quality would be so messed up it wouldn't matter anymore. Some converters sound better at 192khz which deceives a lot of people, but that's just the quality of the conversion itself not the frequency range difference.
Yes it is. external Headphone amps/dacs are generally nonsense as well. Most mobile devices has sufficient audio processing unless the drivers specifically need amplification which most consumer grade units don't. Lossless is a subjective basket being that music production sometimes derives sounds from less than stellar sources. I.E Where was the instrument from a rompler sampled from? at what rate? Was a sample pulled from Vinyl, or an obscure, out of production album that someone uploaded to youtube? are the headphones neutral monitors, or capable of being adjusted to be so? The Audiophile experience is often placebo, namely because most audiophiles don't understand the music production process in the first place.
@soundahfekz321 A lot of truth to all of that, for sure. Source will always matter the most.
I'm old, so I remember when TVs were nothing but CRT screens, and they had a flyback oscillator that me, as a child, could hear running at 15.7 KHz. If I wanted to know if my folks were still awake I would open the door the the TV room and hear the whistle. Now well into my 50s I can hear maybe 10 KHz tops, and the funny thing is it seems mostly old guys my age (but I suspect with more money) want to extend the response out to over 20 KHz. Forget about it.
That's a great example of how hearing changes over time and can be influenced by environmental factors, and it's a real bummer how some of the frequencies we could hear as children fade away as we age! The youth truly is wasted on the young.
Our ears actually can hear better than CD quality, but good video just the same and hope you cover SA-CD at some point as you booted below.
It's a bigger topic than I can cover in one video, but you're definitely making a great suggestion! I'll make sure to do a video on SACD sometime in the future, thanks for watching!
Actually CD quality has greater bandwidth than human hearing. Pat Brown did an experiment ("Is CD Quality Good Enough?") where he tested dozens of audio engineers and nobody - not ONE subject - was able to hear a difference between standard CD quality and hi-res. I think that says something!
@@ViklasSvenske I'd have to see and read about that one. Any links?
@@nicholassheffo5723 YT comment filtering won't let me post the link, but it's at ProSoundTraining, May 31, 2014 by Pat Brown. Title is "Is 'CD Quality' Good Enough?"
I have a hi-res audio player, with 24 bit 48 kHz .wav or .FLAC. I remaster my CD's or singles in a way that the original digital mastering could not
What about flac?
FLAC and ALAC (Apple's version of it) are both compressed lossless codecs. A codec means "compression-decompression". So both are file formats that compress and then decompress the audio on the fly. The reason they're popular is because you can keep the lossless quality of WAV or AIFF, but in a smaller file format with files that are often half the size. FLAC (and ALAC) are fantastic, but they have nothing to do with the quality of the audio itself, they just store the audio. Think about FLAC, ALAC, WAV, AIFF, and any other lossless format as all containers that hold the thing itself, but the qualities of that thing are not necessarily entirely determined by the container. MP3, AAC, and OGG are lossy formats so they will reduce the quality of the sound they store. FLAC supports 16bit, 24bit, and 32bit audio (for bit depth), and it supports 44.1khz all the way up to 600,000khz+, so the range of what it can store is basically infinite. I'm actually surprised the format doesn't get used more often than it does, as it's my favorite. Assuming you're building a music catalogue or collection, I highly recommend storing all of your files in 44.1khz 16-bit, FLAC or ALAC with full compression (8) enabled.
Hi-res stops at 24 bit 48 kHz. 24 bit depths gives me sonic possibilities at remastering that weren't there at 44100 Hz
Question: it’s it worthwhile recording from yt music for playback in public space for background music vs just recording the yt audio/video version of it?
I think that answer depends on exactly what you're doing with it. If you were a DJ just downloading songs from TH-cam Music for playback then I don't see an issue with it, the quality isn't great but it's not horrible either. But when you're streaming it from TH-cam the quality is actually significantly lower. You have to google what quality they stream at, but I remember it being low. Depends on the event, but for most situations it probably won't be a big deal.
What Spotify hasn't even started with lossless? even though they've been talking about it for years
Great video, thank you. Although your last statement about collecting music in CD Flac file format, I disagree for old music the best is to put together a vinyl collection of original first release records. 😊
Thank you so much for watching and commenting. I do agree that vinyl is awesome for old music, especially music recorded on tape. Nothing better than the real thing :)
And then Philips and sony brought out SACD. which was not PCM based at all but DSD although many albums apparently converted from PCM. others opted for DVD audio instead. but it went to the grave. although you can find some discs still
DSD was really cool, especially because the masters for it were particularly well done. A big reason for this was they knew the target market were a more caring and sophisticated listener. Wish they'd bring it back to popularization!
