> where do you draw the line?

You draw the line at the limits of human hearing, which turn out to be about 20 bits from softest audible sounds to "hearing damage inside a minute"and a sample rate of around 48 ksa/s.

CDs weren't quite there. Modern stuff recorded with 24 bit depth and high bit rates is effectively lossless.

One of the goods things about the limitations of the CD format is that we learned a lot about masking noise via dithering (deliberately introducing more noise) and noise shaping.

I've heard studio noise floor is typically about 15-20 dB(A), and typical quiet living room is usually more than 30 dB(A). In my previous appartement in the city I measured something in range of 45-50 dB, which was actually pretty high and surprising to me - and funny, I didn't hear that noise live until recording and hearing back.

Dithering gives about 120 dB of useful dynamic range, so I can't imagine a case where someone wanted more. You can't go much lower than the noise floor of the environment and you don't want to go too far above 100 dB, because this already isn't pleasant (nor good for your ears; 85 dB is considered the upper safe SPL limit for longer exposure). I'd risk stating that even without dithering, the default 96 bit depth is good enough in typical listening conditions.

Shaped dither can increase dynamic range, and anyway 120dB range is rather unnecessary for music.

https://web.archive.org/web/20200310174634/https://people.xi...

> In that case, any digitization is lossy

Yep, that’s part of my point, but CD audio in particular isn’t that great of bit depth or sampling rate, even by standards of the time for uncompressed digital audio. There’s also the point that the “audio wars” era covered the era of CD audio and that’s certainly another major audio quality factor.

Only one of these points is a thing. Even if you could find some situation where the dynamic range of a CD was inadequate, shaped dither or even complete garbage uniform dither would render that moot.

No producer or technician has any basis to complain about CD audio quality for listening by humans; the deficiencies of their CDs are entirely their fault.

> No producer or technician has any basis to complain about CD audio quality for listening by humans

Ok, so since I can still identify sound frequencies >22khz despite being now 40yrs old, I’m not entitled as a human to critique the CD audio format, especially as I can tell you the exact moment when the audio range is being clipped in the songs I enjoy?

It is highly unlikely that you can hear difference with >22kHz when listening to music. It could be just intermodulation distortion introduced by playing near-ultrasonics on normal equipment.

https://web.archive.org/web/20200310174634/https://people.xi...

Okay, even if you could hear frequencies in that range in an actual released musical recording, and not just at near-pain-threshold levels in isolation, 48kHz has become more popular since it's a more convenient "computery" number. Tell me you can hear frequencies above 23kHz, below the pain threshold, let alone not in isolation.

I think you’ve gotten off track here. I’m not claiming higher bit rates and sampling rates than CD audio are inherently better in every case, just that CD audio is not what I’d call lossless.

Do we need to cover the Nyquist frequency again? Do you think that all of the audio engineers that designed the CD audio standards were tyros who were just making shit up and guessing?

> Do we need to cover the Nyquist frequency again? Do you think that all of the audio engineers that designed the CD audio standards were tyros who were just making shit up and guessing?

Do we need to cover the site rules again? Is it assuming best intent to try and straw man my argument to serve what purpose, boost your ego?

1) of course the audio engineers knew what they were doing who created the CD audio standard.

2) of course they also had constraints from picking “best”, cost among them, and designed something that satisfied the demands placed on them as best as they could.

How does any of these tangents y’all have went on disprove the only point I made, that CD audio is not lossless (close, but no cigar) in the audio sound quality sense.