I can add an old data-point. Sometime in the 2000’s Linn Records was one of the first record company to issue high-res. downloads. I wanted to listen to these files not only on my libing-room audio system but also on my iPod. So when I first started buying 192-KHz 24-bit files, I used my audio editor’s re-sampler, which wasn’t bad, to down-sample to 44.1KHz and 48kHz, and a Wave Arts plug-in’s dither engine for dithering down to 16-bits, and I did a few experiments. At that time the biggest difference in sound I could hear was not the down-sampling whether to 96kHz, 88.2KHz, 48KHz of 44.1KHz. Those were very subtle differences, if any, but when the 24-bit file was dithered, the change was easily noticeable. And it was noticeable whether I dithered or just truncated. But, I could only hear any of these differences with then high-end headphones, not over my mid-fi amp and speakers. Are modern dithering engines better? Maybe, they measure superbly, but I haven’t repeated the experiment.
I'd be shocked if it was purely the change from 24 to 16 bits of dynamic range that caused an "easily" noticeable difference! But depending on how the original samples _used_ that 24 bits of resolution and exactly how the dithering was performed, I can imagine ending up with an ostensibly 16 bit sample that actually has nearly everything crushed up into only about 10 bits of range and leaving the rest mostly unused. Now that definitely would be easily noticeable!
@@jhonbus I'm sorry you find it so easy to trash my observation. I can imagine many things too, but I don't think I'm imagining that you may be thinking that I am lying about the precision of my then auditory perception. Or that I am a clumsy experimenter. Either could be true, the latter may be likely. That said, I could never repeat the observation today, my hearing has deteriorated, but unless you've done your own experiment with a reputable DAW's built in dither facility, or a high quality plug-in, I'm pretty sure I don't care what you imagine. I mentioned my observation because it went against today's prejudgement in some quarters on the relative importance of bit-depth and sample rate. If anybody cares, and I suspect nobody does, when I did the experiment I thought the audibility of the change in bit depth from 24 bits to 16 bits would be less audible than reduction in sample rate in 24-bits. My initial prejudice was shown then to be wrong. It's one experiment. BTW in my experiment I used an early predecessor of Zynaptiq's Triumph audio editor and an early Wave Arts Final Plug for dither. Today I would use the GoodHertz GoodDither 3 plug-in, which is easy to recommend and might produce a different result,
Many years ago, I purchased the Koss Electrostatic Headphones, & they were wonderful! `Only,' about $250 as I recall, a little over 50 years ago! Today, still available, list price : $1,000, recently on sale for $800. If you want or need to use headphones, I recommend these, I'm also old, but have tried to take care of my ears, & can still hear differences in clarity, distortion & detail! Koss makes a number of, `standard,' headphones, & still appears to have one Electrostatic. If you could order, try, & then return if you're not pleased, I'd recommend this! Good luck!
I have been wondering this for a while. Thank You Paul for the explanation. I will stick to ripping used CDs into Roon with EAC or buying the DSD versions and not worry about it.
Remember that the distortion is increasing with lower volume in digital audio. I used to visit HiFi-stores in the 80s and 90s using a Denon test CD for evaluating da converters on CD-players. I still have the CD. There is a track recorded at peak -60dB. Crank the volume up and it becomes easy to hear distortion from using lower peak levels. In music the weakest sound is the sound of the room. A spacious recording(church music) will sound bigger or smaller depending on how good the weak sound can be reproduced.
I believe one of the big reasons Octave Records sound "better" than other similar recordings is, the mastering process. OR care about the mastering process. They could likely release their titles only at 44.1 16-bit and MOST people would never know any better. Because they care about and pay attention during the mastering process, NOT because you can buy the files in DSD quality. I'd still choose the higher resolution version and format just in case, but really, I can't hear a difference w/ WAV/FLAC at 44.1 16 and DSD 256 in side by side comparisons when mastering is done well.
I imagine many just use the recording mode handed to them and they dont dabble. I learned a hell of a lot trying to fit high quality audio onto mp3 players (flac players). Needs to be optimised. For example i could get a full dvd down to 700mb with no loss in detail at all. Just very well tuned. I'd do the same for audio and with audio it was a lot tougher, you had to keep the file sizes high, like skipping 80% of the flac compression. Still vastly smaller file sizes than wav but theres not much to play with. I then learned a lot about brainwaves and processing over the next decade. Fairly certain nobody is optimising for brainwave sampling and synchronisation. Love it or hate it and despite having $10,000 speakers your brain is still pretty much mp3 quality. You have to respect the equipment. Were not built with 100% range on a fully analogue input. Were mp3 🤣 brain has its own (terrible) sample rate and systems for conversion. For true good quality audio you should work backwards towards the recording rate not from the ears but from the brains imaging of the audio.
Another straightforward no nonsense clear description of the topic in hand. I've promised myself that I'll finally buy a few Octave Records SACD titles in 2023.
I absolutely agree: Going from 16-bit to 24-bit makes no difference. Going from 44.1kHz to 48kHz ~ 192kHz makes a difference because it makes the trebles sound more clean & clear. (But the audio system must be able to reproduce that).
@nicksterj Actually, that's not true. That's not how sound works. There's a reason why music production mastering EQs have knobs above 20kHz. Also, I've made my own tests: I have a 24-bit 48kHz album; I've converted it to 24-bit 44.1kHz and I was able to hear a difference in trebles (48kHz sounds "brighter").
@nicksterj No. But, uplifting the trebles above 20kHz will affect the sound below 20kHz. Please do your own researches: If you have a good audio interface, or a DAC, and a pair of good speakers (not headphones), buy a high resolution (.wav) song from HDtracks, then convert it to 44.1kHz (.wav), using an audio editing program, like Sound Forge. Compare the two tracks. Then, we'll talk.
@nicksterj I hear nothing. But, that's normal because you've been downsampling the high-res tracks. Please redo the test, but now use the 44.1kHz track on top of the high-res track. Then switch the polarity of one of them.
@nicksterj Hello! I just got home. I've done my test. I am right. The 48kHz track doesn't null completely with the 44.1kHz track. I will make a video & I will post it on TH-cam in a few days.
I have a Yamaha CD/DVD player that supports SACD. You can play two channel SACD hybrid and toggle between SACD and regular CD. SACD sound is simply richer. I've got about 70 SACD CDs and enjoy them - along with about a 1,000 CDs that also sound just fine.
Higher sample rates and greater bit deth are used in the recording process to give editing headroom. Strictly speaking there are dynamic range audible differences from 16 bit to 20bit or 24 bits. In practice this difference is trivial because you can't have the conditions to produce such a dynamic range in an audio system. You would need a completely silent room (no room noise), a dac/amp chain that's capable of producing more than 115db of dynamic range (note that I say produce and not just decode) and a music recording with such dynamic range which probably doesn't exist.
And with dithering and noise shaping, standard 16-bit CD-quality audio can approach 120 dB of dynamic range across the most musically important frequencies, at the expense of adding some noise above 16 kHz that is "only" about 80 dB below full scale, which most gray-haired audiophiles can't hear anyway, and still far exceeds the normal dynamic range of recorded music.
Dont forget your brain processes the data too. A slightly unbalanceable conversion rate will give you crap sound in your head. If they havent maximised the recording to optimum (using all available data) then thats just being a cheap producer surely.
Great discussion. Expanding my knowledge base. Everything is a tradeoff. I fully acknowledge there is a sound benefit in well recorded high res recordings. Another factor is phase noise. Every improvement in a clock lowering phase noise is clearly audible on all the aspects leading to natural dynamic sound. With the best oscillators available, the lower the clock rate, the better the performance and resulting sound. So I chose to optimize my digital streamer/DAC to play 44.1 16 bit with the best possible 5MHz oscillator. I trade off the ability to play high res, but achieve more natural sound. And most of the stuff I have was recorded on CD media. When I buy audiophile recorded material, I am offered lowest price for a version down sampled to 44.1KHz. Win win.
I like the 24 bit vs 16 bit answer, I completely agree. No musical application requires more than 96 dB of dynamic range and processing 50% more data adds cost with no benefit. As for the higher sampling frequency discussion I'm more skeptical. Obviously DSD requires higher sampling frequencies than PCM. So were not comparing apples and oranges the following is just about PCM. I asked MS copilot "in digital audio why would you need to sample at a frequency higher than twice the upper boundary of human hearing?" Its reply was: Great question! In digital audio, the Nyquist Theorem states that to accurately capture and reproduce a sound, you need to sample it at least twice the highest frequency present in the sound. Since the upper boundary of human hearing is typically around 20 kHz, you would need to sample at a minimum rate of 40 kHz. However, in practice, sampling rates are often set higher than twice this upper limit for a few reasons: 1. Anti-Aliasing Filters: To avoid artifacts caused by frequencies above the Nyquist limit, anti-aliasing filters are used. These filters are not perfect and require some headroom, so a higher sampling rate helps ensure the filter performs effectively. 2. Headroom for Processing: Higher sampling rates provide more data points, which can be advantageous for audio processing tasks like mixing, editing, and effects, leading to better overall sound quality. 3. Perceptual Differences: Some argue that even though humans can't hear frequencies above 20 kHz, higher sampling rates can still improve audio fidelity and the listener's experience, due to the more accurate representation of the original analog signal. Thus, common professional audio sampling rates are 44.1 kHz, 48 kHz, and even higher like 96 kHz or 192 kHz. // Here's my take on those 3 points. 1. Yes some headroom is necessary for the anti-aliasing filters which is why the Red Book CD standard is 44.1 kHz not 40 kHz. You certainly don't need 148 kHz headroom (192 kHz - 44.1 kHz) to have filters than work well beyond human perception. 2. For mastering yes, keep the high sample rate all the way through the process but once the mastering is complete, the end user will not benefit from anything higher than 44.1 kHz. 3. Everyone is different. Maybe some people can hear dog whistles. For me, my 64 year old auditory system can't hear above 16 kHz and I doubt that in a blind test I could in any way sense those ultrasonic frequencies.
Yeah, it is a repeated question, but I'm glad because this time I paid attention and understood why I hear differences between sample rates 96 and higher compared to lower rates, but I don't hear any difference between 16 and 24 bit. Thank you.
The best dacs can effectively process around 20 bits of resolution. The format is 16 or 24 or 32 bit. The sample rate is limited to the file. R2r dacs do best with lower sample rates. Delta sigma do best with higher sample rates. Which is better? Don't care I have both and today they are both amazing. Enjoy some fucking music. But remember if you have an r2r they don't need high sample rates because they basically are an algorithm (it does crazy math in milliseconds and nanoseconds) . The delta sigma estimates what the signal will look like. That means the more it looks at the signal the better the estimation. The r2r depends on its instructions. The more accurate the resistive ladder inside it and the better its fpga chip can process math the more clarity and more precise it recreates the signal.
Clients have still asked me for squashed masters. I ask them to listen to a less squashed version and they just don’t care for it. It’s always “I need it to be as loud as ______”. Should ever take one of those unsquashed masters and play it in a venue I know it’ll blow everyone away. Upon listening one artists self master I then played a random tune which was being marketed hard. His tune was dynamic and had groove. The random song was fatiguing on the ears even though it was only 3:30 long. My recommendation was to push the limiter just a little harder so it shaved 1.5 to 2 db off and then make up that gain.
It matters if one can hear the difference and if that difference is worth paying for. I use Qobuz, and on my system the difference between 16 aand 24 bit makes such a difference that it's woth the money.
In my own experience, listening on an Astell&Kern DAP and comparing the 16 bit cd quality directly to the 24 bit with 44.1kHz there is a remarkable difference. For example, Donald Fagen - Kamikariad the separation of instruments is night and day. The drums even change location. The hi hats move from center to the left. More notable detail in the cymbals. The sticks making contact with the cymbals comes through. All of the instruments have more detail. I have found this with all of the albums that are only available in 24 bit 44.1kHz. I probably only have around 10 albums or so at 24 bit that were not available with 96 or 192 kHz, but I am happy with them and on direct comparison I do notice a significant difference. I wish everything could be 192kHz or more. However, I have no complaints. Well, I do have one complaint and this is it - they have not released Kansas Masque, Leftoverture, Point Of Know Return, Song For America, Monolith and the original Kansas album as 192kHz 24 bit digital or even DSD. This music is epic.
A greater bit-depth gives you a more cohesive sound image with better fading. I prefer listening to 24/48 or 32bit DXD. With 24bit DXD things sound disconnected -- there being a lack of faded depth to match the directional information that DXD sampling provides.
The best reason to pay extra for the 24 bit version is that sometimes that indicates an audiophile remaster or remix. Qobuz or Tidal may not mention that difference, so listen to both versions. Don’t pay for it if you don’t prefer the sound🤷♂️
Have listened to thousands of songs on Qbuz streaming to my Rossini Apex DAC. I do not know what the question asker is talking about for more expensive for higher files. I pay one monthly price and can listen as much as I want for a small fee. The 24/96 files most sound incredible. Some 16/44 sound incredible but more rare to find. I ordered octave record SACD. I did notice a difference from my regular CD"s but also there are qbuz recordings that blow away the sacd for quality. So what format it's in helps your odds of finding a good recording but not absolute. It's more of a listen and search and save when you find the gems and that is part of the fun of the process.
Most 24/44.1 albums have truly either been recorded in that quality or higher and then mastered at the quality. That is an incredibly common sample rate for popular music to be recorded at now. It is usually 24/44.1 or 24/48 with the occasional 24/96.