I don't know about your ears but I listen to all my CD music as 128kbpsMP3's on a Windows PC connected to an AV Receiver at 24bit88.2 Studio Quality.
Compression or weak MP3 format or not, the music is light years sounding better than the same CD I ripped from into the PC DVD-ROM drive or through a direct 4kBD player.
Infact it sounds too rich that if I play it even higher than 24bit96 the audio feed from the PC will brick the high end AVR.
I can agree that HighRes Audio was developed for hardware. Even the HEVC DD48k 5.1 is much richer than through a BD player just because Windows is configured backend at 88.2.
In saying that Cold Play knows more about sound recording quality is like saying Doctors know about medical technologies and chemistry, they don't... tech companies and pharmaceuticals create medicines and MRI tech.
what about dsd? i have a lot of thoughts on dsd i have no idea if it sounds better or more faithfully recreates live music in large part because files recorded in dsd often are kinda niche auidiophile type music and that type of music is loaded to sound good coming out of a stereo. look for a well known track and it would be a master tape transfer to dsd which is unlikely to yield benefit over cd which is the best format imo. the way dsd is implemented in theory more accurately recreates a sine wave as well, as opposed to the 'little steps' approach of more traditional digital formats.
also i wish he would tell people that vinyl is not a thing that would make a great video. everything here is more true of vinyl
Vinyl is awesome when it's real. The problem is most vinyl today comes from digital prints, so it's basically pointless lol. The entire point of vinyl originally was to convert recordings from tape into an actual analog product without ever having to digitize it. This is technically better and worse than digital in a myriad of ways, but mostly it sounds better and warmer to our ears.
As far as DSD goes it's not actually better in terms of technology, but the masters were better leading everybody to think the format was better. There is a good article on the technology that I recommend reading here: www.mojo-audio.com/blog/dsd-vs-pcm-myth-vs-truth/
Apple music does support hi res lossless but I do agree hi res isn't worth it cd quality is the best standard and that should be the standard.
It only does if you're using their streaming service, but if you buy an album on iTunes then it's VBR 256kbps AAC.
It was really educational. Can I connect with you for some more information please.
Thanks! Please do connect with me, would love to talk with you.
Already in the 90s, recordings were made in record studios with 20 bit
I agree with every word you've said. It's a huge marketing ploy but unfortunately the general public has fallen for it hook line and sinker. It's become an urban myth and trying to convince people otherwise is like whistling in the wind. Keep up the good work. 👍😢
Thanks for watching and commenting :)
There are those who are critical of sound quality when it comes to streaming. these people think, for example, that streaming in their own network or playback from physical media sounds better
Hopefully, everyone is aware that the term "scam" can apply to anything. It's basically a person's opinion of something. Taxes can be a scam. Airplanes can be a scam. Greek literature can be a scam.
This is an excellent point. But then almost everything on the internet is an opinion anyways, outside of scientific research. But even research can be heavily biased by opinions, nothing is for certain but death.
I got so aggravated with your video I only made it about half way through. Lol
Hopefully you explained that Hi Res is a standard and that 99.9% of people shouldn't care about it. It's really only for people who are building a full Hi Res system.
Personally, I'm more than happy with a non Hi Res system.
I did explain exactly that, and that it's mostly marketing taking advantage of people just to push people to buy upgrades they don't need.
Thank you for this video.
I totally understand the math and logic behind these numbers - yet the reality looks a bit different. Probably for completely other reasons, as there are many more factors involved than just these numbers. At least in the few examples and tests I have done myself, HiRes can _definitely_ sound much better than a CD. Even though technically it should not.
I own a few albums both on CD as well as in a 96kHz/24bit FLAC version from an online HiRes audio file seller.
With some of these albums I cannot make out any difference between the CD version and the "HiRes" audio version. In those cases they clearly sold merely an upsampled version of the CD. And I agree, a total scam.
However, I _do_ have a few albums where the HiRes version _definitely_ has a lot more fine detail, that is nearly muffled or sounds totally flat, like the instruments being recorded under a felt hat or something, on the CD version in comparison. Maybe the CD version was poorly engineered? Or despite being brand-new and having no scratches the error correction destroyed the sound due to a poor CD pressing? I doubt my CD player is to blame as other CDs do sound exactly like their HiRes audio file version. So I don't know the exact reasons, but I _can_ hear a clear difference.
