|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
I'd read Dan's comments in archives of various web boards this spring, but I believe this white paper was only posted recently on Dan's site.Lots of material here -
Mike
also see his site, www.lavryengineering.com, and see the "support" page, where this white paper and other good info reside
Follow Ups:
The major weakness of Dan Lavry's main arguments against the higher sampling rate (eg file size storage costs & reduced accuracy at higher sampling rates) is that they become increasingly less valid as technology continues to improve.Even with today's technology, though, conversion chips that sample at higher rates, everything else being equal, generally are constrained to generate and to be designed into environments having less jitter which is an absolute improvement. I think this is analogous (yes, I still prefer analog) to the situation where many DVD players sound better with CDs than dedicated CD players(not high end) do - the allowable worst case performance is more tightly constrained for the higher throughput media and repro equipment in certain audibly significant aspects.
Plus, I haven't been convinced that getting sampled information more correctly placed in time, even down to the low microsecond range, doesn't improve the musicality of reproduction since good musicians cue to under millisecond intervals and occasional phase lock even happens between different musicians playing at the same fundamentals and harmonics.
whilst i don't necessarily completely disagree with Dan, i thought the paper was a bit simplistic or too "theoretical" - it ignores practical and real world considerations that may favour 192/24 over 96/24 over 48/24.in any case, the original reason for high sampling rates and high bit depths was not necessarily because "more is better" - it was to give recording studios headroom during the recording and mixing process.
therefore, given that the recordings are already in a particular sample rate/depth, why not preserve that in the consumer delivery format? i would prefer no resampling or bit depth flattening, regardless of whether there is any theoretical benefits from high sampling rate in the first place.
Both Hi-rez music and Classic Records have two sided DVD-As with 24/96 on one side and 24/192 on the other.....Listen to them....make up your own mind based on true science (that is, your own ears/the best instrument).....I have nothing against theories and science....but the ear can tell what is more accurate. Listen on a stock machine....then listen on a higher resolution player. You will hear....hearing is believing.
As does 176.4 on the "new" Neil Young DVD-As. I'd be much happier if I could understand why they sound better.
Wish I hadn't taken my back medicine, cause everything sounds good.
I don't have a sore back, so it should work even better on me. ;-)
Regards,
Geoff
the Classic Records "HDAD" won't give a "fair" comparison since the 96/24 tracks are downsampled from the 192/24.a fair comparison should have 96/24 and 192/24 versions independently recorded, and one shouldn't be a resample of the other.
192 is an exact multiple of 96 has same bit-depth and no dithering is necessary.Can you explain to me why an analog -> 96/24 ADC should differ from analog -> 192/24 -> 96/24?
You can just point to a book / paper if you don't have time to explain and I try to look it up.
.
Thanks for the linkWhat I like about it is that it relates sampling theory with real-life signal processing. I'm not sure the argument about larger files and transmission issues is relevant, in a world where the price of storage is rapidly heading towards $.50 per Gb, and where bandwidth is a commodity.
The argument about sonic distorsions would be far more serious, but I have no way to assess the point.
I have to admit that I haven't yet found a major instance when the sound was obviously better in a 24/192 capture, or at least not as obvious as in 24/96 over redbook.
mt
This post is made possible by the generous support of people like you and our sponsors: