|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
99.249.86.176
In Reply to: RE: Hi-res is dead. Apparently. 12 reasons why. posted by jusbe on November 18, 2015 at 02:09:21
That is, you only need to sample at twice the frequency you intend to reproduce to capture that frequency perfectly. And very, very few, mostly very, very young people can hear 20kHz much less higher. Thus the adequacy of the 44.1kHz Redbook standard.However the trick part, I would surmise, has always been the filtering of frequencies above the maximum target frequency, (22kHz for Redbook), that are just noise. The advantage of high resolution is in this realm, i.e. it is easier to design filters with fewer artifacts, (such as phase shift or ringing), when the digital frequencies are higher to begin with.
I might suggest, as a non-expert, that advances in DAC design are almost entirely a matter of filtering with fewer artifacts. BTW, I'm extremely pleased with my Schiit Bifrost Multibit with its "closed form" filtering; the level of realism is astonishing, at least at that price point.
I love the music of Dmitri Shostakovich
Edits: 11/19/15 11/19/15Follow Ups:
"That is, you only need to sample at twice the frequency you intend to reproduce to capture that frequency perfectly. And very, very few, mostly very, very young people can hear 20kHz much less higher. Thus the adequacy of the 44.1kHz Redbook standard."
Maybe in theory, but a lot of us can hear the time domain distortions at HF, even if it's below 20 kHz.... If we couldn't, there wouldn't be a large group of people who believe no filtering (non-oversampling or "NOS") was the best approach to CD playback. I prefer "short" filters personally.... Which don't adhere to Nyquist as well as "long" filters, but I prefer how they sound.
"However the trick part, I would surmise, has always been the filtering of frequencies above the maximum target frequency, (22kHz for Redbook), that are just noise. The advantage of high resolution is in this realm, i.e. it is easier to design filters with fewer artifacts, (such as phase shift or ringing), when the digital frequencies are higher to begin with."
If the sample rate is high enough, I think all filtering could be analog, which would significantly simply the process.....
"I might suggest, as a non-expert, that advances in DAC design are almost entirely a matter of filtering with fewer artifacts. BTW, I'm extremely pleased with my Schiit Bifrost Multibit with its 'closed form' filtering; the level of realism is astonishing, at least at that price point."
At the 44.1 kHz sample rate, there is no filter that is "good enough" to where it would be indistinguishable from 15 ips analog tape.... The filters have advanced to where the obtrusiveness at HF has become tolerable in a lot of cases. But still not completely transparent. (Vinyl tells me this whenever I listen to it.) Schiit is one of the better recent products.
Had the CD been sampled at 48 kHz, we may not have had half the problems we've been having.... The Nyquist frequency would have been high enough (24 kHz vs. 22.05 kHz) where the HF artifacts could have been significantly lower.
The high-rez movement was based on the notion that the resolution of 16/44 Redbook CD wasn't sufficient for true high-fidelity audio playback. I agree with that notion personally, but I've yet to experience a high-resolution digital format that I prefer over CD.
Feanor: " However the trick part, I would surmise, has always been the filtering of frequencies above the maximum target frequency, (22kHz for Redbook), that are just noise. The advantage of high resolution is in this realm, i.e. it is easier to design filters with fewer artifacts, (such as phase shift or ringing), when the digital frequencies are higher to begin with. "Todd: " If the sample rate is high enough, I think all filtering could be analog, which would significantly simply the process....."
This is the approach take by the new PS Audio NuWave DSD; it oversamples, does DAC using delta-sigma, then uses an analog filter.
Feanor: " I might suggest, as a non-expert, that advances in DAC design are almost entirely a matter of filtering with fewer artifacts. BTW, I'm extremely pleased with my Schiit Bifrost Multibit with its 'closed form' filtering; the level of realism is astonishing, at least at that price point. "
Todd: " At the 44.1 kHz sample rate, there is no filter that is "good enough" to where it would be indistinguishable from 15 ips analog tape.... The filters have advanced to where the obtrusiveness at HF has become tolerable in a lot of cases. But still not completely transparent. (Vinyl tells me this whenever I listen to it.) Schiit is one of the better recent products.
...
The high-rez movement was based on the notion that the resolution of 16/44 Redbook CD wasn't sufficient for true high-fidelity audio playback. I agree with that notion personally, but I've yet to experience a high-resolution digital format that I prefer over CD. "The Schiit approach is quite different from PSA's NuWave approach. Schiit oversamples and uses a "closed-form" filter that preserves the original PCM data points, then uses a high-accuracy R2R (ladder) DAC. I haven't heard the NuWave but the Bifrost Multibit is remarkably good, especially with 16/44.1. I'd love the hear Schiit's more up-market Gungnir Multibit or Yggdrasil that use the same principles, but they are out my price range unfortunately.
I love the music of Dmitri Shostakovich
Edits: 11/22/15
I am extremely pleased with my Schiit Gungnir Multibit. I have never heard 44/16 sound so good. However, the sonic benefits of true hires recordings are very discernible (listen to the sampler from AIX records). Having said that, the reproduction of redbook standard music by the Gumby is so enjoyable that I don't feel the need to rebuy stuff I already own. I do occaisionally buy hires for new releases that I can discern the provenance of the recording but I'm very satisfied with the sound of well recorded music from my existing redbook digital library.
For the multibit upgrade.
nt
I love the music of Dmitri Shostakovich
is that the signal fed to AD converter has no component above half of the sampling frequency.
So filtering technique is also needed in A to D conversion.
It is this theory about high sample rates for facilitating filtering that has informed my current digital player, I think.
Big J
"... only a very few individuals understand as yet that personal salvation is a contradiction in terms."
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: