New Video! Nyquist-Shannon (digital sound part 2)
Thanks for your patience these last weeks as I worked through some deadlines. Enjoy the continuation of the Digital Sound subject! Please tell me if there's something I explained weirdly. I think I've contextualized this all now but ya never know...
(forgot to add:) Captions, cards, and the end-screen are coming soon, as per usual. Those take a while and I like to get this up for you ASAP
Comments (11)
I was wondering why 44.1 as chosen and how all that data was stored back in the day (would love a more in depth vid on it sometime) great work :-)
Monty's videos are wonderful (and I watched them ages ago so it's possible I forgot :D) but I don't remember the lightbulb in my head going off as strongly as when you explained that the low-pass filter is what fulfills Nyquist-Shannon and ensures the audio samples come out as the original waveform. I knew that the waveform was actually what was generated and the stairstep or direct line was a lie, but I don't think I understood the low-pass filter's critical role in making that happen. Even going back and re-watching Digital Show and Tell I feel like gaps have been filled in that I didn't understand previously. Excellent as always!
Damn. You're really earning the $1 I send you once a month.
I just realised jay formans one of ur patrons, thats awesome, also at some point can you explane why a 5GB ipod holds like 3000mins of audio, wheres a 750mb cd only holds 78, i know its due to compression but if like an indeph explanation
Simple bitrate; Red Book CD audio *must* be 1,411kbps, over 4x that of even the biggest MP3s.
Alec, in Part 2, would you please talk about why many CD players (from the '80s and '90s anyway — I haven't looked at any modern examples) advertise that they have a "1-bit DAC" as though that's some kind of impressive feature. It always seemed strange to me that CD audio has 16-bit samples, yet somehow a "1-bit" DAC is able to reproduce an accurate analog signal from the 16-bit samples. What's going on here? Thanks!
1-bit DAC's are now ubiquitous, but since they're no longer new, I guess they're not being advertised. Wikipedia has a page on delta-sigma modulation, which is used to produce the necessary waveform. It's also a very attractive way to generate audio from an FPGA, as it only takes one I/O pin (for mono) and a simple analog filter. <a href="https://en.wikipedia.org/wiki/Delta-sigma_modulation" rel="nofollow noopener" target="_blank">https://en.wikipedia.org/wiki/Delta-sigma_modulation</a>
Very nice explanation. Already knew about it, but it still surprises me how many people blindly believe there are staircases in the reproduced audio. That representation is always an intermediate state. But what makes me more angry is when companies promoting hi-res audio by showing the staircase scope image for CD resolution and the perfect sine wave for the hi-res. And CD audio is really good enough for human hearing. It's not the format causing why today's music sound bad, but the mastering.
And don't forget the whole 96-kHz (and even 192-kHz) craze. Unless you're reproducing music intended to satisfy your cat's sense of hearing, there is no reason to retain all those super high frequencies, yet Dolby (et al.) have somehow succeeded in scamming the market into believing that a 96-kHz sampling rate produces a superior audio signal versus 48-kHz.
Hey we have the same scope!
So good. And there’s so much rabbit hole you could dive into on this stuff. Keep it comin’!