Georgia Tech Professor of Electrical and Computer Engineering in the Digital Signal Processing Technical Interest Group here... This is the first explanation of sampling, aliasing, upsampling/downsampling, antialiasing filters, etc. I've seen/heard, in print or in video, in 20 years that (a) didn't make me cringe at any point, and (b) completely blew me away with how illuminating it was.
I can say exactly the same! I have designed and built a DSP guitar amplifier. I have seen a lot of videos and articles that are incorrect and / or misleading. This one is superb!
@@antoniomonteiro1203 How did you find the process of building that, Antonio? Were there any resources/instructions/information you found particularly helpful? Something I've been thinking about trying my hand at. I'm pretty handy with various projects, but an Amp would be new territory for me and I'm not quite sure where to start.
I've been dabbling into DSP myself, and I'd like some more info about this. I understand the basics of harmonic distortion and I understand what aliasing is. But does this unwanted effect, of reflected frequencies, happen due to processing in the time domain or in the frequency domain? I haven't tried to make a saturation plugin myself, but I'd imagine I would simply limit each sample in the time domain, with a transfer function that compresses the curve towards the extremes, so that a sinusoid would start to approach a square wave. Surely this would add third order harmonic distortion, but wouldn't cause any aliasing, would it? Sure, if I kept on adding odd harmonic sinusoids, and went above nq frequency, then there would be aliasing, but why would I do that? As you sure can tell, I haven't actually done any non-linear processing (coding) myself... :)
You are not right, some things actually are related to production and music producers to use tools right, among them, your CPU performance limits you to do things right......
I've been trying to grasp this sample rate stuff for years. And along comes Dan. I'm not sure anyone could possibly do a better job of teaching this. Color me awestruck. I finally get it! And if I can get it, anyone can!
Then you mustn't have tried. There are tons of resources on the basics of digital signal processing. There have even been great introductory videos for people that cannot read, like xiph's "Digital Show & Tell" published in 2013.
Same here man! He also presents in a manner which I want to listen to and not fall asleep to. Even though his voice would be GREAT for audio books haha.
I know I'm going to get my ass flamed & I would not trust this resource for anything historical nor political, but what's wrong with this definition ? en.wikipedia.org/wiki/Sampling_(signal_processing)#Sampling_rate
This is, in my mind, the clearest and most informative demonstration and explanation of sample rates, aliasing, and the tricky bits of digital audio I've ever encountered. I'm also digging the Reaper mixer setup.
Dude, forget aliasing and oversampling.....this is a masterclass on the “inner lives” of our DAWs. This is yet another reason that I will always support fabfilter. Their plugins are among the cleanest in the industry, they’re infinitely usable and they’re pleasant on the eyes over the course of a long day. But even after all that, there’s content like this. I fear the day you guys decide to release a DAW. At that point, I’d have a pretty serious decision to make regarding my 20+ year investment in Pro Tools’ ecosystem....
@@Drew.DrivesYT This is *just* a question…seriously, nothing antagonistic AT ALL. Don’t read into this as anything other than curiosity. Do you do this for a living?
it should be at least a blind test, when you see changin text 44 oversampled/48 bla bla, it tricks your brain and you "hear" difference. but it the reality there's no difference maybe (now you need to do a blind test yourself, or remain a fool). then go watch pensado's place where he, with a friendly smile on his face, swears that he can hear a difference in audio cables. also every single itl where he lies to people then every single mixing tutorials channels where there is not a second of useful info and then think
@Gonzalo: Definitely among the best "processor" VST's I've ever used, but then I've not really tried many (any?) others since discovering FF! I guess thats a good AND bad thing lol!
The most important feature for high sample rates which you did not talk about, is the ability to pitch down audio with much better quality. Other than that, awesome video!
I can see this being relevant for slow motion video. I'm curious, what are some other areas where you would significantly downpitch as to where this would have a relevant impact?
@@thorwaldjohanson2526 basically in any modern electronic music production/sound design environment, for instance drum'n'bass and jungle producers have been severely downpitching/pitching/timestretching all kinds of sounds since the 90's, also the whole "chopped and screwed" hip hop sub-genre is BASED around slowing songs down for example. Furthermore, changing the speed of a sound is basically one of the easiest ways of transforming it and making it unrecognizable, hence giving you creative options.
@@thorwaldjohanson2526 Auto-tune. The way it works is that it cuts audio into the tiniest chunks and sort of copy-pastes them and stretches them to alter pitch, length and vibrato. With 44.1 kHz, it sounds weird very quickly even with very mild auto-tuning. Even just having the recording in 48 kHz makes a big difference in my experience, now you can get away with correcting pitch and timing without it being noticeable. 96 kHz allows for extreme stuff if you're into that, but if you got somewhat capable musicians and are only fixing a few off notes or beats, 48 kHz works well for me (in Logic's own auo-tune system).
@@Hamachingo Thanks for the responses guys :). I'm curious, are you usually recording in 192khz 24bit and then downsample if need be, or do you start directly with 48khz? Or do you record different instruments at different settings? And are there any disadvantages besides filesize, to record at the maximum bitrate?
I've been a strong proponent of working at 44.1 project sample rates for over a decade. Based largely on the understanding that most music producers have sample libraries comprising 44.1 samples, and that crappy realtime SRC up to 48 inside most DAWs was incurring a needless quality loss from the minimum phase anti-aliasing filter during realtime SRC. But your explanation of the cumulative effect of many oversampling stages with a gentler anti-aliasing filter at 48 has opened my eyes. Your A-B comparison at the end, of 44.1 oversampled versus 48 oversampled across the entire project was mind-boggling. My hearing is relatively destroyed at this point in my career, but even with my trashed ears, I can clearly hear the difference in that final AB demonstration. I would never have had the patience to set up a test project of that size to show the cumulative difference at 48 vs 44.1 like that. THANK YOU. (I'll be working at 48000 project rates from now on, despite the fact that I don't do soundtrack work and my sample library comprises mostly 44.1 samples.)
@Agent К_видео The difference is subtle, and falls into the category of "ear training" for certain things, plus a LOT depends on your monitors and room characteristics (or your headphones). I'm used to listening for "how much clipping is too much clipping" because I use clippers a lot and push limiters hard. The aliasing artifacts and IMD in the 44100 version has some similar "dirt" and "grit" around the edges of the transient sounds in the mix (like with clipping), and the mids and highs feel a little more "congested". By contrast, the 48000 version has a more "open" feeling in the mids and highs, and not as much of that dirty/gritty edge to the transients.
@@Baphometrix Agree with your assessment and my opinion was that neither were necessarily "better" I listened to the comparison with different sets of monitors, and for some it felt like the 44.1 version with OS was slightly more mid/low-mid forward and it made it feel like a more focused mix. No doubt that the 48 one felt a bit wider/open, but that's not always what I want/need in a mix.
I think you missed the point of that final comparison, which was that the A/B differences were negligible. I think the changeable nature of the music made it harder to notice the lack of difference between the formats.
Thank you, thank you, thank you, thank you. I've been telling people (mostly amateur engineers and audiophiles) this stuff for years but this is the first video I've seen that actually illustrates it perfectly with smart examples and incontrovertible proof. High sample rates for the final, consumer delivery format are useless. Even for basic tracking, they don't really matter because the majority of microphones don't pick up much of anything over 18k anyway. The only reason to use them is in mixing and it is far more practical to just use over-sampling rather than record everything at a higher sample rate. Dan Worrell is the best.
