AP Mastering: "Nyquist theorem debunked: why 44.1kHz sounds bad"

[Sokel]

Click bait for sure,I watched it.
All he did is compared late 70's digital gear with today's Delta-Sigma DACs to conclude that they are perfectly OK while the early stuff suffered from the known child diseases.

Enormous yawn.

 

[AnalogSteph]

Even early digital is quite arguably better than he's giving it credit for. I've heard some great classical recordings from the very late 1970s to early 1980s, even in original CD form and with potential need for SRC from 16/50. (What do you reckon they would have used, the very first Studer SRC or D/A out to A/D in?) Mind you, these were recorded with equipment that could be 100 grand. Early digital could be good but wasn't going to be easy or cheap at that point.

Either way, the comment section might be the most interesting part of this video, as it raised some interesting issues:

A number of folks on the recording side - but not everyone - seem to be aware that any kind of nonlinear effect VSTs (compressors, limiters, tape simulation etc.) can generate some nasty aliasing due to their output not being inherently band-limited, and some will oversample these to good effect. One guy applies audio band-limiting to 18 kHz to a 48 kHz signal instead - kinda backwards but essentially the same approach.

Working at 44.1 kHz straight does not seem to be entirely extinct, despite the fact that many ADCs do not deliver adequately "bulletproof" filter performance to guarantee an aliasing-free 20 kHz bandwidth (at least 48 kHz is often implied, and more can be beneficial) and with high-quality software SRC being available for free. AKM entirely dropping such a filter option on their current top-end ADCs compared to the previous AK5397 and AK5394A was a bit of a mistake IMHO. With CS5361/81/CS4272 filters being a "good but 48k+" option and limited/decreasing traction for TI's midrange to high-end options (PCM420x, PCM 422x), it was up to ESS to save the industry's bacon when it comes to this admittedly somewhat niche problem.

 

[Bergante]

I hope he debunks Shannon. I hate to spend money on large disks.

 

[Old_School_Brad]

Science fiction

Nah, it's just drama in exchange for money -designed to attract clicks and views through deliberate controversy. To me, that suggests the creator prioritizes YouTube ad revenue over accuracy. And if they occasionally decide to create content in good faith, it’s hard to fully trust them.

 

[Ktacos]

AP mastering

This guy is really good as being consistently incorrect. Also a really good example of how a lot engineers really don't know shit about anything works. He was smart to go into mastering, that's where the easy money is. I'm often a little bummed out at how success in the music industry often seems to have more to do with how well you can fool people.

 

[Blumlein 88]

This guy is really good as being consistently incorrect. Also a really good example of how a lot engineers really don't know shit about anything works. He was smart to go into mastering, that's where the easy money is. I'm often a little bummed out at how success in the music industry often seems to have more to do with how well you can fool people.

Also disturbing is that people who have been fooled in audio turn into pretty hard core spreaders of those beliefs. Their true belief makes them seem trustworthy. This helps them convince other people of misbegotten ideas. Such people are not reprehensible in the same way as those who know the truth and are pulling a con for selfish reasons. Both are detrimental.

 

[Waxx]

The NQUIST theory is not perfect, and neighter is pcm sampling. But it's by far the best we have, and in reality the resolution of digital is not limited by that theory, but by the converson to analog trough the devices that are availeble. But that limit is so high these days that it's not a big issue because that limit is already above the resolution of our ears, and a lot of other stuff is way more limiting than the NQUIST theory and the devices that are build on that tech. Speakers are way less clean, and are the real limiting factor in audio, not the electronicsn, and certainly not the higher quality ad/da conversion of today.

But i think it's true, NQUIST is not perfect, but it's good enough to not matter anymore in the bigger equation. Can we better it? Probally, but it won't happen in a clickbait youtube video without strong mathematical/scientific proof of why and how. And untill we have a scientifc proven better system we will keep the NQUIST theory as base of digital audio.

 

[Mnyb]

Edit: practical sampling is not perfect but NYQUIST is , it's math an provable , practical implementations does not exactly reach the theory but a such insignificant level that it does not matter

 

[kemmler3D]

I'm often a little bummed out at how success in the music industry often seems to have more to do with how well you can fool people.

