|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
65.19.73.221
In Reply to: RE: CS4398: extreme opposite of the large tap length Chord DAC? posted by knewton on November 19, 2015 at 10:02:53
There is no definition of "objectively perfect" that can be applied to a DAC. This is a technical characteristic of the format. The format can not fully capture the analog signal that was sent to the analog to digital converter. In the process, information is thrown away. Depending on the specific method of throwing away information different playback strategies (DAC filtering or lack thereof) may offer different approximations to the original. However, there is no way to choose between these approximations on an objective basis. Indeed, there are multiple dimensions to the distortions and different listeners may weigh the different dimensions differently.
I have played around with different filters on recording and playback and can confirm the interaction of these filters. The relationships involved can be explained theoretically, can be measured in the time and frequency domains, and can (on some recordings) be heard as differences in tonality, air, sound stage and imaging.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
Follow Ups:
[There is no definition of "objectively perfect" that can be applied to a DAC. This is a technical characteristic of the format...]
Tony, I'm not sure what you refer to by that declaration. At any rate, what I wrote was "objectively the better", not objectively perfect. I read you to be a technically educated person, so, is there really any question that the closer a band limited digital audio signal's reconstruction filter approaches an ideal sinc function at Nyquist, the objectively better it is, all else being equal (quantization noise floor, band ripple, etc.)?Whether or not it's subjectively the better is a different question.
_
Ken Newton
There is no "objectively better" when one is measuring distortion along multiple dimensions, unless all the available objective measurements come out better for one of the choices. Given the nature of filter tradeoffs this is mathematically impossible. The tradeoffs are an inescapable result of the filtering required to conform to the Nyquist requirement. One such tradeoff is duration of ringing vs. high frequency extension.
If the recording is made with a good apodizing filter there will be no energy at the Nyquist limit and, indeed, no energy in a guard band close to the Nyquist limit. In this case, and only in this case, then the "correct" reproduction of the recorded data will be a perfect sinc filter. However, this reproduction will not correspond to the original analog signal, nor what the producers heard if their DAC used something other than a perfect sinc filter. My experience shows that recordings made with apodizing filters will sound best when played back with a perfect sinc filter, but most recordings are not made this way. Most recordings are made with an imperfect sinc filter and most of these will sound better when played back with an apodizing filter in the DAC. However, playing back an apodized recording with an apodizing DAC will produce an inferior playback, because of unnecessary high frequency rolloff and transient smearing.
These effects are audible and appreciated by recording engineers who have taken the time to train themselves to hear the distortions and related tradeoffs. Filtering associated with the 44.1 kHz sampling rate has been proven to be audible by double blind listening tests.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
I tend to agree with you. No of taps seems to me to be a marketing theoretical toll and does not relate to how a dac sounds.
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: