|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
67.188.250.219
In Reply to: RE: PCM digital recorder posted by tailspn on May 01, 2014 at 13:37:12
. . . why do I always see (via spectrographic software) the ultra high-frequency dusting of noise from DSD sourced recordings, whereas most of the PCM sourced recordings look clean up in the same regions. It may be all PDM at some level, but, obviously something is causing this difference - right?
BTW, your statement "recording engineers who record acoustic music originating from a natural space (unlike a studio recording) prefer DSD to converted to PCM as more like the analog mix" is one I think no one could possibly take issue with. But by the same token, I'd think that most engineers would prefer PCM (if the recording started out that way) to converted-to-DSD. I mean, just on general principle, that's the way I feel.
Follow Ups:
The issue is whether or not the high frequency noise has been filtered out or not. When converting from DSD to PCM it is possible to filter out all of the high frequency noise and produce a clean looking spectrograph and some conversion software does this. However, this comes at the cost of steep filtering and this usually creates worse artifacts than the high frequency noise assuming one's analog amplifiers are suitably designed.
When DSD is converted to PCM there will be noise starting just above 25 kHz from the original modulator, creating this dilemma. If the source was one bit DSD at double speed this noise will be moved up the f-scale and filtering will be less critical. If multi-bit sigma delta modulation is used then the DSD noise will be reduced (6 dB per bit) and so will require less steep filtering.
Keep in mind that the only problem with the PCM format is its (generally) low sampling rate. It is this low sampling rate that creates the need for filtering on recording. Conversely, the only problem with DSD is the limited number of bits, creating a lot of high frequency noise that will be difficult to filter out, especially since the SACD standard for DSD was set by Sony and Philips at too low a sampling rate. They made the same mistake with SACD that they made with 44.1, namely setting the sampling rate based on technology in place at the time of their decision rather than optimizing the system for technology expected to arrive shortly, according to Moore's Law.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
Hi Chris,
Tony hit the nail right on the head. It's the necessary decimation filter when converting PDM to PCM at a lower sampling rate that diminishes the DSD originated high frequency noise. But with this comes the phase shift effects and non linear time delays of any low pass filter, exacerbated by the steepness of the filter.
The point I was trying to make was if all A/D converters begin with PDM, and the type recording (classical acoustic music and jazz) requires only editing and level balancing, with no mixing or effects processing (all of which require PCM conversion), then leaving the recording and listening chain all PDM (DSD) less deteriorates the sound quality available from the microphones.
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: