4 ms·
Wouldn't lossless Audio need to be some kind of analog? Or some fourier transformations? Isn't the quantization of digital already a loss over the continous nat
by KingOfCoders 5y ago
Wouldn't lossless Audio need to be some kind of analog? Or some fourier transformations? Isn't the quantization of digital already a loss over the continous nature of sounds? I have no clue about audio.
- tootie 5y agoShort answer: no. Digital audio can reproduce analog audio flawlessly already. Chris "Monty" Montgomery has some excellent videos on the topic that I was coincidentally listening to just this morning. This Verge article sums it up and links to his really excellent talks. https://www.theverge.com/2015/11/6/9680140/chris-montgomery-digital-audio-hi-res-explainer https://www.theverge.com/2015/11/6/9680140/chris-montgomery-...
- KingOfCoders 5y agoThe point was not "flawlessly" but lossless.
- tootie 5y agoLossy and lossless has to do with the encoding not the signal processing. I'm just saying it is entirely possible to replicate an analog signal with a digital intermediary. Losslessness depends on the software.
- sergio 5y agoI understand that the lossless part is only the compression from the original digital file: what you play is bit by bit what the publisher uploaded. Just like BMP vs JPEG. Not about maintaining the analog resolution.
- bobbylarrybobby 5y agoIn this context, lossless just means "the same quality as it was originally recorded in, which is sufficiently high for all practical purposes". And since our hearing only goes up to 20kHz, if you record at a sample rate of 44.1kHz -- just above the Nyquist frequency -- you'll be able to faithfully reproduce the original audio to within the precision afforded by your bit depth (generally 16 or 24 bit). Sort of like how a RAW image is considered lossless because it contains every single pixel captured, even though it can only represent finitely many distinct colors and cannot zoom in infinitely far with no degradation.
- ZoomZoomZoom 5y agoSmall correction: "bit for bit identical sound data as it was rendered at the final stage of production (mastering)"†. Technically, our hearing plays no role in this, the requirement is just preserving the same sampling rate and dynamic range. †NB: industrial standard is preparing a master pack of different files from the same final render, including lossy mp3.