3 ms·
CD may be 40 years old, but human ears didn't change at all in that time. If you can tell CD from other recordings, it is almost certainly because the mastering
by cormacrelf 7y ago
CD may be 40 years old, but human ears didn't change at all in that time. If you can tell CD from other recordings, it is almost certainly because the mastering (which is not something intrinsic to the standard) was deliberately changed for CD because it could reproduce sound better, so no format-specific mastering hacks were needed to make it sound good. Face it, you don't have golden ears.
For example, if a studio today were to master something for cassette, it would dramatically turn up the high end to compensate and probably bypass some saturation or exciters because the format has its own. Similarly, any studio not tweaking their masters for the perfection of CD was probably doing the material a disservice. Heck, today, some tracks are deliberately mastered to sound good on iPhone earbuds.
(The usual article cited in these discussions is https://people.xiph.org/~xiphmont/demo/neil-young.html https://people.xiph.org/~xiphmont/demo/neil-young.html, which happens to make this point briefly under "Different media, different master" and "Better masters". Edit: another fun article is https://www.soundguys.com/high-bitrate-audio-is-overkill-cd-quality-is-still-great-16518/ https://www.soundguys.com/high-bitrate-audio-is-overkill-cd-..., with the fantastic pull quote, "There is no safe listening level to hear the difference between these files".)
CD doesn't need to be the pinnacle of digital technology, because our ears don't know any different beyond it.
- numpad0 7y agoDid you know: Nyquist sampling theorem assumes, and only applies to, pure sine waves. Apparently this is a novice level knowledge in DSP community at StackExchange: https://dsp.stackexchange.com/questions/45032/suitable-sampling-rate-for-triangle-wave https://dsp.stackexchange.com/questions/45032/suitable-sampl...
- unlinked_dll 7y agoThe Shannon-Nyquist theorem doesn't assume or apply just to sines (that's a gross mischaracterization of the proof and it's implications). It assumes ideal samplers and reconstruction filters can exist, which they can't. In the real world, ADCs and DACs' inherent filtering, noise, and distortion characteristics are quantifiably superior to any analog storage media and reproduction system.
- jdsully 7y agoRecords don't have infinite bandwidth either. I'd actually love to see a comparison of vinyl vs digital reproducing a triangle wave, it would be a lot more illuminating than blindly quoting Shannon Nyquist.