I duno, i can still clearly hear the difference between anything recorded in the mid 80's or before, and those recorded digitally. Im going to have to try this guys stuff and compare. But i think it kinda lost its soul when it all went digital. Something was lost.
@@luminousfractal420 anything recorded, mixed or mastered digitally below 24/88.1 I would say that there is some merit to what you are saying especially older digital recordings. Hi res digital recordings especially recorded using DSD would be incredibly hard to distinguish between analog and digital especially on vinyl.
20 bits of dynamic range represent 120dB and is the practical limit of consumer audio electronics and still well beyond what makes sense for music. Thus at least 4 of the 24 bits aren’t making much actual sense. So why not use 20 bit streaming and files? Because data formats tend to get defined in quantities of bytes and 3 bytes of 8 bits are 24 bits.
Don't break that woofer!!!🤣 Yes, as a musician myself, compression is kind of an evil. It takes so much emotion out of music, but I do understand that so many systems and people don't really appreciate the variance.
Paul I have a somewhat obscure or esoteric question: I own a (mostly) tube Audio Research system. According to an Audio magazine review from 1986 it was recommended that I leave my SP 11 on all the time. I guess this is because it is a hybrid preamp. How are the tubes "protected" in this preamp? Additionally I have a servo for my infinity IRS betas that I think I should be leaving in all the time since it does not contain tubes. What are your thoughts? Thanks, Jim PS I am slavering over your DAC equipment. One day! I will have to visit
Yes. 24 bit audio is meant to allow adjusting the gain after the fact. Is similar to how a raw image allows a photographer to not worry about white balance because it can be adjusted in the computer.
@@housepianist It kinda does because the more people argue the more it can push people away. R2r designs cost a lot more to build. Delta sigmas also cost a lot to perfect. But there's cheap ones. When you don't have the money the lowest fruit looks best and people talk shit. The price of both the r26 and a26 is almost the same. That's because most likely to make any dac sound good it costs a certain amount to overcome different challenges. Different approaches different strengths different weaknesses. I own the x26 pro and r26. I prefer r2r but Gustard has proven to me delta sigma is not shit anymore. So it is possible.
@@V1ralB1ack No one is really arguing here. It was a question about sampling rates. What you need to understand is that not everyone tunes in to this channel on a regular basis so it’s hard to know what has and hasn’t been discussed. I suppose one could search through all the videos that Paul has been done to see if *maybe* a similar question has been asked but no one is going to do that. And besides, it kind of adds a bit of acknowledgement if someone’s question is selected and answered. Lastly, no one is addressing your specific inquiry so I’m not sure what that has to do with anything. My point was simply stating that these kinds of topics will come up again and again. That’s OK. If people are pushed away, so be it. But I tend to doubt that. After all, you’re here! 😂
If a device (streamer / DAC) benifits from a higher sample rate (or bit depth) above CD quality, it's more likely that the device can benifit more from technical improvement, was about the way Hans Beekhuyzen (on youtube) put it recently. I agree. I cannot judge DSD, as my streamer needs to convert it to CD quality, as my DAC will at the most play 48/24. Octave does make fine recordings which sound great at CD quality, yet lacks tracks that I like. And really, most recordings sound great, like Radiohead The Bends, which sounds remarkable, yet is challenging for equipment.
Don't laugh when I tell you I have a $70 DAC. I'm a poor man. When I purchased it and plugged a 1990s Sony CD player into it, the sound is much better than the internal DAC. Someone on the net stated that old tech CD players are rated 16/41, but don't match up to the resolution standard and are actually around 13 bit. This is the reason many proclaim the necessity of 24.
I've noticed more than a few files on Qobuz at 24 bit 44 1kHz. I've got no comment on sound quality- just that I have noticed those files are on Qobuz and it's curious.
Oh Paul - your arguments and assumptions here are really frustrating. Your "Where did they get it" question is an issue for me - casting aspersions on Qobuz and other record companies is problematic. Most modern studios record at 24bit/96kHz, then it gets mastered for CD release - why shouldn't the record companies also make those original files available for those that want them? CD is a compromise - and as most people don't even own CD players the idea of having music released via downloads as "CD quality" makes zero sense. Wherever possible I'll buy the high resolution file, because why wouldn't I?
i Can d efinitely hear the difference between 16 bit 44.1 and 24 bit 96khz. Even in my poorly treated room with my diy speakers. I think beyond that there's no discernable difference though
16 bits or 96 dB is more than needed for any music. Paul mentions dynamic range in the 70s for his audiophile recordings. Also, why push peaks to 0 dB? Have peaks at -3 to -6dB. I see pop recordings with 10 dB dynamic range and almost continuous intersample overs. It's loudness wars and total crap.
the theory is the stronger the signal, the further away from background noise, aka high s/n ratio but doing so by makeup gain after heavy compression simply destroys original dynamic range
Engineers will tell you is the only difference is dynamic range by just adding 24db of dynamic range. Although this is true, what they don't tell you is the number of bits starting at the quietest db to the loudest is exponential. The number of digital steps in 24bit towards the louder end of the exponential graph is much much more. Both the number of digital steps and the volume change of the db the louder the signal get is exponential thus the confusion about dynamic range. Every 3db takes about 2x the power and every 6db is about twice the volume. 24bit has 16777216 digital steps & 16bit is only 65536. Because both db and number of steps per additional bit are both exponential the result is about 6db per bit but the loudest db in 24bit has massively more digital steps. There is an advantage of up scaling 44.1khz to 96khz on certain DAC's and digital equipment to tighten up the clock and have less clock jitter on playback. Less clock jitter on playback impruves imaging and isolation in the stereo image, especially on modest systems, even though there is no frequency advantage. Wether someone needs higher quality files or components is entirely up to the listener and what the listener can hear in their enviroent and system.
I was always under the impression that 16/24 was the steps of the signal. Like analog but it's sawtooth so 24 bits would be a smoother rise in the amplitude. Is that not true?
"Pearl Acoustics" (UK) has a great video actually corroborating everything Paul is saying, and giving more detail and info, along with reference sample music to download and test with. The main topic revolves around "vinyl or digital", but with a lot more accurate information than CD bashing, or Vinyl bashing. th-cam.com/video/MlccCTy4PiQ/w-d-xo.html
I can only speak for my setup, but I have a dedicated FLAC player and there is a huge chasm between 16-bit and 24-bit. The 24-bit will beat every 16-bit FLAC file out there in my opinion. I have run my FLAC player through my car audio at times, and when a 16-bit file comes out, it is obvious that the file lacks in comparison to a 24-bit. I believe this is the same video that Paul released a couple of weeks ago, so my same comment may be further down the section.
Yes it is because anything digital has to be converted to analogue to allow you to listen to the music. That's why every digital format needs a DAC. Your speakers are electromechaical transducers that vibrate the air we breath, so without analogue, digital music would not exist. Simple.