Yeah I read an interview with one of the engineers that was on the CD team. Phillips Sony, there was a Toshiba guy there as well according the interviewee. It was fascinating reading.
Ending a very technical and precise video essay with the punchline "If it sounds good, it is good" is just killer. :D Thank you so much Dan for explaining these things. I’m one of those audio guys who know next to nothing about the technical side but feel like I really should know more. Compared to the photography and videography world, us music producers seem to worry little about technical aspects that do play a role (and we tend to know about them even less, especially compared to our camera wielding counterparts). I’ve wondered about why that is. I love how well you explain the technical aspects so that even a layman like me can understand most of it, and you approach it with a healthy pragmatism that both the "I only care about the creative side and therefore tech doesn’t matter" as well as the borderline-esotericists miss out on. I’ll probably have to revisit the video a few times before I grasp it entirely, but for now: I’ve already learned a lot with a single watch! Cheers!
@@rickbiessman6084 yup I remember it clearly i worked music retail back in them days. In fact the irony of both slogans is palpable. by "we hear you" are they referring to how they like to slap people with lawsuits when they are found out how Behringer infringe the copyright / patents of other manufactures I wonder? because they are certainly listening out for that hahah
probably the most honest video a company has ever made. Not even speaking of the clear and good information told by somebody with a great voice to listen to. Thank you for doing this!
For the parts of the lecture/video that are easier to "grasp" even for laymen interested in this topic, I bow deeply and take my hat off to you!!! Thank you, Sir Worrall, for this extraordinary informative content 👏
Great video and explanation. Matches up with what I've been doing for years. I record and edit/mix almost everything at 48kHz. With popular music, the space and processing savings of 48k are noticeable and any harmonic content above 20k is pretty much buried under other instruments anyway. Plus, producing high sample rate content is pointless anyway since pretty much all forms of music compression (size) will bandpass limit the content anyway before applying the lossy compression. However, there is still one thing that I record and edit at higher sample rates - that being orchestral music. With typically few microphones (allowing acoustic mixing of the band/orchestra prior to microphone capture), the space and processing savings is unnecessary. There are some instruments that do have harmonics, and even natural energy, above 20k. I find it best to capture and edit at the higher sample rates and then apply bandwidth filters upon final mixdown. This avoids the issue of those frequencies above 20k turning into aliased content in the below 20k range during initial tracking/capture or in the editing stage.
geez, this is the greatest example of edutainment (in fact, mere education) that I've come across in years. so clearly put and illustrated that well it should serve as an obligatory role model for anyone teaching or just having to explain any complex matter.
I am seriously getting goosebumps! I have been trying to figure out why cutting off frequencies above certain values and saturating would interplay in weird ways. I now can at least PARTIALLY understand some of it.
Absolutely unreal as to how much I just learned in this half hour... Couldn't be happier to have been school by the great Dan once again! Thank you so much!!!!
Dan, your presentations convey not only technical knowledge but also, and more importantly, much wisdom about how the skills fit into the whole project and to this work in general. Much thanks for an EXTREMELY well-planned and produced presentation -- the density of useful knowledge per unit time is off-the-charts! If there were awards given for this sort of thing, you would sweep the whole lot, every time!
Aligns with my own experiences and lessons over the years: bandpass everything, all audio is bandpassed at some stage anyway so you better do it.. before it is too late and the signal encounters something that can't cope with its demands. Distortion and heat will be the result of that and you will get degraded audio signal. Also: Hi-res record releases are not meant for consumer consumption, you once again have to deal with bandpassing the signal before it hits your speaker. The end result is intermodulation distortion in the analog chain too. Most of it will be turned to heat but it is never a good idea to push too wide of a sausage thru a too small hole... A good speaker has ultra- and infrasonic filtering but one of the principles i have picked out is that it is far better to cut the signal before amplification than after. If for nothing else, components are smaller and the circuit design can be considerably better.
About these high res 'record' releases (I'm assuming digitized recordings of analog master tape playback then of course, just stating this because in common parlance the term record is often read as synonymous with vinyl plate prints); To deny that they are *meant* for consumer consumption (as opposed to what consumption?) is simply contrarian. You know about these releases, so you must know that their resellers have a consumer demographic. I consume, as a consumer - which includes every producer because none of us only uses music materials for work aka in your terms "producer production": just to source, process, stock etc. - I consume hi-res consumer content recreationally, like the vast majority of their target demographics. And I'm not sure if you're simply generalizing a part of this market, or if you're making a case against the release of hi-res digitizations of any music source material, for any recreational consumers. Certainly when taking what you're saying literally, that is what you are doing; but of course communication is too complex to know for sure. In any case, consider the consumption of hi-res releases, by a consumer who bit-streams it to a high quality digital-to-analog converter, going into an analog (pre)amplifier which easily handles the output signal of the aforementioned converter, ending up, in any circumstance which assumes no use of destructive ignorent elements in the signal chain between, playing back through high quality speakers or headphones. What about this is 'not meant' to be?
After reading and arguing about this stuff on the forums for so long, these videos (especially your latest ones) feel like a big therapeutic release, because all the stuff I've learned by doing a bunch of tests in my home studio are finally put into a form that is tangible and easily shareable.
Honestly, your videos are near perfection. Straight to the essential point, very pleasant to hear. And most of all, supported with scientifical demonstration that make us have confidence in the knowledge you transmit. I believe that your videos made me subconsciously create a growing percieved value for the fabfilter bundle, which I bought recently
Dan, for years, I have searched for this conclusion. I spent two weeks chasing my tail trying to mix a solo acoustic guitar song: each corrective tweak created more deficiencies than what it solved. No matter what I did to "improve" the mix, it simply sounded like there were both dead spots and artificial bright spots that nothing could fix. After watching this video, I tried a 'what-if'. I lowered the DAW session sample rate from 96kHz to 48kHz and made no other adjustments. Immediately, as if someone flicked on the sonic fullness switch, my mix sprang to life. Life will never be the same, thanks to this video. 48 kHz is my new sample rate. If a project calls for a track to make its way into the movie domain, I will simply upsample the 48 kHz master to 96 kHz to meet that format. Thank you for this most excellent 'put the sample-rate argument to bed' video. Keep up the good work!
I'm sorry to say this with a delay of one full year. But are you aware that if you resample anything from 48 kHz to 96 kHz, the only thing you get is your files taking up twice the space they originally took, without any improvement in sound quality whatsoever?
Best video and information on sample rates and oversampling to date. I’ve been so confused for so long, so I feel like I’ve watched and read them all. I feel much more comfortable now. Thank you so much.
Converter design legend, Dan Lavry, wrote a white paper over a decade ago integrating his explanation of this issue with a somewhat math-heavy explanation of the Nyquist-Shannon Sampling Theorem. But I suspect this excellent vid from DW will reach more people and leave less head-scratching among those of us who didn't absorb much trigonometry in school.
So much useful stuff in just one video, it's unbelievable : - Understanding why digital saturation/compression doesn't sounds right compared to analog stuff (i.e. guitar amps for example) - What is aliasing, with clear examples - Which effects are concerned and you need to oversample, with audible comparisons - Advice on what to do for the best quality without destroying your CPU - The incredible custom guitar chain with EQ for the cab, then 3 stages of saturation for the amp. I HAVE to use this ! - At the end of the day, technical stuff or not, if it sounds good it IS good. Maybe the best advice of this video. Thank you so much Dan Worrall, for this free valuable resource.