If as a consumer my only choices for a certain album are either a CD version or a HiRes FLAC audio file at 96kHz/24bit - with no 44.1kHz FLAC version sold - then the HiRes version _can_ sound much better than the CD version, and hence would still be worth buying. Despite the pure numbers not making any sense.
Perhaps the difference then is not the kHz but the bit resolution. 44.1kHz recordings are often done at 16 bit, while 48kHz or 96kHz typically use 24 bit.
If a 44.1kHz/16bit recording is upsampled to 96kHz/24bit it will not sound any better, and likely even worse due to the resampling.
However, if the original source was a 48kHz/24bit recording, the CD at 44.1kHz/16bit might indeed sound a bit more muffled and flatter than a HiRes 48kHz/24bit FLAC file - even if that one was upsampled to 96kHz/24bit.
That is the only explanation I can think of, for the clear difference I hear in some HiRes audio files, compared to "CD quality".
Great comment. So if you're hearing a difference between those sampling rates it's not because of the rates themselves, as you stated, it's from other factors.
Some converters just operate better at 192kHz than they do at 44.1kHz. Some word clocks operate better at higher rates, some operate better at lower rates. There are least a dozen factors that can make a higher (or lower) sampling rate sound better than another, and it's notoriously difficult to get files for blind testing on this topic because of those differences in filtering and aliasing.
However, when the tests are done purely scientifically with extreme care to get exactly matching files, people cannot tell the difference audibly between 44.1 and 192. So the key here is making sure your recordings are tracked well, mixed well, and mastered well, because the sample rate won't matter much in the end if at all.
Thanks for your great comment, I read everything you wrote and thought it was very well written.
@@JoshChristiane To make matters worse, HiRes audio file vendors never disclose any technical details. What was their original source? A 192kHz/24bit file? Or merely a 44.1kHz/16bit file? What was the hardware and software used to upsample or downsample these files?
At the end of the day they make you buy the proverbial cat in the bag. It's a lottery game.
As we do not have enough data to make a truly informed decision, we are left with having to buy HiRes files hoping for the best. That for whatever reason they may - or may not - sound better than the CD.
And like with the lottery, sometimes we win, and indeed the HiRes files sound crisper with more detail than the (now comparatively) crummy CD.
And sometimes we lose and both the HiRes and CD version sound exactly the same. And it was all a waste of money.
Lastly I would also like to bring this item into the discussion:
As a fan of Howard Shore's soundtrack for The Lord of the Rings, I have three versions of it: the CD version, the Audio-DVD and the Audio-Blu-ray version.
The dynamic range and clearness of the Audio-DVD version compared to the CD version is truly stunning! It literally sounds 10x better than the CD. With the CD it sounds like you are sitting in the auditorium of a concert hall, listening to the orchestra in the distance. Yet with the Audio-DVD version you are literally sitting _inside_ the orchestra, amongst the musicians!
And surprisingly, the Audio-Blu-ray version yet again sounds noticeably better than the Audio-DVD version! Perhaps not another 10x better, but still noticeably better again! To continue the earlier analogy, now it seems you are actually sitting on an orchestra player's lap - even hearing him breathing. The clarity of the sound is just amazing!
So why is that?
Is it merely a better remastering? Yet they are all ultimately from the very same original source. Or is it better mastering equipment? Possible, as these three were released many years apart from each other.
Or is it perhaps indeed somehow related to the higher bitrates and sampling frequencies, with the CD at 44.1kHz/16bit, this Audio-DVD at 48kHz/24bit, and this Audio-Blu-ray at 96kHz/24bit?
Frequency-wise according to the math all three versions should sound exactly identical. Yet they are miles apart from each other!
So in the end, in reality, it is still _very_ _well_ worth it to buy HiRes audio versions. In some cases they really _do_ sound amazingly better - for whatever reason...
Yet every time we buy them, it is a lottery game. It may sound amazingly better - or not at all better than the CD.