I feel like an idiot playing background music for a party. The wife asks why my fancy stereo can't play soft easy listening music. I respond by saying I'll just ask the musicians sometime.
HDTracks and similar websites are generally overpriced and not worth it IMHO. Unless the music you're looking for is hard to source and/or you want to burn it to a CD.
The high bitrates need tying in with the brains natural frequency. Multiples or divisions like radio antennaes. Not for any recieving but for the processing our brain does. It takes samples much like a digital track and fills in the gaps with our inbuilt ai. Same way our eyes only take in about 5% of what we think we see and the rest is created in house within the brain. If those samples the brain works with are delivered at a rate that allows smooth processing we will get something that sounds completely natural. If its off your going to get clunky sound. Dont forget the last bit of sampling equipment in the process..the brain. Were a digital system ourselves. Chemical system but digital in its processing nature. Avoid heavy use of 20-21000hz (sorry i forget the exact frequency, was 20years ago) too as that can cause brain damage at high volumes, notch filters 👍. The last stage isnt your ears. it gets processed by the brain and fed to us. Thats where you should start when setting up a recording. Binaural whatever...for a crisp reproduction you need it tuned to what the brain can do. Not your ears
For me, life starts at 96khz, those 24 bit and 16 bit... Meh. I do sometimes feel that when editing existing audio files, the 24 bit does help as opposed to the 16 bit. And I do indeed hear that 'studio standard' 24/48 sounds better than 16/44. Not nearly good enough, but it's noticeable. But generally, CD quality is just not enough for me. I do wonder where someone finds 24/44.1 files though, never saw one. It's always 24/48 or 24/more than 48.
I have seen this exact video a week ago. Uploaded twice by mistake perhaps? If this is not the case, it would imply that my understanding of life is true, haha 😅
@@squicker 44.1kHz sampling does not come anywhere close to covering the human hearing because it does not account for the fact that we can hear many different soundwaves happening at the same time - and we need to sample them individually to capture the full audible dimension of the sounds we can hear. This is the true reason DSD sounds so much more impressive than standard CD PCM 16b/44.1kHz stuff.
So here is the rube. No difference between 16 and 24 you say. But A 10,000 buck power cable make all the difference. Really time to come back to reality. There really is a lot of junk being sold for big dollars that do nothing but look better. I feel super high end do a few things. Look great, build quality is amazing and impart a beyond nature sounding system. That is all.
If you have a CD ripped into 16bit or 24, you will hear almost no differences. with 24 bits, more data goes per second from server to user, and therefore 24 bit is more expensive. But is the music 100% uncompressed from the start and which codec compression do they use? That is a better question!
You won't hear the difference - the 96dB of 16 bits is the full range of dynamics the ear can hear. A differential test between 16 bit and 24 bit would have to be amplified a hell of a lot to produce any audible sound output at all.
The loudness war in digital mastering adds up to stupidity. Hopefully this is not a phenomena in classical or jazz recording, in which dynamics are a major component in such music.
Whoa whoa whoa! What kind of crazy talk is that, sir?!? *NOTHING* is a waste of money when it comes to acquiring better audio/gear! Of course, if you can spend even more and upgrade that 44.1 to 88.2, or 196.4, then it's even better! 📈
Well, I have very picky and sensitive ears, so every effort to satisfy them is worth the outrageous marriage-threatening cost. Besides, after so many threats, if my wife didn't divorce me by now, she'll be probably going to stick around! 👍
@Douglas Blake That's what my friends said about my gold-plated demagnetized optical fiber cables. I guess they're just not that much of a critical listener as myself... ( -.-)
No dsd ain’t dramatically better than 16/44.1. Aes 2007 reported nobody can perceive higher than cd quality. Going higher will lead to deafness quickly. Do a double blind test and let’s see if you can hear higher than 16/44.1. All this talk of 5.6mhz is complete nonsense. Nobody can perceive more than 16/44.1. Absolute rubbish that 96khz is a minimum
When I record at 24 bit (usually 88k) I believe it sounds ‘Smoother’ to me. Perhaps some is the plug-ins that I use, maybe having more info to work with the end result is better? Just like having ‘excessive’ sample rate, I can see where having excessive dynamic range could give more information to work with- especially when editing.
When you "record" content at "24-bit" therefore a file is produced, how do you really know it's a 24-bit file? It may sound better but is the file really 24-bit? how do you check that?
@@wilcalint when I’ve done so, it was with ProTools and I simply selected 24bit for the setup. Then set the sample rate of course. To check? ProTools will say what bit depth and sample rate is if you ‘inspect’ the file.
Paul nailed this one including how most modern sound engineers dynamically crush modern music as part of the loudness war.
I can add an old data-point. Sometime in the 2000’s Linn Records was one of the first record company to issue high-res. downloads. I wanted to listen to these files not only on my libing-room audio system but also on my iPod. So when I first started buying 192-KHz 24-bit files, I used my audio editor’s re-sampler, which wasn’t bad, to down-sample to 44.1KHz and 48kHz, and a Wave Arts plug-in’s dither engine for dithering down to 16-bits, and I did a few experiments. At that time the biggest difference in sound I could hear was not the down-sampling whether to 96kHz, 88.2KHz, 48KHz of 44.1KHz. Those were very subtle differences, if any, but when the 24-bit file was dithered, the change was easily noticeable. And it was noticeable whether I dithered or just truncated. But, I could only hear any of these differences with then high-end headphones, not over my mid-fi amp and speakers. Are modern dithering engines better? Maybe, they measure superbly, but I haven’t repeated the experiment.
I'd be shocked if it was purely the change from 24 to 16 bits of dynamic range that caused an "easily" noticeable difference! But depending on how the original samples _used_ that 24 bits of resolution and exactly how the dithering was performed, I can imagine ending up with an ostensibly 16 bit sample that actually has nearly everything crushed up into only about 10 bits of range and leaving the rest mostly unused. Now that definitely would be easily noticeable!
@@jhonbus I'm sorry you find it so easy to trash my observation. I can imagine many things too, but I don't think I'm imagining that you may be thinking that I am lying about the precision of my then auditory perception. Or that I am a clumsy experimenter. Either could be true, the latter may be likely. That said, I could never repeat the observation today, my hearing has deteriorated, but unless you've done your own experiment with a reputable DAW's built in dither facility, or a high quality plug-in, I'm pretty sure I don't care what you imagine. I mentioned my observation because it went against today's prejudgement in some quarters on the relative importance of bit-depth and sample rate. If anybody cares, and I suspect nobody does, when I did the experiment I thought the audibility of the change in bit depth from 24 bits to 16 bits would be less audible than reduction in sample rate in 24-bits. My initial prejudice was shown then to be wrong. It's one experiment. BTW in my experiment I used an early predecessor of Zynaptiq's Triumph audio editor and an early Wave Arts Final Plug for dither. Today I would use the GoodHertz GoodDither 3 plug-in, which is easy to recommend and might produce a different result,
Many years ago, I purchased the Koss Electrostatic Headphones, & they were wonderful! `Only,' about $250 as I recall, a little over 50 years ago! Today, still available, list price : $1,000, recently on sale for $800. If you want or need to use headphones, I recommend these, I'm also old, but have tried to take care of my ears, & can still hear differences in clarity, distortion & detail! Koss makes a number of, `standard,' headphones, & still appears to have one Electrostatic. If you could order, try, & then return if you're not pleased, I'd recommend this! Good luck!