I'll have to watch again. Excellent content as always. It's not only a big learning factor, your video conclusions nearly always relaxes me, helping not to worry too much about possible quality losses. That's so great and I thank you very much.
I didn't think I'll grasp this concept, but I did. The explanation is stellar. Nothing was overly complicated with word salads, the examples were presented and explained in a way that's easy to understand. Great stuff right there!
I love these rabbit holes what take you right back where you started, but with more knowledge of why your are good where you're at. I could hear the difference between all the sample rates/oversampling combos. I think I've concluded that for my uses, mixing in 48kHz with oversampling where possible will do juuuuust fine. I'm actually considering getting Saturn now, but I'm not sure I need it with the other saturation plugins I have. I haven't even looked to see if they have oversampling. Mixing is one of those things where you can easily get lost in the details.
Dan, Your videos give me a deeper understanding of complex topics and sooth me into a state of zen relaxation all at once. Your voice is amazing and easy to listen to and your approach to teaching these concepts is incredible. I have learned so much from your channel and I am looking forward to what you teach us next. Thank you so much for all your hard work it is much appreciated.
Thank you for quality explanations! Hope the 6 files (3 sample rates and OS on/off) could be provided in 44.1kHz 16bit wave format for off-line comparison.
My goodness, what a tutorial! Dan completely grabbed my attention for 30 mins when I thought I am not in a position to watch the entire video. Thanks a lot, FabFilter for the awesome plugins and Dan for the awesome tutorials.
Wow, what a teacher! I fell asleep halfway through this video, woke up still in awe, and continued watching until the end. I'm going to make a playlist full of this man's videos.
The point of higher sample rates in my industry is so that you can pitch audio down without losing all high frequencies. Its VERY important for sound design
When Dan speaks … I'm glad he finally addressed this issue. Chris "Monty" Montgomery at Xiph.org did an excellent couple of articles and videos on this years back, but they did not go nearly into the kind of practical depth that Dan does, and they somehow went over a lot of people's heads. This is why so few companies can touch FabFilter. They do the kind of thing you'd expect from companies like UAD, but really get from Airwindows and FabFilter.
Again you guys have done yet another video that's lightyears ahead of the rest, explaining a complex subject in clear way that doesn't dumb it down too much.
Alguien por aquí ha dicho que deberíamos de pagar por vídeos así. Estoy de acuerdo! De lo mejor que he visto sobre esta eterna discusión acerca del sample rate en años. Cuestión zanjada en mi opinión. Thank you so much Dan for such a valuable information!! 👏👏👏👍👍👍
Fabfilter, yall literally make the best, most informational, deep yet approachable videos of anyone in the game! this stuff is the deep magic in the days of digital audio
Thanks for a fantastic video Dan Worrall!! I have been skeptical if the the higher samplingrates made any bigger difference. When using 96 and resampling down to 44.1, I never heard a better result than going from 48 to 44,1. Now I can stop wondering and just stick to 48k with a great explanation and relief :-)
This video demonstrated that you will hear a difference, assuming you apply digital processing of the signal to a common degree. What did you think the point was of him advocating the use of functions which use 8x the project sample rate, featured in the distortion plugins?? It just also demonstrated that you're better off using higher sample rates on a per-process/plugin basis, to avoid using resources on processes which cannot benefit from it (meaning not setting your entire project sample rate, based on the highest denominator which is your most aliasing-prone plugin(s) because it affects the CPU power needed for everything which doesn't audibly affect the signal as well. This has always been the case in out-of-the-box processing, because every analog signal processor, such as the type of distortions he's emulating here, was already 'upsampled' separately, by way of being an analog process, while the drum machine that the band used, was not unnecessarily digitally upsampled at the source, or at any digital point in the chain that didn't benefit from upsampling. But you will certainly hear the difference when you apply a digital process that aims to recreate the type of analog warmth that these distortion DSPs are doing. He's literally explained to you that you want very high sample rates for that type of processing, just that it's handy to have this function available and contained within each plugin, which in effect is just a way of getting closer to the way out-of-the-box mixing and mastering is done, assuming the inclusion of analog gear, which is virtually any music that isn't chiptunes.
@@JohnnyNatrium Thanks for your reply Johnny. I got it and that was exactly my point. Using plugins upsampling.. and not considering the AD/DA when using outboard or recording audio:-)
bandwidth limiting when using multiple stages of distortion to prevent aliasing.... that's what i've been doing! also it is incredible how much of a different 41.1 to 48 makes this is an amazing demonstration
Amazing video! Could you also do the same video for resolution as that has the same myths. So many DAWs support 32bit floating now but so many still use 24bit. Would be great to know the math behind this as well.
My leaked experimental version of Ableton 13 supports 64bit flying point audio which it isn't even possible to record at yet, except on the prototype interface I stole from MI5's top secret recording studio (2 miles below London underground) when the guards were distracted by a girly mag I threw on the floor, so I'll see y'all later!
CPUs might actually handle 32-bit float better than 24-bit int depending on the algorithm. 24-bit actually has a ton of information, you probably won't hear it the difference, but I would not say the word "always" when it comes to computers There's a good chance some DAWs will store 24-bit samples as 32-bit integers since CPUs handling 32-bit data better than 24-bit, plus it would increase mixing quality
The point is made at the end that higher sample rates can result in a higher chance of increased intermodulation (lower quality content), is this touched on at any other point in the video? Also he said that "higher samples rates do nothing to improve the quality of audible content below 20khz" isn't this contradicting what was shown earlier in the video in Acustica? Really great video by the way, I loved it.
He touched on this when using multiple instances of harmonics-generating plugins (saturation, compression) in a row: even at 96kHz the first plugin adds harmonics that will 'bounce back' off the nyquist freq. The second plugin then generates harmonics off of those 'false' harmonics and so on. The reason the chance of this happening more often in higher samplerates is human error: because you tend to not worry about aliasing and thus don't turn on the oversampling algo because you're already working in a high samplerate (so you think..) Or at least, that's how I understood it...
@@gijscoolen6439 Ah ok, yeah it's the reasoning behind it being said that higher sample rates can be lower quality that puzzles me. Seemed to imply more than just forgetting to use oversampling. The statement "higher samples rates do nothing to improve the quality of audible content below 20khz" as mentioned in my first comment still puzzles me too.
I think the point is that increasing samplerate doesn't automagically mean better audio fidelity and cleaner signals. A lot of things have to align in the stars (as it were) for this to 'just happen' and that no matter what, with digital audio, you're going to have work around how digital audio naturally exists. Also, the point he made about increasing the sample rate was also that you'll use far more CPU with marginal results in many cases. It's like the differences between 4k computer monitor and 8k computer monitor. Your computer has to work 4 times as hard to give you a picture that literally no one can tell the difference unless it's blown up to the size of a movie theater screen and even then you'll only notice the difference once you get 5 feet from it, and no one likes watching a movie theater screen from 5 feet away.