People get so transfixed with Hi Res, this is not the most important thing. The one thing imho that makes a track sound great is the quality of the recording you are listening to. For years i was a subscriber to Qobuz listening to most tracks at 24 bit 192 khz through my HD600s and Chord Mojo and yes most tracks sounded great, but a lot did not and some music videos on TH-cam sounded better than those badly recorded tracks. For the last 6 months I have been listening to Spotify through my Sony XM5s through Bluetooth LDAC. I have Spotify set to Normal best streaming quality with Dolby Atmos turned off. Through my Samsung smartphone i have the HD audio turned on in the Developer Options playing through 16 Bit, 92khz at 660 kbps. In the Sony Sound Connect App I set my favourite EQ settings to Custom 2. I am telling you honestly that listening to Spotify this way the sound is just as good as when listening to Qobuz through more expensive equipment 🤷♂️
Absolutely 100% correct. The quality of the mixing and mastering is so much more important that the entire topic is bordering on pointless. Even if you're listening to super low quality MP3's at 128kbps, a good mix will still sound vastly better than a bad mix, lol. So by far the most important thing is the mixing and mastering.
Hi-Res allows for less steep filters, therefore less damage to the audio signal. The filters are not only applied in DA conversion, but also in the studio‘s AD conversion. Gentler filters = better audio.
Studies have shown that the human hearing resolution is ca. 7 microseconds. You claim that more than 44.1 kHz sampling frequency is not necessary. However, CD quality only means a time resolution of 20 µs. Only at 192 KHz sampling do we get down to 5 µs. From this I conclude that Hi-Res can make sense over this point of view. Granted, not every listener will have that ability, but prof. musicians and conductors with trained listening may have.
I hear what you're saying, but plenty of studies and blind tests have been conducted during research on this topic. Every single one I've ever read concluded that professionals and amateurs alike could not tell a difference. I'm not saying there are not differences, of course there are... But they're just no longer noticeable.
People mention filtering and such, but mostly that's based on old information. Modern converters work quite differently to produce extremely high quality filtering even at low rates. It's so good today that I doubt any living person could tell a difference in a blind test even between low-quality modern converters.
@ Sadly, you offer a totally uninformed opinion. Rather than try to learn about and understand what may be different in High Res, you wildly speculate about something with insufficient understanding. But you are young and still have time to catch up.
Same marketing tactics using the term Digital Antennas to sell old UHF/VHF antennas for OTA TV reception by confusing unsuspecting customers.
I actually didn't know that, how interesting! Thanks for letting me know.
@ actually the only digital of the whole thing is the signal, which is decoded by tuner. Same thing with color antennas, no such thing, the signal is using the same spectrum bands.
yeah but i still have noise wenn i rip a cd to alac or wav. I looked at spectrograms of my alacs and they all had noise. i didnt convert anything i just converted it from one format to another. i converted one cd to flac. still noise in the spectrogram
That's often from the CD press itself. I used to make my own CD's (from actual bounced prints) and depending on how you do it, it can cause noise. Dithering adds noise intentionally to help reduce bit error noise when going from 24bit to 16bit for CD standards, so it's possible that's also what you're hearing. This is why I buy digitally (7Digitial, iTunes, Qobuz) then download 44.1 16 bit CD quality FLAC or ALAC for storage. Nothing wrong with physical CD's though, which ripper are you using? I just use Foobar2000 and have had good results from it.
Must also consider dynamic range. Most standards unacceptably compress the dynamic range.
My biggest concern and complaint is mastering itself, not any file format. Nothing destroys dynamic range today more than the loudness wars, we need to find a solution to that.
@@JoshChristianeI went back to CDs for this reason as I got better equipment, a lot of the masters on streaming are using the brickwalled remasters. An 80s CD with proper dynamic range sounds better to me than a poorly remastered 96khz-192khz lossless file.
I don't know much about the current high-res players but I want to make the sure the audio file is at least the same SCD quality and I want to make sure the player and whatever I'm listening on can reproduce The full quality of the recording and I'm listening to you I mean that's what's important here. I would hope that a music player that's listed as high res can play back a audio file and it's full quality now as far as headphones being high res I don't know about that, I can understand a music file and an audio player being high res but headphones I mean you got awesome headphones decent headphones and sucky headphones so if you get a set of really good headphones it should be order do the job, I don't know if I've ever seen a set of headphones labeled high res but any truly good to the headphones should be able to play back a high-res file and make it sound awesome. But 44.1 versus 48 You're probably not going to hear much of a difference, 192 versus 320, you're probably not going to hear much of a difference so I get what this video is about I think he's putting out a good message here
You are right, it is all about the chain and it doesn't really matter if a device says "Hi-Res" if it can't actually reproduce that quality. In the end the source material and how it's processed (mixed, mastered) will matter by far the most.