I have been wondering this for a while. Thank You Paul for the explanation. I will stick to ripping used CDs into Roon with EAC or buying the DSD versions and not worry about it.
Remember that the distortion is increasing with lower volume in digital audio. I used to visit HiFi-stores in the 80s and 90s using a Denon test CD for evaluating da converters on CD-players. I still have the CD. There is a track recorded at peak -60dB. Crank the volume up and it becomes easy to hear distortion from using lower peak levels. In music the weakest sound is the sound of the room. A spacious recording(church music) will sound bigger or smaller depending on how good the weak sound can be reproduced.
I believe one of the big reasons Octave Records sound "better" than other similar recordings is, the mastering process.
OR care about the mastering process.
They could likely release their titles only at 44.1 16-bit and MOST people would never know any better. Because they care about and pay attention during the mastering process, NOT because you can buy the files in DSD quality.
I'd still choose the higher resolution version and format just in case, but really, I can't hear a difference w/ WAV/FLAC at 44.1 16 and DSD 256 in side by side comparisons when mastering is done well.
It's only the mastering you hear - never it's SACD or sampling rate
I imagine many just use the recording mode handed to them and they dont dabble.
I learned a hell of a lot trying to fit high quality audio onto mp3 players (flac players). Needs to be optimised. For example i could get a full dvd down to 700mb with no loss in detail at all. Just very well tuned. I'd do the same for audio and with audio it was a lot tougher, you had to keep the file sizes high, like skipping 80% of the flac compression. Still vastly smaller file sizes than wav but theres not much to play with. I then learned a lot about brainwaves and processing over the next decade. Fairly certain nobody is optimising for brainwave sampling and synchronisation. Love it or hate it and despite having $10,000 speakers your brain is still pretty much mp3 quality. You have to respect the equipment. Were not built with 100% range on a fully analogue input. Were mp3 🤣 brain has its own (terrible) sample rate and systems for conversion. For true good quality audio you should work backwards towards the recording rate not from the ears but from the brains imaging of the audio.
Another straightforward no nonsense clear description of the topic in hand. I've promised myself that I'll finally buy a few Octave Records SACD titles in 2023.
For the author it's a waste of money, for everyone who hears a difference it's more than worth it.
I absolutely agree:
Going from 16-bit to 24-bit makes no difference.
Going from 44.1kHz to 48kHz ~ 192kHz makes a difference because it makes the trebles sound more clean & clear.
(But the audio system must be able to reproduce that).
@nicksterj
Actually, that's not true.
That's not how sound works.
There's a reason why music production mastering EQs have knobs above 20kHz.
Also, I've made my own tests:
I have a 24-bit 48kHz album; I've converted it to 24-bit 44.1kHz and I was able to hear a difference in trebles
(48kHz sounds "brighter").
@nicksterj
No.
But, uplifting the trebles above 20kHz will affect the sound below 20kHz.
Please do your own researches:
If you have a good audio interface, or a DAC, and a pair of good speakers (not headphones), buy a high resolution (.wav) song from HDtracks, then convert it to 44.1kHz (.wav), using an audio editing program, like Sound Forge.
Compare the two tracks.
Then, we'll talk.
@nicksterj
I hear nothing.
But, that's normal because you've been downsampling the high-res tracks.
Please redo the test, but now use the 44.1kHz track on top of the high-res track.
Then switch the polarity of one of them.
@nicksterj
Alright.
I'll make a nulltest myself.
I'll get back to you with my results.
@nicksterj
Hello!
I just got home.
I've done my test.
I am right.
The 48kHz track doesn't null completely with the 44.1kHz track.
I will make a video & I will post it on TH-cam in a few days.
I have a Yamaha CD/DVD player that supports SACD. You can play two channel SACD hybrid and toggle between SACD and regular CD. SACD sound is simply richer. I've got about 70 SACD CDs and enjoy them - along with about a 1,000 CDs that also sound just fine.
You impart so much wisdom Paul. The Gandalf of audio wizardry.
Higher sample rates and greater bit deth are used in the recording process to give editing headroom.
Strictly speaking there are dynamic range audible differences from 16 bit to 20bit or 24 bits. In practice this difference is trivial because you can't have the conditions to produce such a dynamic range in an audio system. You would need a completely silent room (no room noise), a dac/amp chain that's capable of producing more than 115db of dynamic range (note that I say produce and not just decode) and a music recording with such dynamic range which probably doesn't exist.
And with dithering and noise shaping, standard 16-bit CD-quality audio can approach 120 dB of dynamic range across the most musically important frequencies, at the expense of adding some noise above 16 kHz that is "only" about 80 dB below full scale, which most gray-haired audiophiles can't hear anyway, and still far exceeds the normal dynamic range of recorded music.
@@vwestlife yes, CD 16/44.1 is capable of exceeding the abilities of human hearing. It is all down to the Producer at the time of recording.
Dont forget your brain processes the data too. A slightly unbalanceable conversion rate will give you crap sound in your head. If they havent maximised the recording to optimum (using all available data) then thats just being a cheap producer surely.
Great discussion. Expanding my knowledge base.
Everything is a tradeoff. I fully acknowledge there is a sound benefit in well recorded high res recordings.
Another factor is phase noise. Every improvement in a clock lowering phase noise is clearly audible on all the aspects leading to natural dynamic sound. With the best oscillators available, the lower the clock rate, the better the performance and resulting sound.
So I chose to optimize my digital streamer/DAC to play 44.1 16 bit with the best possible 5MHz oscillator. I trade off the ability to play high res, but achieve more natural sound.
And most of the stuff I have was recorded on CD media. When I buy audiophile recorded material, I am offered lowest price for a version down sampled to 44.1KHz. Win win.
I like the 24 bit vs 16 bit answer, I completely agree. No musical application requires more than 96 dB of dynamic range and processing 50% more data adds cost with no benefit.
As for the higher sampling frequency discussion I'm more skeptical. Obviously DSD requires higher sampling frequencies than PCM. So were not comparing apples and oranges the following is just about PCM.
I asked MS copilot "in digital audio why would you need to sample at a frequency higher than twice the upper boundary of human hearing?"
Its reply was:
Great question! In digital audio, the Nyquist Theorem states that to accurately capture and reproduce a sound, you need to sample it at least twice the highest frequency present in the sound. Since the upper boundary of human hearing is typically around 20 kHz, you would need to sample at a minimum rate of 40 kHz.