This was a great video and you broke it down in layman's terms where people could grasp the concept. In reality the better the original recordings are captured or managed can also play a very important role in eliminating some of these issues you are pointing out especially if it means being able to use less plugins. In my home studio I try to use plugins sparingly not only to limit how much processing power is required by the computer but just to keep things simpler. I also run a Hybrid DAW station with an analog mixer with the option to run external compressors and limiters when needed which allows me to capture a really decent analog drum mix where the DAW can be just more of a multi-track recorder and the same goes for capturing bass guitar and using an external compressor/limiter rather than using plugins to manage the bass guitar's dynamics and as long as you're mindful of the gains during recording, you're golden. For me and my needs this works but for a modern day commercial recording studio this probably wouldn't fly or at the least they would be limited especially dealing with many different styles of music due to the dynamics they may see like speed metal vs classical music along with music production for a movie etc and the effects required to make it deliver.
Not for all videos. This video uses the Opus codec with a 20 kHz filter. The 20 kHz tone at 4:59 is actually clearly there (I checked with a spectrum analyser), allbeit a bit attenuated.
not only Fabfilter makes some of the best plugins out there but they also do incredible and honest technical videos with the legendary Dan ! Thank you !!
Thank you for this demo, i've record in many studios and it concur with the knowledge i have acquired, and i do my own mix in 48khz, however after hearing your mixes at 44.1, 44.1 with oversamplig and 96khz, i hear a difference in the very highs specially on the cymbals and hit hat in 96khz like a bit more of "air" than the other exemple
i was arguing with a buddy because he swears hi res is the only way to go about it, if i only found this video sooner i could of saved my self a headache, really informative. and cleared some things up for me too. i have used fabfilter as my go too compressor and EQ for a few years but never use the over sample features at all. didnt actually hear the difference but now i SEE why its a feature
This video is a staple in the hi-res audio debate of whether it's actually worth anything, and you do a great job at explaining how it's mostly just a bunch of people who don't know the details who obsess over hi-res. For example, I was one of those people, and then I watched this video lol.
Before I invested in my mixer and other outboard hardware, I did a lot of research into what was considered the optimum rate against cost benefit and return. I wanted to record bands multitrack during live performances and master them afterwards into usable acceptable marketing material. I had used my much loved and trusted Yamaha 01x for almost 20 years but it has been obsolete for over a decade since Yamaha refused to update the "firewire" drivers since Windows 98 so there after I have to run later versions of Windows in development/test mode for the mixer to work !! For whatever reason I never found this tutorial, not that it would have made any difference, it is by far the best explanation over any other I watched of sample rates and diminishing returns on investment. I can see some highly regarded producers seriously argue with your opinion/conclusion, but like you I just chose to ignore their snobbish professional opinion of higher is better, understanding it didn't matter. So I bought my Behringer x32 over the more expensive Midas 32 variant or Allen and Heath knowing the quality might not quite be the same but still good enough for what I wanted to achieve. I also own a Liquid 56 so any extra quality could be overdubbed later, that is housed in my home studio cabinet that is way to heavy to be considered mobile. I just connect my Surface Pro through USB and I have a full 32 channel mobile recording mixer/studio with full remote control. I bought this start of lockdown as I was offered it at a great price including the stage boxes and interconnecting cables. Now I know some might comment on the Midas pre amps, I know, really I do, they are better, but I got everything for less than 1/3 the cost of the Midas 32 new.
Georgia Tech Professor of Electrical and Computer Engineering in the Digital Signal Processing Technical Interest Group here... This is the first explanation of sampling, aliasing, upsampling/downsampling, antialiasing filters, etc. I've seen/heard, in print or in video, in 20 years that (a) didn't make me cringe at any point, and (b) completely blew me away with how illuminating it was.
Excellent.
I can say exactly the same!
I have designed and built a DSP guitar amplifier. I have seen a lot of videos and articles that are incorrect and / or misleading.
This one is superb!
@@antoniomonteiro1203 How did you find the process of building that, Antonio? Were there any resources/instructions/information you found particularly helpful? Something I've been thinking about trying my hand at. I'm pretty handy with various projects, but an Amp would be new territory for me and I'm not quite sure where to start.
@@jackallenproductions There is a lot of information and I believe it is not practical to put it here. How can we communicate by other means?
I've been dabbling into DSP myself, and I'd like some more info about this. I understand the basics of harmonic distortion and I understand what aliasing is. But does this unwanted effect, of reflected frequencies, happen due to processing in the time domain or in the frequency domain? I haven't tried to make a saturation plugin myself, but I'd imagine I would simply limit each sample in the time domain, with a transfer function that compresses the curve towards the extremes, so that a sinusoid would start to approach a square wave. Surely this would add third order harmonic distortion, but wouldn't cause any aliasing, would it? Sure, if I kept on adding odd harmonic sinusoids, and went above nq frequency, then there would be aliasing, but why would I do that? As you sure can tell, I haven't actually done any non-linear processing (coding) myself... :)
Finally, someone explains the whole sample rate thing in a way that non-DSP engineers can actually understand. Thank you Dan
You are not right, some things actually are related to production and music producers to use tools right, among them, your CPU performance limits you to do things right......
this!
Meanwhile a student who is studying DSP algorithms discovers the world of music production (me)
Dan, you're amazing. I would replace Alexa and Siri's voice with yours in a heartbeat.
😃 good idea
Lux Elliott or my wife’s.....
this should happen. I want Dan's voice as an option for Siri.
Dude me2 - and imagine they would give you so much well articulated and informed information
@@LoveMeBack that cracked me up thanks
"If the song is good and the mix is good, no one will care about a bit of aliasing!" Phew ... sanity is finally restored.
Yes the differences were so subtle I don't know if I could pass a blind AB test.
Dude I don't care if the mix sounds good or nah. I just want to be able to brag about having 14 compressor on my vocal track.
@@michaelvenne9386that’s just dumb
@@Gang-25j yeah.
its not just aliasing.. there is sound difference..
I've been trying to grasp this sample rate stuff for years. And along comes Dan. I'm not sure anyone could possibly do a better job of teaching this. Color me awestruck. I finally get it! And if I can get it, anyone can!
Then you mustn't have tried. There are tons of resources on the basics of digital signal processing. There have even been great introductory videos for people that cannot read, like xiph's "Digital Show & Tell" published in 2013.
@@xnoreq no need to rain on his positivity dude....he's just a happy camper!
@@xnoreq oooh aren't you the clever girl
Same here man! He also presents in a manner which I want to listen to and not fall asleep to. Even though his voice would be GREAT for audio books haha.
I know I'm going to get my ass flamed & I would not trust this resource for anything historical nor political, but what's wrong with this definition ? en.wikipedia.org/wiki/Sampling_(signal_processing)#Sampling_rate
This is, in my mind, the clearest and most informative demonstration and explanation of sample rates, aliasing, and the tricky bits of digital audio I've ever encountered. I'm also digging the Reaper mixer setup.
Best digital audio aliasing explanation ever. Thank you!
EVER!
..and Nyquist frequency.Thank you!
PLEASE Someone has to make a 10 hours version of Dan Worrall saying "However" and "kilo Hertz".
I'll need it to fall asleep at night
And that person is you.
Lol it's funny I'm watching this before turning the light out and going to sleep.
