4 ms·
I agree that if we’re pursuing sheer audio fidelity, CD quality is objectively superior to vinyl in every way. However (as others mention), many vinyl listeners
by electrograv 7y ago
I agree that if we’re pursuing sheer audio fidelity, CD quality is objectively superior to vinyl in every way. However (as others mention), many vinyl listeners are not after objective sound fidelity perfection: For them, the flaws are part of their “retro” charm.
However, one note about CD quality — it may not be as perfect as you think! A peer-reviewed meta-analysis has shown that humans can hear the difference between CD quality (44khz/16bit) and “high-res audio” recordings (e.g. 192khz/24bit)! The perceptual difference is relatively small, but measurable (assuming high quality speakers and signal chain):
https://phys.org/news/2016-06-people-difference-high-resolution-audio.html https://phys.org/news/2016-06-people-difference-high-resolut...
I doubt 44khz sampling rate is a problem, but fixed point 16 bits per sample does seem fairly constraining, given how our perception of linear audio loudness corresponds to exponential increases in signal strength.
Also, if anyone tries to test it yourself, beware that it is actually quite hard to do so properly:
1. Any listening tests in a web browser will just resample to CD quality automatically anyway, because they just play into your OS’s shared audio mixing system, which mixes all audio sources before they reach your DAC, which usually resamples them all to CD quality in this process.
2. The above point applies to ANY app that doesn’t take exclusive control over your PC’s audio output capability, and even then there are many pitfalls where you can almost accidentally trigger a default CD quality resampling stage somewhere in the pipeline.
3. Be careful to ensure the audio file you’re playing wasn’t just a regular CD quality file that was simply upsampled to 192/24 or whatever high res rate it claims to be. Sadly, this happens a lot. Your audio file MUST be recorded AND mastered (edited) in the studio at every stage using high res formats.
4. Make sure your audio signal path doesn’t have any components introducing dynamic range (loudness) compression (many default to this, these days), or any other resampling, equalization, etc. stages.
5. Your DAC may not even support anything higher than 44/16 data streams, so resampling may be inevitable!
6. Make sure you played the audio samples on extremely high quality speakers (often costing many thousands of dollars, making this whole point moot/irrelevant for most people anyway). The vast majority of consumer headphones and speakers can’t render a low vs high bitrate MP3 very differently, let alone the subtle differences we’re talking about here.
- cosmotic 7y agoAgreed; The default signal processing and lack of transparency on the sample rate and depth conversions through the whole system are a pretty frustrating aspect. The dust/scratch factor of vinyl has a much bigger impact than 16 vs 24bit or much of the signal processing issues.
- microcolonel 7y agoDo you know where to find the actual meta analysis that article is talking about? They mention the journal, but not the title. I'm not really sure what their criteria were. To me, it would seem that in order to compare the two, you would basically need to resample the 24/192 recording, with dither, and then resample it back to 24/192; and you would need to choose a high quality resampling filter. Anything other than that I wouldn't really consider a comparison. Furthermore, given that the measured sensitivity is more likely on the frequency front than on the dynamic range front, did they establish a distinction between 44.1kHz and 48kHz?
- ddeck 7y agoThe study is here: (pdf) https://qmro.qmul.ac.uk/xmlui/bitstream/handle/123456789/13493/Reiss%20A%20Meta-Analysis%20of%20High%20Resolution%202016%20Published.pdf https://qmro.qmul.ac.uk/xmlui/bitstream/handle/123456789/134...
- microcolonel 7y agoThank you! It really is interesting, it seems like there's an opening for a better-designed experiment. Wonder if there's anyone at the local universities interested in that sort of thing.
- archontes 7y agoIn addition to this, audio engineers mastering albums often have different masters for vinyl vs. CD resulting in different dynamic ranges. It's not just a matter of reproduction. Humans make the products different.