FTFY ...

 

[boxerfan88]

A 100hp car is sufficient to get us to the grocery store and back. But don't we love having a 500hp sports car, or 1000hp supercar for the grocery store run.

16/44k1 is sufficient for 20-20k bandwidth, but don't some love having 32/768k

.

 

[voodooless]

A 100hp car is sufficient to get us to the grocery store and back. But don't we love having a 500hp sports car, or 1000hp supercar for the grocery store run.

16/44k1 is sufficient for 20-20k bandwidth, but don't some love having 32/768k

.

That analogy doesn’t work. You’ll have to imagine that in practice the supercar drives, feels and handles exactly the same as the 100hp car. So in practice, there is zero advantage in owning a supercar. The question is: what is then so “super” about it?

 

[boxerfan88]

That analogy doesn’t work. You’ll have to imagine that in practice the supercar drives, feels and handles exactly the same as the 100hp car. So in practice, there is zero advantage in owning a supercar.

Glad you understood my point.

The question is: what is then so “super” about it?

Because it has 1000hp; bloody difficult to get in/out of; sits so low and uncomfortable; noisy and loud; ride that is harsh to the ass; what else .... very few can afford it..
That's why it is so "super"

.

 

[voodooless]

Because it has 1000hp; bloody difficult to get in/out of; sits so low and uncomfortable; noisy and loud; ride that is harsh to the ass; what else .... very few can afford it..
That's why it is so "super"

Clearly you didn’t understand my point

 

[earlevel]

I watched the whole thing, so some of you won't have to. Really, it's one thing after another for the length of the video, but a few themes reoccur, especially noise floor. For instance, he never comes out and says something like, "We all know how noisy digital audio is, I'll tell you why..." instead, he just tells you why there's this terrible noise floor—see it on the chart he drew?

He thinks the theorem holds for the sample rate being twice, or more, the highest frequency. Of course, we know it has to be more, but the fact he does lets him show that at half the sample rate, a sine wave could land on the zero crossing and get nothing.

Then he gets hilariously wrong by inferring, from a graph, that a sine wave near half the sample rate will modulate in amplitude (beat). This isn't true, of course. The sad thing is it's so easy to check. Just record or generate such a sine wave in a DAW, play it back, and view with an oscilloscope.

He keeps saying over and over how what is taught in universities is wrong. As if professors and student would never bother to look and notice that their sine ave had this ridiculous problem that he suggests. He would have us believe that engineering students turn a blind eye to the theorem failing, and just accept and regurgitate it decade after decade.

 

[Hayabusa]

I just saw this suggested video on youtube from the channel AP mastering. I shouldn't be surprised as he put out another video where he was highly critical of some of ASR's highest rated speakers and saying that one of those terrible single driver monitors are great while ignoring multi-tone performance.

This time, he seems to have an axe to grind with the legendary Monty Montgomery video and how it has rightly been used to defend digital audio. His biggest gotcha's are kind of ridiculous and it appears that he just parrots the texas instrument's marketing material for hi-res DACs and ADCs.

Amplitude above and below the 0 line (1 out of 65,536 and the negative): That's -90dBfs. OK, not a big deal unless you plan to set your gain to +80dB on an incredibly quiet recording.

Another is that quantization error and aliasing exists, therefore delta sigma/oversampling with slow or no filters are the magic cure. While he's technically correct, he fails to quantify how insignificant this quantization error is. 16/44.1 without dither is good for -98dB or 0.00126% THD without dither and -120dB@3KHz with a modern shaped dither.

The funny thing is, most audio ADC's have a noise floor right at this 100dB mark and each track has a summing of this ambient and hardware related noise floor. This noise summing is far more significant than the influence of quantization error generated by 16/44.1. I do still believe that in the music producing world, there is certainly some value in 24/48K tracks, especially trying to keep that noise floor summing down. But, the final master can be 16/44.1K without any audible loss in quality.

What are your thoughts?

click bait

 

[antcollinet]

You’ll have to imagine that in practice the supercar drives, feels and handles exactly the same as the 100hp car.

And sounds, and looks. The only differences is the supercar (that looks, sounds, drive, handles feels exactly the same as the compact) has a display on the dash that says "very fast", and it consumes more petrol.