Haha. Gold
I fell asleep watching this vid.
th-cam.com/video/yaiS3f3P-2M/w-d-xo.html ok, its not 10 hours, but hey, Merry Christmas
Dude, forget aliasing and oversampling.....this is a masterclass on the “inner lives” of our DAWs. This is yet another reason that I will always support fabfilter. Their plugins are among the cleanest in the industry, they’re infinitely usable and they’re pleasant on the eyes over the course of a long day. But even after all that, there’s content like this. I fear the day you guys decide to release a DAW. At that point, I’d have a pretty serious decision to make regarding my 20+ year investment in Pro Tools’ ecosystem....
I would love to see Fabfilter make a DAW! :-)
OMG - can you imagine a FF DAW?! Who says Utopia can't exist?
I think Avid/PACE is a strong enough motivator to move on. There's plenty of better DAWs out there already.
@@Drew.DrivesYT
This is *just* a question…seriously, nothing antagonistic AT ALL. Don’t read into this as anything other than curiosity. Do you do this for a living?
@@manifestgtranswer: you definitely don’t need protools too make a living in music/mixing..stop the bullshit
I feel like I should have paid for this. Anyone else?
it should be at least a blind test, when you see changin text 44 oversampled/48 bla bla, it tricks your brain and you "hear" difference. but it the reality there's no difference maybe (now you need to do a blind test yourself, or remain a fool).
then
go watch pensado's place where he, with a friendly smile on his face, swears that he can hear a difference in audio cables.
also every single itl where he lies to people
then every single mixing tutorials channels
where there is not a second of useful info
and then think
@Gonzalo: Definitely among the best "processor" VST's I've ever used, but then I've not really tried many (any?) others since discovering FF! I guess thats a good AND bad thing lol!
I was going to make my own comment but this comment sure seems like what I was thinking so there you go!🤗😇🤪
@@gonzalob3348 Yep, bought most of them a while back. They are really all one needs (and yes, I have UAD, Waves etc)
@@C0D3O Genuinely curious. What kind of lies does he tell?
The most important feature for high sample rates which you did not talk about, is the ability to pitch down audio with much better quality. Other than that, awesome video!
Igelkotte Good point, thanks.
I can see this being relevant for slow motion video. I'm curious, what are some other areas where you would significantly downpitch as to where this would have a relevant impact?
@@thorwaldjohanson2526 basically in any modern electronic music production/sound design environment, for instance drum'n'bass and jungle producers have been severely downpitching/pitching/timestretching all kinds of sounds since the 90's, also the whole "chopped and screwed" hip hop sub-genre is BASED around slowing songs down for example. Furthermore, changing the speed of a sound is basically one of the easiest ways of transforming it and making it unrecognizable, hence giving you creative options.
@@thorwaldjohanson2526 Auto-tune. The way it works is that it cuts audio into the tiniest chunks and sort of copy-pastes them and stretches them to alter pitch, length and vibrato. With 44.1 kHz, it sounds weird very quickly even with very mild auto-tuning. Even just having the recording in 48 kHz makes a big difference in my experience, now you can get away with correcting pitch and timing without it being noticeable. 96 kHz allows for extreme stuff if you're into that, but if you got somewhat capable musicians and are only fixing a few off notes or beats, 48 kHz works well for me (in Logic's own auo-tune system).
@@Hamachingo Thanks for the responses guys :). I'm curious, are you usually recording in 192khz 24bit and then downsample if need be, or do you start directly with 48khz? Or do you record different instruments at different settings? And are there any disadvantages besides filesize, to record at the maximum bitrate?
I've been a strong proponent of working at 44.1 project sample rates for over a decade. Based largely on the understanding that most music producers have sample libraries comprising 44.1 samples, and that crappy realtime SRC up to 48 inside most DAWs was incurring a needless quality loss from the minimum phase anti-aliasing filter during realtime SRC. But your explanation of the cumulative effect of many oversampling stages with a gentler anti-aliasing filter at 48 has opened my eyes. Your A-B comparison at the end, of 44.1 oversampled versus 48 oversampled across the entire project was mind-boggling. My hearing is relatively destroyed at this point in my career, but even with my trashed ears, I can clearly hear the difference in that final AB demonstration. I would never have had the patience to set up a test project of that size to show the cumulative difference at 48 vs 44.1 like that. THANK YOU. (I'll be working at 48000 project rates from now on, despite the fact that I don't do soundtrack work and my sample library comprises mostly 44.1 samples.)
@Agent К_видео The difference is subtle, and falls into the category of "ear training" for certain things, plus a LOT depends on your monitors and room characteristics (or your headphones). I'm used to listening for "how much clipping is too much clipping" because I use clippers a lot and push limiters hard. The aliasing artifacts and IMD in the 44100 version has some similar "dirt" and "grit" around the edges of the transient sounds in the mix (like with clipping), and the mids and highs feel a little more "congested". By contrast, the 48000 version has a more "open" feeling in the mids and highs, and not as much of that dirty/gritty edge to the transients.
@@Baphometrix Agree with your assessment and my opinion was that neither were necessarily "better" I listened to the comparison with different sets of monitors, and for some it felt like the 44.1 version with OS was slightly more mid/low-mid forward and it made it feel like a more focused mix. No doubt that the 48 one felt a bit wider/open, but that's not always what I want/need in a mix.
I think you missed the point of that final comparison, which was that the A/B differences were negligible. I think the changeable nature of the music made it harder to notice the lack of difference between the formats.
Dan definitely knows he's the best. Sometimes I get on TH-cam simply to hear this man talk about things I can't comprehend.
This by far the smartest marketing video for FabFilter. And what makes it so unique is the fact, that everything he's explaining is 100% true!😂😂😂
Thank you, thank you, thank you, thank you. I've been telling people (mostly amateur engineers and audiophiles) this stuff for years but this is the first video I've seen that actually illustrates it perfectly with smart examples and incontrovertible proof. High sample rates for the final, consumer delivery format are useless. Even for basic tracking, they don't really matter because the majority of microphones don't pick up much of anything over 18k anyway. The only reason to use them is in mixing and it is far more practical to just use over-sampling rather than record everything at a higher sample rate.
Dan Worrell is the best.
Yeah I read an interview with one of the engineers that was on the CD team.
Phillips Sony, there was a Toshiba guy there as well according the interviewee.
It was fascinating reading.
Not all plugins have oversampling built in. They generate (fold back) aliasing in the audible frequency area when working on lower sample rates.
This video should be mandatory for anyone learning audio engineering.
Ending a very technical and precise video essay with the punchline "If it sounds good, it is good" is just killer. :D
Thank you so much Dan for explaining these things. I’m one of those audio guys who know next to nothing about the technical side but feel like I really should know more. Compared to the photography and videography world, us music producers seem to worry little about technical aspects that do play a role (and we tend to know about them even less, especially compared to our camera wielding counterparts). I’ve wondered about why that is.
I love how well you explain the technical aspects so that even a layman like me can understand most of it, and you approach it with a healthy pragmatism that both the "I only care about the creative side and therefore tech doesn’t matter" as well as the borderline-esotericists miss out on.
I’ll probably have to revisit the video a few times before I grasp it entirely, but for now: I’ve already learned a lot with a single watch! Cheers!
my only issue with that saying, is that it used to be Behringer's slogan, and I mean early Behringer when they were REALLY bad haha.
@@Bthelick Was it? I had no idea :D I only knew their slogan "We Hear You".