However, in practice, sampling rates are often set higher than twice this upper limit for a few reasons:
1. Anti-Aliasing Filters: To avoid artifacts caused by frequencies above the Nyquist limit, anti-aliasing filters are used. These filters are not perfect and require some headroom, so a higher sampling rate helps ensure the filter performs effectively.
2. Headroom for Processing: Higher sampling rates provide more data points, which can be advantageous for audio processing tasks like mixing, editing, and effects, leading to better overall sound quality.
3. Perceptual Differences: Some argue that even though humans can't hear frequencies above 20 kHz, higher sampling rates can still improve audio fidelity and the listener's experience, due to the more accurate representation of the original analog signal.
Thus, common professional audio sampling rates are 44.1 kHz, 48 kHz, and even higher like 96 kHz or 192 kHz.
//
Here's my take on those 3 points.
1. Yes some headroom is necessary for the anti-aliasing filters which is why the Red Book CD standard is 44.1 kHz not 40 kHz. You certainly don't need 148 kHz headroom (192 kHz - 44.1 kHz) to have filters than work well beyond human perception.
2. For mastering yes, keep the high sample rate all the way through the process but once the mastering is complete, the end user will not benefit from anything higher than 44.1 kHz.
3. Everyone is different. Maybe some people can hear dog whistles. For me, my 64 year old auditory system can't hear above 16 kHz and I doubt that in a blind test I could in any way sense those ultrasonic frequencies.
Yeah, it is a repeated question, but I'm glad because this time I paid attention and understood why I hear differences between sample rates 96 and higher compared to lower rates, but I don't hear any difference between 16 and 24 bit. Thank you.
The best dacs can effectively process around 20 bits of resolution. The format is 16 or 24 or 32 bit. The sample rate is limited to the file. R2r dacs do best with lower sample rates. Delta sigma do best with higher sample rates. Which is better? Don't care I have both and today they are both amazing. Enjoy some fucking music. But remember if you have an r2r they don't need high sample rates because they basically are an algorithm (it does crazy math in milliseconds and nanoseconds) . The delta sigma estimates what the signal will look like. That means the more it looks at the signal the better the estimation. The r2r depends on its instructions. The more accurate the resistive ladder inside it and the better its fpga chip can process math the more clarity and more precise it recreates the signal.
This was really good. Thanks finally someone to make it clear.
I have a host of song on cds and when I hear them on movie soundtracks they sound so much better!. The setting is at home using the same equipment.
An octave recording blows away all but a few of my others. Hands down and no comparison.
Thanks, Lonnie!
Clients have still asked me for squashed masters. I ask them to listen to a less squashed version and they just don’t care for it. It’s always “I need it to be as loud as ______”. Should ever take one of those unsquashed masters and play it in a venue I know it’ll blow everyone away.
Upon listening one artists self master I then played a random tune which was being marketed hard. His tune was dynamic and had groove. The random song was fatiguing on the ears even though it was only 3:30 long. My recommendation was to push the limiter just a little harder so it shaved 1.5 to 2 db off and then make up that gain.
It matters if one can hear the difference and if that difference is worth paying for. I use Qobuz, and on my system the difference between 16 aand 24 bit makes such a difference that it's woth the money.
Very clear😊😊😊as usual🎉🎉🎉as always😊😊😊
In my own experience, listening on an Astell&Kern DAP and comparing the 16 bit cd quality directly to the 24 bit with 44.1kHz there is a remarkable difference. For example, Donald Fagen - Kamikariad the separation of instruments is night and day. The drums even change location. The hi hats move from center to the left. More notable detail in the cymbals. The sticks making contact with the cymbals comes through. All of the instruments have more detail. I have found this with all of the albums that are only available in 24 bit 44.1kHz. I probably only have around 10 albums or so at 24 bit that were not available with 96 or 192 kHz, but I am happy with them and on direct comparison I do notice a significant difference. I wish everything could be 192kHz or more. However, I have no complaints. Well, I do have one complaint and this is it - they have not released Kansas Masque, Leftoverture, Point Of Know Return, Song For America, Monolith and the original Kansas album as 192kHz 24 bit digital or even DSD. This music is epic.
...even if upsampled, there is an advantage in mixing at 96k.
A greater bit-depth gives you a more cohesive sound image with better fading. I prefer listening to 24/48 or 32bit DXD. With 24bit DXD things sound disconnected -- there being a lack of faded depth to match the directional information that DXD sampling provides.
Wasn't this posted a few weeks ago? 🤔
Yes, on the Octave channel. Sometimes there's crossover.
@@Paulmcgowanpsaudio Thank you for clarifying, I didn't notice it was another channel. 😄
The best reason to pay extra for the 24 bit version is that sometimes that indicates an audiophile remaster or remix. Qobuz or Tidal may not mention that difference, so listen to both versions. Don’t pay for it if you don’t prefer the sound🤷♂️
What a masterclass.
Have listened to thousands of songs on Qbuz streaming to my Rossini Apex DAC. I do not know what the question asker is talking about for more expensive for higher files. I pay one monthly price and can listen as much as I want for a small fee. The 24/96 files most sound incredible. Some 16/44 sound incredible but more rare to find. I ordered octave record SACD. I did notice a difference from my regular CD"s but also there are qbuz recordings that blow away the sacd for quality. So what format it's in helps your odds of finding a good recording but not absolute. It's more of a listen and search and save when you find the gems and that is part of the fun of the process.
He may have been commenting on files that you purchase on Qobuz
Digital audio has lots and lots of little bits. And little 1s and 0s too. So lovely!
Most 24/44.1 albums have truly either been recorded in that quality or higher and then mastered at the quality. That is an incredibly common sample rate for popular music to be recorded at now. It is usually 24/44.1 or 24/48 with the occasional 24/96.
I duno, i can still clearly hear the difference between anything recorded in the mid 80's or before, and those recorded digitally. Im going to have to try this guys stuff and compare. But i think it kinda lost its soul when it all went digital. Something was lost.
@@luminousfractal420 anything recorded, mixed or mastered digitally below 24/88.1 I would say that there is some merit to what you are saying especially older digital recordings. Hi res digital recordings especially recorded using DSD would be incredibly hard to distinguish between analog and digital especially on vinyl.
20 bits of dynamic range represent 120dB and is the practical limit of consumer audio electronics and still well beyond what makes sense for music. Thus at least 4 of the 24 bits aren’t making much actual sense.
So why not use 20 bit streaming and files? Because data formats tend to get defined in quantities of bytes and 3 bytes of 8 bits are 24 bits.
Don't break that woofer!!!🤣
Yes, as a musician myself, compression is kind of an evil. It takes so much emotion out of music, but I do understand that so many systems and people don't really appreciate the variance.