"Hi-Res" is mostly Scam...
What ?
As a speaker builder, musician, mixing and mastering artist i can say.
That most modern music is only 12 Bit, after the mastering stage...
Cause 1Bit is 6dB in dynamics...
So 16 Bit is 96 of Dynamics.
The lowest ( level ) tone in a song is around -35 dBFS and fades are down to -60 / -70 dBFS...
The highest ( level ) tones are going up to -5 dBFS...
So you have a dynamic range of 12 Bits = 72 dB...
Classic music, real classic music without compression on.. so with the brutal original dynamics... can benefit from High Resolution Media...
Okay kHz... 44,1 vs 192...
My experiences are... the higher the kHz, the more unnatural the sound is... it gets more and more clinical sterile...
So 16/44 is good enough for me...
WHEN the mastering was a good one...
AND THAT is the biggest problem today... over 80 to 90% of music out there... today... is sonic garbage !
I know, when i hear a modern produced song, what went wrong in point of physics, psychological and technical points... beside the stupid loudness of today...
Cause of this mess... Vinyl is at moment, although it is the worse format... the better choice over Hi-Res...
Why ?
Cause the cutting engineer need tracks which are less loud than the CD / Streaming versions...
Cause of that... dynamics came back and more of the original mix sound...
Cause for CD and Streaming, the life is squeezed out of the mix, to BE THE LOUDEST ON THE STREET !!!
A complete stupid mind set in the biz...
Why ?
Cause near everywhere audio is normalized... Even on CD... cause you are the normalization algorithm with you hands... Cause you dial the Volume knob of your amp...
16/44 on CD, when the Mastering was a good one, is fine and can sound warm like Vinyl... without the technical restrictions of Vinyl...
I hope this helps a bit :-)
Totally agreed with most of this. This was a good analysis with some excellent overall points made. I know it's a nuanced topic, and many confuse converter quality with format quality, so that alone can be deceptive. But in the end quality can only get so good before something sounds real, of course there will be a diminish of returns.
There are formats that have even higher sampling rates. and there the idea is that it is precisely the number of all of them that gives higher sound quality. and also maybe the idea is that it should sound more natural or analog two ex DXD 352.8 kHz DSD DSD 64. 2.822 MHz
FYI In respect of pixels on a camera phone they just double up and multiply the resolution as you realistically are probably getting half meg on a so called typical 40 meg camera. It's impossible to obtain these resolutions from a small lense regardless of the sensor, it's just marketing hype as usual 🙃
As for the Hires yes and no. The main reason for up sampling is to add more ingredients in the music designed to then be digitally enhanced in the detailing specific to the quality of the speaker or headphones etc. I do agree though in normal daily hearing frequencies 16 bit is ample. Thanks 👍
Thanks for the comment! Totally right on about cameras though, adding more pixels absolutely does not give you more "real resolution" unless the lens can actually resolve that information. But for example on a lot of small smartphones with cheap glass that's not even possible, so definitely some heavy marketing at play there.
Wow that's very informative Josh !
Glad you enjoyed it! I appreciate your kind comment :)
@@JoshChristiane Thanks my friend, cancelled Qobuz as I found albums from 77/78 in 24 bit & are now aware of upsampling 👍
Thank you for making this so understandable
Thank you for watching, I appreciate it. :)
I do a simple test with my focal monitoring speaker and switch to youtube from Qobuz and between the AAC from youtube and the FLAC there especialy on the same album with the same master source the FLAC win hand on whaterver resolution it is especialy with a 48 khz aac 16 bit and the FLAC 48 khz 24 bit , so ok if you don't compare it's Impossibles to listen to what you realy loose , but on cymbals , high pitch and medium frequencies you loose a lot, FLAC 16 bit or 24 is more than ok at 44,1 khz the sampling rate is not that important but the bit use to digitized is more relevant and squeeze all that to lossy codec make thing worse AAC and MP3 is like compare a PNG to GIF there is some lose of color .
I busted out my Sony discman and started buying CD's again. This is all I need.
Amen. I love the Sony discmans haha. I could see myself walking around with one of those, and little kids asking what it is.
@@JoshChristiane if they ask, tell em it's a portable waffle warmer
A person who plays a copy of a master tapes from Real to real tape probably thinks it sounds much better than a CD