@@rickbiessman6084 yup I remember it clearly i worked music retail back in them days.
In fact the irony of both slogans is palpable.
by "we hear you" are they referring to how they like to slap people with lawsuits when they are found out how Behringer infringe the copyright / patents of other manufactures I wonder? because they are certainly listening out for that hahah
IISGIIG … “The Golden Rule” (read like Andy Samberg from the end of 3-Way … … …)
probably the most honest video a company has ever made. Not even speaking of the clear and good information told by somebody with a great voice to listen to. Thank you for doing this!
Dan, probably the best video I've seen on tis subject.....ever.... Thank you . I'll refer friend and colleagues to this from now on
For the parts of the lecture/video that are easier to "grasp" even for laymen interested in this topic, I bow deeply and take my hat off to you!!! Thank you, Sir Worrall, for this extraordinary informative content 👏
Great video and explanation. Matches up with what I've been doing for years.
I record and edit/mix almost everything at 48kHz. With popular music, the space and processing savings of 48k are noticeable and any harmonic content above 20k is pretty much buried under other instruments anyway. Plus, producing high sample rate content is pointless anyway since pretty much all forms of music compression (size) will bandpass limit the content anyway before applying the lossy compression.
However, there is still one thing that I record and edit at higher sample rates - that being orchestral music. With typically few microphones (allowing acoustic mixing of the band/orchestra prior to microphone capture), the space and processing savings is unnecessary. There are some instruments that do have harmonics, and even natural energy, above 20k. I find it best to capture and edit at the higher sample rates and then apply bandwidth filters upon final mixdown. This avoids the issue of those frequencies above 20k turning into aliased content in the below 20k range during initial tracking/capture or in the editing stage.
What do you think about recording the sounds of a keyboard through its headphone jack? Thanks.
geez, this is the greatest example of edutainment (in fact, mere education) that I've come across in years.
so clearly put and illustrated that well it should serve as an obligatory role model for anyone teaching or just having to explain any complex matter.
I am seriously getting goosebumps! I have been trying to figure out why cutting off frequencies above certain values and saturating would interplay in weird ways. I now can at least PARTIALLY understand some of it.
Absolutely unreal as to how much I just learned in this half hour... Couldn't be happier to have been school by the great Dan once again! Thank you so much!!!!
The best 29 minutes dedicated to sample rates on the internet. Excellent, thank you!
That's "PROFESSOR Dan" from now on. So clear. So complete. So concise. So the best! Thank you!
That was utterly engrossing, educational and absorbing. I got so much from this, thanks Dan.
I’m a former Audiovisual Systems Engineer from Barcelona and this has never been explained so clearly to me. Thank you. Best tutorials ever.
Dan, your presentations convey not only technical knowledge but also, and more importantly, much wisdom about how the skills fit into the whole project and to this work in general. Much thanks for an EXTREMELY well-planned and produced presentation -- the density of useful knowledge per unit time is off-the-charts! If there were awards given for this sort of thing, you would sweep the whole lot, every time!
This is actual KNOWLEDGE, AND IT'S FREE!!!!!
Thank you for making the world better!
Aligns with my own experiences and lessons over the years: bandpass everything, all audio is bandpassed at some stage anyway so you better do it.. before it is too late and the signal encounters something that can't cope with its demands. Distortion and heat will be the result of that and you will get degraded audio signal.
Also: Hi-res record releases are not meant for consumer consumption, you once again have to deal with bandpassing the signal before it hits your speaker. The end result is intermodulation distortion in the analog chain too. Most of it will be turned to heat but it is never a good idea to push too wide of a sausage thru a too small hole... A good speaker has ultra- and infrasonic filtering but one of the principles i have picked out is that it is far better to cut the signal before amplification than after. If for nothing else, components are smaller and the circuit design can be considerably better.
I hope one day to be as wise as you 🙏🏽
About these high res 'record' releases (I'm assuming digitized recordings of analog master tape playback then of course, just stating this because in common parlance the term record is often read as synonymous with vinyl plate prints); To deny that they are *meant* for consumer consumption (as opposed to what consumption?) is simply contrarian. You know about these releases, so you must know that their resellers have a consumer demographic.
I consume, as a consumer - which includes every producer because none of us only uses music materials for work aka in your terms "producer production": just to source, process, stock etc. - I consume hi-res consumer content recreationally, like the vast majority of their target demographics. And I'm not sure if you're simply generalizing a part of this market, or if you're making a case against the release of hi-res digitizations of any music source material, for any recreational consumers.
Certainly when taking what you're saying literally, that is what you are doing; but of course communication is too complex to know for sure. In any case, consider the consumption of hi-res releases, by a consumer who bit-streams it to a high quality digital-to-analog converter, going into an analog (pre)amplifier which easily handles the output signal of the aforementioned converter, ending up, in any circumstance which assumes no use of destructive ignorent elements in the signal chain between, playing back through high quality speakers or headphones.
What about this is 'not meant' to be?
I am an audio engineer with 15 years of experience, perfect video congrats! Highly understandable, perfect explanation bravo !
After reading and arguing about this stuff on the forums for so long, these videos (especially your latest ones) feel like a big therapeutic release, because all the stuff I've learned by doing a bunch of tests in my home studio are finally put into a form that is tangible and easily shareable.
His voice is so calming while also providing a complex yet easy to understand explanation
Honestly, your videos are near perfection. Straight to the essential point, very pleasant to hear. And most of all, supported with scientifical demonstration that make us have confidence in the knowledge you transmit.
I believe that your videos made me subconsciously create a growing percieved value for the fabfilter bundle, which I bought recently
I've watched four thousand videos and read ten thousand articles about this and you've finally demonstrated what REALLY matters. Thank you.
This explanation is really incredible. Thank you Fabfilter and Dan Worrall for educating the mixing masses!
Dan is the master! The pedagogical content of these videos is no less than amazing.
Dan, for years, I have searched for this conclusion. I spent two weeks chasing my tail trying to mix a solo acoustic guitar song: each corrective tweak created more deficiencies than what it solved. No matter what I did to "improve" the mix, it simply sounded like there were both dead spots and artificial bright spots that nothing could fix.
After watching this video, I tried a 'what-if'. I lowered the DAW session sample rate from 96kHz to 48kHz and made no other adjustments. Immediately, as if someone flicked on the sonic fullness switch, my mix sprang to life. Life will never be the same, thanks to this video. 48 kHz is my new sample rate. If a project calls for a track to make its way into the movie domain, I will simply upsample the 48 kHz master to 96 kHz to meet that format.
Thank you for this most excellent 'put the sample-rate argument to bed' video. Keep up the good work!
I'm sorry to say this with a delay of one full year. But are you aware that if you resample anything from 48 kHz to 96 kHz, the only thing you get is your files taking up twice the space they originally took, without any improvement in
sound quality whatsoever?
I think you misunderstood the original post.
I wasn't bored for a second. This is just phenomenal. As payment for this vast knowledge I shall buy Pro Q3, since the Thanks Giving sale is on!
ready to put it on linear phase at maximum and max oversampling? :D
Best video and information on sample rates and oversampling to date. I’ve been so confused for so long, so I feel like I’ve watched and read them all. I feel much more comfortable now. Thank you so much.
