[TheOtherHobbes]

This is easy to disprove by downsampling from a 24/192 source to 16/44.1 Even if the downsampling is (close to) ideal there are obvious differences.

In fact if you can't hear the difference between 24/192 and 16/44.1 you shouldn't be working in audio. (Doesn't apply to consumers. Does apply to musicians and engineers.)

It's like being colour blind.

And if you don't understand the math behind quantisation, you shouldn't be posting pseudo-scientific videos where you use an oscilloscope and a cheap spectrum analyser - both tools with very limited resolution - to "prove" your point.

16 bit isn't enough for hard, objective reasons. One is that the noise spectrum of quantisation is not simple. Most people assume it's something close to plain white noise, but it really isn't. It's actually a very complex spectrum with some prominent peaks at specific subdivisions of the sample rate. Those frequency peaks are significantly above audibility. 24-bit quantisation shrinks them below audibility.

The other is that most people can hear dither/noise-shaping at 16-bits. That adds a single bit of noise which should - if you're being very literal - be far below the threshold of audibility. But it clearly isn't.

These two facts are related.

The more complex reason is that listening is an active perceptual process. The brain does a huge amount of processing to separate sources and place them in a perceptual field which includes information about perceived object type, distance, and ambience cues. Some of those cues are very quiet, and we don't hear them linearly.

So using sine waves as some kind of perceptual reference for audibility is nonsensical. We hear much more complex signals in an active way, and if there's information missing in the quiet parts - which there is with limited quantisation - then the signal simply isn't accurate.

[leothetechguy]

I agree with most of your points, but saying you shouldn't work in audio if you can't tell the difference between 192khz and 44.1khz is a bit elitist imo. And saying you're color blind if you can't tell the difference is like saying you're blind if you don't have 20/20 vision and shouldn't draw. You can always use meters to check for aliasing artifacts.

It's not like all of your samples and virtual instruments are 192khz or even 96k. Many are 48khz or even 44.1k.

I think there are many cases where people never need to go above 44.1khz unless you maybe have saturation on the master bus. I agree that good dithering is important though and think that there hasn't been enough research on that so far.

[mschulkind]

But surely the difference is 16 vs 24 and nothing to do with the sample rate?

[hdgvhicv]

The larger problem with 44.1 is

But it depends what you're sourcing from. If you source 44.1 then you will have a worse recording if you change it to 192. If you source at 48k then you just waste samples. If you’re recording analog inputs at 192k in a crappy adc then you will have a worse outcome than a good adc at 48k (or 44.1k)

Same with bit depth - the adc is far more important.

[spider-mario]

> 16 bit isn't enough for hard, objective reasons. One is that the noise spectrum of quantisation is not simple. Most people assume it's something close to plain white noise, but it really isn't. It's actually a very complex spectrum with some prominent peaks at specific subdivisions of the sample rate.

What you are describing is the result of blunt truncation. If you use the most basic (“uniform” or “rectangular” a.k.a. “RPDF”) dither, the spectrum is in fact flat, as demonstrated by the video you are likely alluding to and calling “pseudoscientific” (https://youtu.be/cIQ9IXSUzuM?t=12m50s). If you sum two uniform dithers together, you get what pretty much everyone uses (“triangular” or “TPDF” dither) which, in addition to decorrelating the mean quantisation error from the signal, also decorrelates the standard deviation, eliminating noise modulation and leaving a correlation only in still higher-order moments like skewness and kurtosis.

You can even try it for yourself with SoX. Find a 24-bit track, quantise it with dither to 16-bit, calculate the difference between both tracks, blow up the difference and take its spectrogram and it will be completely flat. Or listen to the difference (mind the volume) and see if you can make out anything meaningful.

    $ sox source.flac -b 16 dithered.flac
    $ sox --combine merge source.flac dithered.flac loud-difference.flac remix 1,3i 2,4i norm -1 spectrogram  # assumes stereo input
    $ open spectrogram.png
    $ open loud-difference.flac

And then remember that this difference would normally sit at roughly -93 dB FS, so to hear it in a typical room, you would have to be listening at deafening levels. You claim that it “clearly isn’t” below the threshold of audibility but it’s not clear how you arrived at that conclusion. You then claim that the audibility of that noise floor is somehow related to what you said before about the effects of undithered quantisation, even though those effects stop being relevant the moment you apply any sort of dither.

> We hear much more complex signals in an active way, and if there's information missing in the quiet parts - which there is with limited quantisation - then the signal simply isn't accurate.

It’s not missing. You can do a similar test where you “bury” your source material in the 16-bit dither noise floor, blow it up again, and you’ll be able to detect it under the noise.

    $ sox source.flac -b 16 quiet.flac gain -100
    $ sox quiet.flac loud-again.flac norm -1
    $ open loud-again.flac