|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
72.208.173.79
In Reply to: RE: Non Oversampling (NOS) DAC's posted by knewton on August 31, 2012 at 05:20:31
"1) There is the well known high-frequency roll-off of about 3dB at 20KHz due to the zeroth-order hold operation of R2R ladder DACs. I don't believe that sigma-delta DACs have this problem due to their high inherent oversampling operation, pushing any such roll-off way up in frequency."
The "rolloff" in NOS products is due to modulation taking place toward the upper limit of the working audio range, which is half the sample frequency. Although the average amplitude is indeed down at 20 kHz, the peak amplitude is roughly the same. (The filtering in oversampling DACs "smooths out" the modulation, raising the average amplitude to "flat" up to 20 kHz.) This could be why people don't "hear" the rolloff in NOS DACs. Some might perceive more HF energy, in spite of the measured rolloff. Although some of the HF energy is a result of the modulation.
This modulation is benign when playing simple music, but I think this is what obscures background information and inner detail in complex music.
Follow Ups:
Uh...no. The roll-off is due to the zero-order hold operation of most DAC chips. Which is more commonly known as, sample and hold operation. It has nothing to do with whether the DAC is R2R (so,ethin which I'd unintentionally suggested due to clumsy writing), or sigma-delta. Nor is it due to modulation. It would be interesting to know where you came by this notion, Todd. Perhaps, you are thinking of sigma-delta modulation. Which is a related topic, but not the cuase of the roll-off effect. The roll-off has to do with the 'stair-step' like shape of most D/A converter outputs.The high-frequency roll-off shape and transition frequency are a function of the sample rate at which the converter is running. The higher the sample rate, the more brief is the hold period. The hold period is the flat 'step' part of the output 'stair-step' waveform. Depending on the implementation of the oversampling utilized with a given sigma-delta converter, the reason such a converter might exhibit less roll-off has to do with it having a higher effective sample rate. This is also the reason why a true NOS converter, one running at the native rate of the input samples and not utilizing any external or internal oversampling interpolation, will exhibit the most severe in-band roll-off effect.
_
Ken Newton
Edits: 09/01/12 09/01/12
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: