D/A and A/D | Digital Show and Tell (Monty Montgomery @ xiph.org)

> 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” ( ▶ this video ). 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

@xiphmont also made an amazing video response to the many responses he received to this article. Using analog equipment he busts a bunch of myths and demonstrates what really happens with digital audio. link or on YT if you can't play it ▶ this video

Changing the sample rate of audio only affects the frequency range. All audio signal is _perfectly_ represented in a digital form. I am ashamed to admit this took me a long time to properly understand. For further reading I'd recommend: link ▶ this video

Did you actually try any searches? Or is this just an excuse to broadcast your feelings about AI? The author of the Ogg format claims a bit more pessimistic range of bit depth: ▶ this video Here are some measurements of type I, II and IV: link link link Here are the specifications of a typical dbx unit: link

Agreed. The fact that a pixel is an infinitely small point sample - and not a square with area - is something that Monty explained in his demo too: ▶ this video

By "perfect" I just meant that digital audio perfectly reproduces the recorded analog signals. ▶ this video I didn't mean to indirectly praise Bluetooth in my post. Bluetooth anything pretty much sucks. Bluetooth audio in particular is pretty bad and full of usability issues. It doesn't seem to have the bandwidth required since the audio gets transcoded to some lossy format. I once set up an mpd music server on my local network and audio quality was perfect. However I encountered significant latency issues. Play and pause had a latency of one second which made it unusable. That was true even for uncompressed audio streams over the network. I got bored before I was able to resolve the problem. Maybe the problem was my network. I should try it again now that I have a much higher performance router running OpenWRT which is capable of traffic shaping.

This is the most lucid explanation of aliasing I think I've ever read. I mostly deal with this stuff in the realm of audio where the matter of analog vs digital rages regularly. The reason one hears "Nyquist says" so often from people defending digital audio is that the pro-analog people are imagining stair-stepped signals in which the information that is "lost" results in a noticeable degradation in sound quality. This (11-year-old) video is the gold standard for addressing this concern: ▶ this video Curiously, no one ever seems to mention the actual problems (like aliasing).

This is simply not true. Nyquist’s sampling theorem tells us that sound sampled at 44kHz will reproduce all frequencies in the range of human hearing. There is no “quantization”. ▶ this video Further the act of mastering and creating the record, and playback using a needle, will inevitably affect the sound somewhat. In the scenario you describe the music is likely to sound very good, but it will never be identical to a CD.

There is no system high-end enough to allow your ears to magically hear 96kHz, or even 48kHz. Unless you’re very young with golden ears it’s extremely doubtful that you can hear a 24kHz tone. It is absolutely snake oil. link is an overview by the creator of Ogg and Vorbis explaining exactly, scientifically, why these inflated sampling rates are snake oil. ▶ this video is a longer video by the same person that debunks a lot of digital audio myths.

In sound, nothing beats personal experience, experimentation and hours of deliberate listening. Spend a couple of hours with a basic compressor built-in your DAW and try to get the feel and understanding of how each parameter affects different sounds (from sine wave and noise to voice to drums, etc.). If you feel it's not fruitful, read, watch videos, listen, work on your stuff and come back to it after a while. Regarding books, there's lots to choose from. For some quick applied wisdom, Bobby Owsinski's Mixing, Recording and Mastering Handbooks (separate) are recommended. Zen & The Art Of Mixing by Mixerman is very good, as well as his classic The Daily Adventures. The Bruce Swedien Recording Method is very good if you're into recording. For Youtube videos, some random things I can vouch for: I'd suggest watching Monty's Digital Show and Tell [1] as a primer for digital audio as a form. Steve Albini has been putting out some very solid and entertaining videos on recording lately [2]. There's also of course a few hyped (but deservedly so) channels such as @DanWorrall, @Producelikeapro or @ProductionAdvice. Nothing I can suggest for courses as I haven't been interested in them for a while. [1] ▶ this video [2] link

> Audiophiles might point out that analog audio sources can theoretically hold details you can't hear in digital. This is true because values in analog media are continuous. Over imperceptibly short periods of time, the numbers that represent different values on a sound wave in a digital recording jump instantaneously from one discrete value to another with no gentle slope between. Analog signals instead form smooth curves, so at some level they can contain tiny variations in slope that are too subtle for the digitization of that signal to represent. > The problem with analog superiority in general is that, when that digital signal is of standard CD quality (44,100 samples per second at 16-bit resolution), whatever details in the analog signal are too subtle to be captured digitally are also too subtle for me or really most people to hear. This so wrong it makes my head hurt. It's 2023 and people still think that digital doesn't produce smooth curves and that typical analog media -- cassette tape no less -- can store more details than a CD. These people need to be tied to a chair, have their eyelids taped open, and force-fed Monty Montgomery's video. ▶ this video

Nyquist-Shannon simply says that you need 2x the sample rate of the highest frequency to perfectly recreate something. So, to recreate a 20 kHz wave that was sampled at 16 bits, you'd need 40 kHz sample rate. Bump that up to 22.05 kHz (for anti-aliasing) and you get CD audio. The bit depth, OTOH, is your signal-noise ratio (noise floor) when sampling/quantizing said audio. It's unrelated to Nyquist-Shannon. More bits gives you less noise when reproducing it. For CD audio, 16 bits was chosen to give an acceptable SNR of 96 dB. Monty Montgomery (of xipf.org) has a nice YouTube video related to this: ▶ this video

Let Monty Montgomery of xiph.org provide an illuminating tear down of “digital can’t really produce...”: ▶ this video Edit: fix auto-incorrected xiph.org domain

link The talk ▶ this video <-- Youtube backup in case it doesn't work (YMMV) I wish I could force any audiophiles to watch it before they waste their money on snake oil. I think a similar point can be made about analogue synthesis, e.g. You pay Moog £1500 for what is ultimately a fancy box for a relatively simple analogue circuit.

A great video which touches on this idea, but mostly in the context of audio: ▶ this video (relevant part around 8 minutes)

Aside from the history and opinion in the piece, there are technical errors that are addressed by this: ▶ this video Anyone interested in the argument over digital audio vs. analog will enjoy the above.

Exactly. I like Monty Montgomery from Xiph's explanation video[1] on Sine waves, sampling and digital stairsteps. [1] ▶ this video

Monty Montgomery's explanation of digital audio might be the best explanatory video I have ever witnessed, ▶ this video

It's worth watching the video in my references [1] - it's a proper hacker's view of digital audio and how it translates to the analogue domain. Even with a 20khz sound you're not going to get any aliasing at 44khz sample rate. Most people's hearing starts to fall off significantly after around 12khz or so (depending on age). Most cymbals have a fundamental frequency around 2-5khz in my experience. Also, because cymbal sounds are pretty close to being noise, it's arguable whether it would matter or not. Although if you listen to a 128 kbps MP3 then you can often hear the top end frequency sounds (like cymbals) becoming 'mushy'. To the vast majority of people once you get to 192 kbps it's inaudible, and then above that most people with 'perfect ears' would struggle. And that's with a lossy format, never mind the lossless formats of CDs, WAV or FLAC which is the discussion at hand. [1] ▶ this video

Try watching this video (hope I found the right one): ▶ this video I think you're right that recording equipment has a long way to go, though; regardless of format I think people can relatively easily distinguish real acoustical instruments from recordings.