Paul I have a somewhat obscure or esoteric question: I own a (mostly) tube Audio Research system. According to an Audio magazine review from 1986 it was recommended that I leave my SP 11 on all the time. I guess this is because it is a hybrid preamp. How are the tubes "protected" in this preamp? Additionally I have a servo for my infinity IRS betas that I think I should be leaving in all the time since it does not contain tubes. What are your thoughts?
Thanks,
Jim
PS I am slavering over your DAC equipment. One day! I will have to visit
Great informative and interesting video regarding bit rates and quality of audio regards mark
Yes. 24 bit audio is meant to allow adjusting the gain after the fact. Is similar to how a raw image allows a photographer to not worry about white balance because it can be adjusted in the computer.
Oh no we're sampling this topic again🤦♂️
Yep. And it’ll come up again. Never hurts to refresh the memory about such things.
@@housepianist It kinda does because the more people argue the more it can push people away. R2r designs cost a lot more to build. Delta sigmas also cost a lot to perfect. But there's cheap ones. When you don't have the money the lowest fruit looks best and people talk shit. The price of both the r26 and a26 is almost the same. That's because most likely to make any dac sound good it costs a certain amount to overcome different challenges. Different approaches different strengths different weaknesses. I own the x26 pro and r26. I prefer r2r but Gustard has proven to me delta sigma is not shit anymore. So it is possible.
@@V1ralB1ack No one is really arguing here. It was a question about sampling rates. What you need to understand is that not everyone tunes in to this channel on a regular basis so it’s hard to know what has and hasn’t been discussed. I suppose one could search through all the videos that Paul has been done to see if *maybe* a similar question has been asked but no one is going to do that. And besides, it kind of adds a bit of acknowledgement if someone’s question is selected and answered.
Lastly, no one is addressing your specific inquiry so I’m not sure what that has to do with anything. My point was simply stating that these kinds of topics will come up again and again. That’s OK. If people are pushed away, so be it. But I tend to doubt that. After all, you’re here! 😂
@@V1ralB1ack what has all the bullshit to do with the topic?
I have that dynamic feature turned OFF in my settings.
Apple has a setting for that for more than a decade you can toggle on/off.
If a device (streamer / DAC) benifits from a higher sample rate (or bit depth) above CD quality, it's more likely that the device can benifit more from technical improvement, was about the way Hans Beekhuyzen (on youtube) put it recently. I agree. I cannot judge DSD, as my streamer needs to convert it to CD quality, as my DAC will at the most play 48/24. Octave does make fine recordings which sound great at CD quality, yet lacks tracks that I like. And really, most recordings sound great, like Radiohead The Bends, which sounds remarkable, yet is challenging for equipment.
very helpful, thx
Don't laugh when I tell you I have a $70 DAC.
I'm a poor man.
When I purchased it and plugged a 1990s Sony CD player into it, the sound is much better than the internal DAC.
Someone on the net stated that old tech CD players are rated 16/41, but don't match up to the resolution standard and are actually around 13 bit.
This is the reason many proclaim the necessity of 24.
I've noticed more than a few files on Qobuz at 24 bit 44 1kHz. I've got no comment on sound quality- just that I have noticed those files are on Qobuz and it's curious.
Oh Paul - your arguments and assumptions here are really frustrating. Your "Where did they get it" question is an issue for me - casting aspersions on Qobuz and other record companies is problematic. Most modern studios record at 24bit/96kHz, then it gets mastered for CD release - why shouldn't the record companies also make those original files available for those that want them?
CD is a compromise - and as most people don't even own CD players the idea of having music released via downloads as "CD quality" makes zero sense. Wherever possible I'll buy the high resolution file, because why wouldn't I?
Hi Paul. Why did you have to re-post? TH-cam issues?
Trying to sell his wares
I find 24/96 tracks to sound the best in my set up
And 10 out of 10 hard drive manufacturer's would agree too. Char Ching.
i Can d efinitely hear the difference between 16 bit 44.1 and 24 bit 96khz. Even in my poorly treated room with my diy speakers. I think beyond that there's no discernable difference though
16 bits or 96 dB is more than needed for any music. Paul mentions dynamic range in the 70s for his audiophile recordings. Also, why push peaks to 0 dB? Have peaks at -3 to -6dB. I see pop recordings with 10 dB dynamic range and almost continuous intersample overs. It's loudness wars and total crap.
the theory is the stronger the signal, the further away from background noise, aka high s/n ratio
but doing so by makeup gain after heavy compression simply destroys original dynamic range
Paul, sell some of those subwoofer drivers separately. There are still some home brew subwoofer builders out there.
LOL. This video plays at a max resolution of 360p. What happened to your video, Paul?
Engineers will tell you is the only difference is dynamic range by just adding 24db of dynamic range. Although this is true, what they don't tell you is the number of bits starting at the quietest db to the loudest is exponential. The number of digital steps in 24bit towards the louder end of the exponential graph is much much more. Both the number of digital steps and the volume change of the db the louder the signal get is exponential thus the confusion about dynamic range. Every 3db takes about 2x the power and every 6db is about twice the volume.
24bit has 16777216 digital steps & 16bit is only 65536. Because both db and number of steps per additional bit are both exponential the result is about 6db per bit but the loudest db in 24bit has massively more digital steps.
There is an advantage of up scaling 44.1khz to 96khz on certain DAC's and digital equipment to tighten up the clock and have less clock jitter on playback.
Less clock jitter on playback impruves imaging and isolation in the stereo image, especially on modest systems, even though there is no frequency advantage.
Wether someone needs higher quality files or components is entirely up to the listener and what the listener can hear in their enviroent and system.
I was always under the impression that 16/24 was the steps of the signal. Like analog but it's sawtooth so 24 bits would be a smoother rise in the amplitude. Is that not true?
@@sub-vibes Appreciated..Thank you
@@sub-vibes Interesting i have to try it, Thanks...
A digital signal is actually analogue electrically.
@@Douglas_Blake how is the transmitting of DSD done, if I interprete your post correctly it is modulated on a DC offset?
@@Douglas_Blake then, how does the DSD stream gets transmitted, if no modulation of a wave or offset is involved?
"Pearl Acoustics" (UK) has a great video actually corroborating everything Paul is saying, and giving more detail and info, along with reference sample music to download and test with. The main topic revolves around "vinyl or digital", but with a lot more accurate information than CD bashing, or Vinyl bashing. th-cam.com/video/MlccCTy4PiQ/w-d-xo.html
I have the same style of speaker as he uses, single driver transmission line, great imaging
Interesting video btw
This video is 360p, it needs to be upsampled...thought I had cattaracts kicking in
I can only speak for my setup, but I have a dedicated FLAC player and there is a huge chasm between 16-bit and 24-bit. The 24-bit will beat every 16-bit FLAC file out there in my opinion. I have run my FLAC player through my car audio at times, and when a 16-bit file comes out, it is obvious that the file lacks in comparison to a 24-bit. I believe this is the same video that Paul released a couple of weeks ago, so my same comment may be further down the section.
