24-bit/192kHz music downloads and why they make no sense (2012)

No need to add any features to Ardour - it already supports any sample rate that your hardware supports.

There are places where (a) double precision (or better) floating point math benefits DSP, but that's nowhere in Ardour (and likely, if one is clear about the definition, in any other DAW either) - certain types of plugins can benefit from this, but they should not impose that cost on the rest of the processing infrastructure (i.e. they should convert internally and then back again, rather than require that the whole host uses 64 or 80 bit floating point.

There is a reasonably good argument for 96kHz because of the possible characteristics of the brickwall filter and its impact on aliasing. This argument gets a lot weaker for even high SRs.

However, both of these are largely theoretical in the sense that very few people, if any, can reliably hear the results in actual produced music or soundtracks. So while it might be a good idea to use higher SRs, whether that actually results in something that can truly be said to "sound better" is much more questionable.

At the highest level, I think that all of this is pretty irrelevant. Most of the music that people consider "great" was recorded with equipment far below the quality levels achievable with mid-priced pro-sumer gear today (mics perhaps being the sole exception). Great music/great sound is, I think, appreciated largely independently of its audio "fidelity". There is a huge difference between an amazingly well-recorded ensemble and a poorly recorded one, but for most people, if the well-recorded ensemble is playing stuff they don't like and the poorly recorded one is playing some stuff they love, all the "fidelity" in the world won't improve the former and the lack of it won't stop them loving the latter.

In addition, with the rise of deeply impressive sample libraries and better and better synthesizers running as plugins, the AD/DA elements of music during "recording"/composition are becoming less and less important for more and more music. That amazing patch someone uses in Onmisphere doesn't get better or worse by using a higher SR, and that incredible string library (e.g. New Albion) is what it is regardless of what your converters can do.

So, in short, I would say: do what you want, but if you're going to try to justify it with science, make sure the science is right and if you're going to try to justify it with "sound", then be aware that for most people the differences won't matter (or even exist at all).

> 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