 

[solderdude]

He did not know that there are manufacturers making their own converter chips.
He states that only DS chips are used and are 1 bit, they aren't they are multibit.
He states that there are no 'old school' converters being used sold and stairsteps don't exist.. tell that to the folks loving NOS filterless DACs (or even emulations of it)
He handily forgot about the obligatory reconstruction filters used in the upsampling process and only talks about post-filtering.
He talks about 20kHz not being accurate at 40kHz sample frequency. Of course it isn't and why sampling frequency is 44.1 because of the steep (digital) filters which are easy to make.

Worthless video with lots of misinformation and some accurate ones like ... 44.1kHz is enough as a format. It just requires to be built good enough which is totally possible with todays engineering.

 

[Ktacos]

I don't think I've ever been this upset at an internet person before, every video is just endless misinformation and people eat it up.

Based on what he's said, I don't think the dude has much in the way of original thought. So much of what he says are practically tropes at this point in terms of the production world. Pay attention and you'll he constantly puts down others who work in a different manner than he does.

I bet his history has been -

1. Tried to make it with his own music, no one was listening (hey it's tough, not a big deal)
2. Probably went to or tried to go to school for the stuff (not necessary imo) and somehow failed, maybe had an instructor he didn't like which is why he thinks people are being taught incorrectly or not at all.
3. Tried working in a studio but was too hard to work or something
4. Thought to himself what's the easiest way to have an involvement in music making process, minimal effort but maximum credit, well it's mastering of course!
5. His rates are super low, $65 per track, personally I wouldn't get out of bed for that. I wonder what his day job is, nothing wrong with that of course but a man of such esteem should be making a living off it right?
6. Oh no, my rates are so bad I need another source of income, oh how about a youtube grift.


Titles every video with clickbait and always hops in the comments to say "wow can't believe this one stirred things up much". Just a completely insufferable person. Dude sucks, I'm going to bed.



(for all you ppl who make music, I loved that he said you can't do professional level mastering in Ableton. His reasoning is that Ableton doesn't display stereo waveforms. Spoiler, it does. Dude doesn't know how to use his tools but will readily tell everyone else that they're using them wrong).

 

[amirm]

I suffered through the whole video. As he admits in his own pinned comment, he was not able to clearly explain his point of view. He just goes on and on and keeps saying this and that sounds awful without ever demonstrating or explaining why. To wit, he says original digital content sounded horrible because of limited dynamic range. He doesn't realize that they beat analogy formats by a huge mile. He also claims that the "new" delta-sigma converters came after he was born. We had Crystal Semiconductor reps come to our company back in early 1980s, demonstrating their DACs. Pretty sure this was well before he was born.

 

[fieldcar]

(for all you ppl who make music, I loved that he said you can't do professional level mastering in Ableton. His reasoning is that Ableton doesn't display stereo waveforms. Spoiler, it does. Dude doesn't know how to use his tools but will readily tell everyone else that they're using them wrong).

I'm a Fl studio user, and I saw that video too. I swear it's just ragebait calling it by the old name of "Fruity Loops". He insisted that no one does mixing and mastering with FL Studio "just because" the pro's don't. Totally stupid conclusion. It's not like ever DAW these days doesn't mix at 32-bit FLOAT, and support VST's to do whatever mastering you need. It's ridiculous. Though, I believe he needs pushback on stuff like this. No Rage, just calm, collective, and objective proof that he is out of touch.

He doesn't realize that they beat analogy formats by a huge mile

The worst part is his criticism is based on how non-analog VST synthesizers have artifacts rather than natural real world audio. Almost every top of the line VST in the music production world will internally oversample. Funnily enough, the noise floor from something like vocals or any other real world recording would act like a layer of dither, because that's what dither is. I'm tempted to generate a +1/-1 level 1KHz square wave, then add a natural recording of noise floor from my audio interface. It won't HUM at -80 dB like he says it will, it will be totally gone. We don't listen to ultra-quiet test tones and we don't record real world audio at -90dBfs. The SNR increases the closer we get to full scale, just like analog. This is basic stuff, and I get why people are a bit irked that I even gave him the time of day.