Yes it is because anything digital has to be converted to analogue to allow you to listen to the music. That's why every digital format needs a DAC. Your speakers are electromechaical transducers that vibrate the air we breath, so without analogue, digital music would not exist. Simple.
I feel like an idiot playing background music for a party. The wife asks why my fancy stereo can't play soft easy listening music. I respond by saying I'll just ask the musicians sometime.
HDTracks and similar websites are generally overpriced and not worth it IMHO. Unless the music you're looking for is hard to source and/or you want to burn it to a CD.
The high bitrates need tying in with the brains natural frequency. Multiples or divisions like radio antennaes. Not for any recieving but for the processing our brain does. It takes samples much like a digital track and fills in the gaps with our inbuilt ai. Same way our eyes only take in about 5% of what we think we see and the rest is created in house within the brain. If those samples the brain works with are delivered at a rate that allows smooth processing we will get something that sounds completely natural. If its off your going to get clunky sound. Dont forget the last bit of sampling equipment in the process..the brain. Were a digital system ourselves. Chemical system but digital in its processing nature.
Avoid heavy use of 20-21000hz (sorry i forget the exact frequency, was 20years ago) too as that can cause brain damage at high volumes, notch filters 👍. The last stage isnt your ears. it gets processed by the brain and fed to us. Thats where you should start when setting up a recording. Binaural whatever...for a crisp reproduction you need it tuned to what the brain can do. Not your ears
All to get rid of the distortions and artifacts caused by the perfect storage system?
caused by the analogue reconstruction filter after DAC output
@@philiptong4978 As required by this perfect system?
consider food frozen in the freezer vs cooked, hot on a plate ready to eat
storage is not the same as consumption (playback for audio data)
@@philiptong4978 Heating can harm some foods and there are foods that are eaten frozen as well. So what's the point?
If you have the option... Record at 24/48 and be happy
For me, life starts at 96khz, those 24 bit and 16 bit... Meh. I do sometimes feel that when editing existing audio files, the 24 bit does help as opposed to the 16 bit. And I do indeed hear that 'studio standard' 24/48 sounds better than 16/44. Not nearly good enough, but it's noticeable. But generally, CD quality is just not enough for me.
I do wonder where someone finds 24/44.1 files though, never saw one. It's always 24/48 or 24/more than 48.
There are many 44.1/24 files. I use an 18 bit DAC, so it will then play 44.1/18.
On torrents
Today’s show is a re-run from a few weeks ago.
I have seen this exact video a week ago. Uploaded twice by mistake perhaps? If this is not the case, it would imply that my understanding of life is true, haha 😅
I have a high end system, cost's around 25k and i cant hear any difference between 16 bit 44.1 khz and 24 bit 192 khz. No difference at all...
Interesting. I hear the difference and I háve múch more cheaper HIFI system.
@@Milo_Molnar That's because high-end audio gear upsamples low-res audio.
OK, now get your hands on some high-quality DSD recordings and report back.
@@squicker 44.1kHz sampling does not come anywhere close to covering the human hearing because it does not account for the fact that we can hear many different soundwaves happening at the same time - and we need to sample them individually to capture the full audible dimension of the sounds we can hear.
This is the true reason DSD sounds so much more impressive than standard CD PCM 16b/44.1kHz stuff.
1:40 It "HAD" a 4 inch voice coil... Before you dropped it Paul 😂😂🤣
LOL... as if Paul at his age can hear a differences 🤣
Exactly. The guy is just somehow Supernatural.
Paul acknowledged that his hearing isn't the same as when he was in his 20's-40's.
Nice woofer!
It appears to me this video needs a higher bitrate. The max resolution I can select for on my home fiber connection is 360p.
So here is the rube. No difference between 16 and 24 you say. But A 10,000 buck power cable make all the difference. Really time to come back to reality. There really is a lot of junk being sold for big dollars that do nothing but look better. I feel super high end do a few things. Look great, build quality is amazing and impart a beyond nature sounding system. That is all.
If you have a CD ripped into 16bit or 24, you will hear almost no differences. with 24 bits, more data goes per second from server to user, and therefore 24 bit is more expensive. But is the music 100% uncompressed from the start and which codec compression do they use? That is a better question!
This video is a repeat to a video a few weeks back.
quertly ! I hope not
You won't hear the difference - the 96dB of 16 bits is the full range of dynamics the ear can hear. A differential test between 16 bit and 24 bit would have to be amplified a hell of a lot to produce any audible sound output at all.
Literally nobody commenting on the giant subwoofer !!!
Get a better dac
...and a better treated room.
The loudness war in digital mastering adds up to stupidity.
Hopefully this is not a phenomena in classical or jazz recording, in which dynamics are a major component in such music.
Is 360p a waste of FCPs and YTs capabilities? 😁
(Yeah, I know. Just another glitch in the system.)
can't wait to see how the subwoofer will be! 😆
_repeat_
Yes it's a waste 😂
Whoa whoa whoa! What kind of crazy talk is that, sir?!? *NOTHING* is a waste of money when it comes to acquiring better audio/gear! Of course, if you can spend even more and upgrade that 44.1 to 88.2, or 196.4, then it's even better! 📈
Well, I have very picky and sensitive ears, so every effort to satisfy them is worth the outrageous marriage-threatening cost. Besides, after so many threats, if my wife didn't divorce me by now, she'll be probably going to stick around! 👍
@Douglas Blake That's what my friends said about my gold-plated demagnetized optical fiber cables. I guess they're just not that much of a critical listener as myself... ( -.-)
@@fsmoura
Hollywood should make a movie.
“When Audiophiles Attack”
No dsd ain’t dramatically better than 16/44.1. Aes 2007 reported nobody can perceive higher than cd quality. Going higher will lead to deafness quickly. Do a double blind test and let’s see if you can hear higher than 16/44.1. All this talk of 5.6mhz is complete nonsense. Nobody can perceive more than 16/44.1. Absolute rubbish that 96khz is a minimum
When I record at 24 bit (usually 88k) I believe it sounds ‘Smoother’ to me. Perhaps some is the plug-ins that I use, maybe having more info to work with the end result is better? Just like having ‘excessive’ sample rate, I can see where having excessive dynamic range could give more information to work with- especially when editing.
When you "record" content at "24-bit" therefore a file is produced, how do you really know it's a 24-bit file? It may sound better but is the file really 24-bit? how do you check that?
44,1Khz 16-bit is 96db of dynamic range. That's a lot. Paul even says so on this video.
I tend to agree. I haven't done a proper test, but going from 16 to 24 bits seems to make more of a difference than doubling the sample rate, to me.
@@wilcalint when I’ve done so, it was with ProTools and I simply selected 24bit for the setup. Then set the sample rate of course. To check? ProTools will say what bit depth and sample rate is if you ‘inspect’ the file.