Converter design legend, Dan Lavry, wrote a white paper over a decade ago integrating his explanation of this issue with a somewhat math-heavy explanation of the Nyquist-Shannon Sampling Theorem. But I suspect this excellent vid from DW will reach more people and leave less head-scratching among those of us who didn't absorb much trigonometry in school.
So much useful stuff in just one video, it's unbelievable :
- Understanding why digital saturation/compression doesn't sounds right compared to analog stuff (i.e. guitar amps for example)
- What is aliasing, with clear examples
- Which effects are concerned and you need to oversample, with audible comparisons
- Advice on what to do for the best quality without destroying your CPU
- The incredible custom guitar chain with EQ for the cab, then 3 stages of saturation for the amp. I HAVE to use this !
- At the end of the day, technical stuff or not, if it sounds good it IS good. Maybe the best advice of this video.
Thank you so much Dan Worrall, for this free valuable resource.
absolutely this
Showing the vibrato aliasing and intermodulation at 20:00 is absolutely brilliant!
I'll have to watch again. Excellent content as always. It's not only a big learning factor, your video conclusions nearly always relaxes me, helping not to worry too much about possible quality losses. That's so great and I thank you very much.
I didn't think I'll grasp this concept, but I did. The explanation is stellar. Nothing was overly complicated with word salads, the examples were presented and explained in a way that's easy to understand. Great stuff right there!
I love these rabbit holes what take you right back where you started, but with more knowledge of why your are good where you're at. I could hear the difference between all the sample rates/oversampling combos. I think I've concluded that for my uses, mixing in 48kHz with oversampling where possible will do juuuuust fine. I'm actually considering getting Saturn now, but I'm not sure I need it with the other saturation plugins I have. I haven't even looked to see if they have oversampling. Mixing is one of those things where you can easily get lost in the details.
No you didn't.
Dan, Your videos give me a deeper understanding of complex topics and sooth me into a state of zen relaxation all at once. Your voice is amazing and easy to listen to and your approach to teaching these concepts is incredible. I have learned so much from your channel and I am looking forward to what you teach us next. Thank you so much for all your hard work it is much appreciated.
That was FABulously amazing! Thank you for the effort creating it and the generosity for sharing it.
Perhaps the most important video on mixing in the digital domain ever made. Dan you are a genius!
Thank you for quality explanations! Hope the 6 files (3 sample rates and OS on/off) could be provided in 44.1kHz 16bit wave format for off-line comparison.
Dan Worrall's knowledge and approach to teaching, visually representing, and offering of effective audible examples is peerless.
Especially well-explained and demonstrated concepts. All of it is valuable, perspective shaping information. Great work.
So far the best tutorials I have ever watched regarding music theory and music production.
My goodness, what a tutorial! Dan completely grabbed my attention for 30 mins when I thought I am not in a position to watch the entire video. Thanks a lot, FabFilter for the awesome plugins and Dan for the awesome tutorials.
Wow, what a teacher! I fell asleep halfway through this video, woke up still in awe, and continued watching until the end. I'm going to make a playlist full of this man's videos.
Please make a Masterclass Dan !!!!
That was a masterclass!
It was indees ! But I’m referring to a complete mixing course or masterclass.
Dan has own TH-cam channel and started being more active recently, keep an eye on it ;)
TzzSmk watched all his videos on his channel😅
@@BBDRecordsStudio Whats the name of his channel? Definitely interested!
This was one of the best educational videos about digital audio i've ever seen. Thank you so much for this.
The point of higher sample rates in my industry is so that you can pitch audio down without losing all high frequencies. Its VERY important for sound design
Nobody has ever explained this in such a way that I could fully understand it. Thanks Dan!
When is the Dan Worral Netflix series coming?
Honestly such a great video. Very informative.
When Dan speaks … I'm glad he finally addressed this issue. Chris "Monty" Montgomery at Xiph.org did an excellent couple of articles and videos on this years back, but they did not go nearly into the kind of practical depth that Dan does, and they somehow went over a lot of people's heads. This is why so few companies can touch FabFilter. They do the kind of thing you'd expect from companies like UAD, but really get from Airwindows and FabFilter.
When I hear the voice of Dan Worrall, I slap LIKE. Because of EPIC explication. Thank you.
Again you guys have done yet another video that's lightyears ahead of the rest, explaining a complex subject in clear way that doesn't dumb it down too much.
dan worrall could probably sell me air if he wanted to
Music is basically just fancy wiggly air
@@DeanLawrence_ftw Now THAT is too funny. And true!
After watching this, I feel like I’ve been to my old college lecture hall again. Amazing work and command of the subject!
If there’s a lot of pitching shifting and tempo change in the project higher sample rate may be beneficial
Love that short anecdotes of your youth/past.... gives the narration a human feel
First time I upvote every voted comment I see. The video is that good!
Fantastic video explanation! Truly eye opening
Who else smashes _Like_ on Dan Worrall videos even before watching them.
SLAP LIKE NOW)
That's silly. But, carry on!
Alguien por aquí ha dicho que deberíamos de pagar por vídeos así. Estoy de acuerdo! De lo mejor que he visto sobre esta eterna discusión acerca del sample rate en años. Cuestión zanjada en mi opinión. Thank you so much Dan for such a valuable information!! 👏👏👏👍👍👍
Best video ive seen this year so far
Fabfilter, yall literally make the best, most informational, deep yet approachable videos of anyone in the game! this stuff is the deep magic in the days of digital audio
The aliasing on that synth patch actually sounded good to me. Then again, I listen to a lot of autechre haha
And people pay a lot of money for MPC60s and EMU SP1200s......
I knew these things already from a mathematical and technical point of view, but I can’t remember anyone explaining it that easy to grasp. Thumbs up!
Reaper 6 dominating the DAW scene xD
Thanks for the wonderful education!!
Wow, I wish every audio processing video were as clear and concrete as this one. Wonderful job, thanks a lot for the learnings.
Thanks for a fantastic video Dan Worrall!!
I have been skeptical if the the higher samplingrates made any bigger difference. When using 96 and resampling down to 44.1, I never heard a better result than going from 48 to 44,1. Now I can stop wondering and just stick to 48k with a great explanation and relief :-)
This video demonstrated that you will hear a difference, assuming you apply digital processing of the signal to a common degree. What did you think the point was of him advocating the use of functions which use 8x the project sample rate, featured in the distortion plugins??
It just also demonstrated that you're better off using higher sample rates on a per-process/plugin basis, to avoid using resources on processes which cannot benefit from it (meaning not setting your entire project sample rate, based on the highest denominator which is your most aliasing-prone plugin(s) because it affects the CPU power needed for everything which doesn't audibly affect the signal as well. This has always been the case in out-of-the-box processing, because every analog signal processor, such as the type of distortions he's emulating here, was already 'upsampled' separately, by way of being an analog process, while the drum machine that the band used, was not unnecessarily digitally upsampled at the source, or at any digital point in the chain that didn't benefit from upsampling. But you will certainly hear the difference when you apply a digital process that aims to recreate the type of analog warmth that these distortion DSPs are doing. He's literally explained to you that you want very high sample rates for that type of processing, just that it's handy to have this function available and contained within each plugin, which in effect is just a way of getting closer to the way out-of-the-box mixing and mastering is done, assuming the inclusion of analog gear, which is virtually any music that isn't chiptunes.
