Improving CD 44.1kHz/16bit audio

[Mr. A]

Encoding CD shows, if you followed all this process, depends very much on not showing up to the party with sound that goes over 20 KHz. All these tricks go right in the toilet if you do that. Worse than creating distortion, the process creates new, clearly audible sounds that didn't used to be there.

Not a smooth road this Nyquist thing.

Koz

I see. But didn't they take care of the aliasing problem with making oversampling a part of the DACs job to do automatically so that did not occur? I read that oversampling was created really so they could get away with building a much smaller and cheaper low pass filter that would have a much much slower roll off but to no consequence because it would still roll off way before it ever gets to the Nyquist frequency plus eliminate the massive phase shifts that come with a high order filter. I thought oversampling was used to make things cheaper. It did say in my reading that the Nyquist frequency was double of the highest audible frequency. I guess that's what they want you to believe?

[kozikowski]

<<<It did say in my reading that the Nyquist frequency was double of the highest audible frequency.>>>

That's widely assumed, yes, but there needs to be Certain Tricks to get that to work. Clearly if you have a sine wave and x2 sampling and the samples end up in the gutters of the wave, there will be no digital output--or the outputs will be zero. That's why noise or "dithering" is added to the sound to keep that from happening. Without that noise--literally non-audio signals--the Nyquist value needs to go back up to 2.6.

My grown-up audio program has settings for that in the CD authoring section. I suspect it's available in Audacity as well, I just don't know where it is. There is a lively debate as to which dithering sounds best [sigh].

Koz

[waxcylinder]

Audacity can have its dithering managed from Edit > Preferences> Quality

WC

[przemek7]

Someone who mentioned about Dither should know that a dither is nothing but low noise that brings up any part of sound that is below dynamic range during transfer from 24bit system (e.g. studio session). So dither has nothing to do with bandwidth.

Now about bandwidth. In order to record higher frequncy that human ear (very yound human) can hear, we need to take at least twice much samples of that frequency, so if one wound like to record 1Hz he needs at least 2 sample per second.
CD sample rate is 44.1KHz at 16 bit means theoreticly that we have bandwidth from 20Hz to 20KHz and dynamic range around 96dB (1bit=6.02dB). Unfortunetly is not so.

The problem is not that 20Khz is archived poorly and is squered, but the fact that other lower frequncies are not camptred much better. For exampe, 10KHz is sampled only around four times during a freq. cycle (44.100 / 10.000 = 4.41). That means that all importent harmonics are gone, so transfering CD to any upper sample rate doesn't have any sanse.

I'd rather do it with Vinyls. Transfering them to the higiest sample rate, like 96KHz at 24bit or better would make sense (96KHz / 10KHz = 9.6 cps).

CDs should be replaced many years ago by DVD-a or SACD, but people who run this business
don't want to invest. Well, what is the point anyway, most people listen mp3.

Cheers,
Sorry for poor English!