Is Hi_Res Digital Audio Dead? :-)
I was forwarded an article today.
It speaks of the frivolous existence of delivering hi-res digital audio. You know, anything above our Redbook Audio CD standard of 44.1kHz 16bit PCM digital audio. OK, so certainly I have a bias here…
If you read the article it discusses a recent year long research project involving blind testing of CD playback vs a Hi-Res format (DSD). However, my immediate reaction was that they were all missing the point!
I do respect and admire the Mix Magazine author, but I have to believe that he knows the presented angle of the research misses the mark. The argument for hi-res audio, for me, has always been based on the increase in available processing power and not in the additional delivery of stored information to the consumer.
My experience has shown that a hi-res recording of the “basic” tracks of a mix really does improve the quality of the final mix. If the DAW has 24 bits of information and over twice as many samples (96kHz) to work with in ALL of the DSP involved, it will be a better mix. ALL of the calculations will be improved, throughout the life of the project.
The use of higher resolution (both sample rate AND bit depth) leads to a better final mix regardless of delivery format.
I’d love to see a research project look at the underlying presence of processing etc in the digital audio world rather than focusing on that last element of delivery format!
| 
I think you could compare this with digital images. Nobody could tell the difference between a 20 megapixel and a 8 megapixel picture if its in a 10 inch screen, but that doesnt mean that you dont need the high res image to produce the final result.
This is a fascinating topic, and it has brought up many heated discussions in the Pro Audio world. Thanks for posting!
Have you ever done blind, level-matched trials to confirm what difference you hear?
Btw, the Meyer & Moran paper doesn’t argue against using >16 bits for audio recording, production and DSP — no one disputes its utility for that, and that use has got a sound technical basis. And dither and noise shaping let the benefits of that trickle down to 16 bit delivery formats.
The ‘mark’ the authors were trying to hit, was the hype about the intrinsically better sound of DSD vs CD as home delivery formats, which they demonstrate simply isn’t supported when they are compared fairly — that is, when it’s not the MASTERING and production that’s being compared, but just the formats. (THey did find that the noise floor difference of DSD and 16 bit becomes audible, as expected, when system gain is boosted up past normal listening levels — 14 db gain, I think.)
Kriman, Yes, that is my main point. Comparing the final on the basis of the storage only seems odd to me.
Steven, yes, it only makes sense that the storage medium would give the same results given the same source material, but that is a bit like watching analog tv on an hd set, I’m not sure of the need to make that comparison…
Here’s an interesting article from a digital audio conversion engineer that suggests that higher sampling rates (192 kHz+) may actually degrade audio quality.
http://www.lavryengineering.com/documents/Sampling_Theory.pdf
Leave a Comment