@@JohnnyNatrium Thanks for your reply Johnny. I got it and that was exactly my point. Using plugins upsampling.. and not considering the AD/DA when using outboard or recording audio:-)
bandwidth limiting when using multiple stages of distortion to prevent aliasing.... that's what i've been doing!
also it is incredible how much of a different 41.1 to 48 makes this is an amazing demonstration
Amazing video! Could you also do the same video for resolution as that has the same myths. So many DAWs support 32bit floating now but so many still use 24bit. Would be great to know the math behind this as well.
My leaked experimental version of Ableton 13 supports 64bit flying point audio which it isn't even possible to record at yet, except on the prototype interface I stole from MI5's top secret recording studio (2 miles below London underground) when the guards were distracted by a girly mag I threw on the floor, so I'll see y'all later!
@@ncshuriken Reaper records in 64 bit FP
CPUs might actually handle 32-bit float better than 24-bit int depending on the algorithm. 24-bit actually has a ton of information, you probably won't hear it the difference, but I would not say the word "always" when it comes to computers
There's a good chance some DAWs will store 24-bit samples as 32-bit integers since CPUs handling 32-bit data better than 24-bit, plus it would increase mixing quality
So many years in the game.. and this rocks my world and completely explains everything regarding sample rates I've always wanted to know.
Dan Worrall!
Worked in massive recording studios and film sets for 14 years. This is the first time I've learned something in nearly 10!
The point is made at the end that higher sample rates can result in a higher chance of increased intermodulation (lower quality content), is this touched on at any other point in the video? Also he said that "higher samples rates do nothing to improve the quality of audible content below 20khz" isn't this contradicting what was shown earlier in the video in Acustica? Really great video by the way, I loved it.
Good question
He touched on this when using multiple instances of harmonics-generating plugins (saturation, compression) in a row: even at 96kHz the first plugin adds harmonics that will 'bounce back' off the nyquist freq. The second plugin then generates harmonics off of those 'false' harmonics and so on. The reason the chance of this happening more often in higher samplerates is human error: because you tend to not worry about aliasing and thus don't turn on the oversampling algo because you're already working in a high samplerate (so you think..)
Or at least, that's how I understood it...
@@gijscoolen6439 Ah ok, yeah it's the reasoning behind it being said that higher sample rates can be lower quality that puzzles me. Seemed to imply more than just forgetting to use oversampling. The statement "higher samples rates do nothing to improve the quality of audible content below 20khz" as mentioned in my first comment still puzzles me too.
I think the point is that increasing samplerate doesn't automagically mean better audio fidelity and cleaner signals. A lot of things have to align in the stars (as it were) for this to 'just happen' and that no matter what, with digital audio, you're going to have work around how digital audio naturally exists. Also, the point he made about increasing the sample rate was also that you'll use far more CPU with marginal results in many cases. It's like the differences between 4k computer monitor and 8k computer monitor. Your computer has to work 4 times as hard to give you a picture that literally no one can tell the difference unless it's blown up to the size of a movie theater screen and even then you'll only notice the difference once you get 5 feet from it, and no one likes watching a movie theater screen from 5 feet away.
people.xiph.org/~xiphmont/demo/neil-young.html
This was a great video and you broke it down in layman's terms where people could grasp the concept. In reality the better the original recordings are captured or managed can also play a very important role in eliminating some of these issues you are pointing out especially if it means being able to use less plugins. In my home studio I try to use plugins sparingly not only to limit how much processing power is required by the computer but just to keep things simpler. I also run a Hybrid DAW station with an analog mixer with the option to run external compressors and limiters when needed which allows me to capture a really decent analog drum mix where the DAW can be just more of a multi-track recorder and the same goes for capturing bass guitar and using an external compressor/limiter rather than using plugins to manage the bass guitar's dynamics and as long as you're mindful of the gains during recording, you're golden. For me and my needs this works but for a modern day commercial recording studio this probably wouldn't fly or at the least they would be limited especially dealing with many different styles of music due to the dynamics they may see like speed metal vs classical music along with music production for a movie etc and the effects required to make it deliver.
in case anyone wondered, youtube is cutting at around 15k or 16k (so you might still hear 20k Hz)
Oof I nearly forgot
Not for all videos. This video uses the Opus codec with a 20 kHz filter. The 20 kHz tone at 4:59 is actually clearly there (I checked with a spectrum analyser), allbeit a bit attenuated.
Check the stream you're receiving on the "Stats for nerds" option. If you're receiving an opus stream, you're fine.
not only Fabfilter makes some of the best plugins out there but they also do incredible and honest technical videos with the legendary Dan ! Thank you !!
Thank you for this demo, i've record in many studios and it concur with the knowledge i have acquired, and i do my own mix in 48khz, however after hearing your mixes at 44.1, 44.1 with oversamplig and 96khz, i hear a difference in the very highs specially on the cymbals and hit hat in 96khz like a bit more of "air" than the other exemple
i was arguing with a buddy because he swears hi res is the only way to go about it, if i only found this video sooner i could of saved my self a headache, really informative. and cleared some things up for me too. i have used fabfilter as my go too compressor and EQ for a few years but never use the over sample features at all. didnt actually hear the difference but now i SEE why its a feature
I only hear this type of music in a Dan Worrall video.
so true lol
@@funkyboy3326 Its a distinct flavor and its good
I NEED that twin 2 patch and midi
@Tronam and to think, I missed that, back in the day.
Thanks so much for helping us mortals with a proper technical explanation of when and where oversampling should be used. Also love the plugs!!!
Finally someone who understands everything about the sound.
This video is a staple in the hi-res audio debate of whether it's actually worth anything, and you do a great job at explaining how it's mostly just a bunch of people who don't know the details who obsess over hi-res. For example, I was one of those people, and then I watched this video lol.
I want dan to teach a 3-hour audio course for everybody at my funeral
Before I invested in my mixer and other outboard hardware, I did a lot of research into what was considered the optimum rate against cost benefit and return. I wanted to record bands multitrack during live performances and master them afterwards into usable acceptable marketing material. I had used my much loved and trusted Yamaha 01x for almost 20 years but it has been obsolete for over a decade since Yamaha refused to update the "firewire" drivers since Windows 98 so there after I have to run later versions of Windows in development/test mode for the mixer to work !!
For whatever reason I never found this tutorial, not that it would have made any difference, it is by far the best explanation over any other I watched of sample rates and diminishing returns on investment. I can see some highly regarded producers seriously argue with your opinion/conclusion, but like you I just chose to ignore their snobbish professional opinion of higher is better, understanding it didn't matter. So I bought my Behringer x32 over the more expensive Midas 32 variant or Allen and Heath knowing the quality might not quite be the same but still good enough for what I wanted to achieve. I also own a Liquid 56 so any extra quality could be overdubbed later, that is housed in my home studio cabinet that is way to heavy to be considered mobile. I just connect my Surface Pro through USB and I have a full 32 channel mobile recording mixer/studio with full remote control. I bought this start of lockdown as I was offered it at a great price including the stage boxes and interconnecting cables. Now I know some might comment on the Midas pre amps, I know, really I do, they are better, but I got everything for less than 1/3 the cost of the Midas 32 new.
hope one day I will send my child to Dan Worrall school...
I'm 7 years into music production & this video felt like I knew nothing about audio. Soooo much in-depth audio knowledge here.
this video taught me AND hurt my feelings