musicandaudioguide
  • All Recordings Are DSD…NOT!


    The discussion over at CA also included a great deal of push back on the DSD vs. PCM debate. One of the writers presented numerous lengthy comments about how modern ADC and DAC converters operate. Recall that the title of the post was "Mark Waldrep is claiming that PCM 24/96 is superior to DSD".


    I highlighted the comments made by this gentleman because he specifically called out both John Siau and myself.

    "Let's look at the recording and playback chain:

    Virtually all ADCs take the analog signal through sigma-delta modulation into DSD format internally as a first step. It can either be left as DSD, or if the studio is 'recording to PCM,' internally the DSD format will be decimated to higher or lower resolution PCM. Note that this conversion from DSD format to PCM (and to lower rate PCM from higher rate) is a lossy one.

    Then nearly all DACs (all delta-sigma DACs) take the bitstream, and if it is lower resolution PCM they first interpolate it to higher resolution PCM, then send it through a sigma-delta modulator to obtain a DSD bitstream. All the interpolation and modulation steps are lossy conversions. The DSD bitstream then goes through the final reconstruction filter for conversion back to analog.

    That's what happens with a 'PCM' recording. Everything you read where people complain about the deficiencies of DSD describes everything you have ever listened to through a DAC, the sole rare exception being if you listen only to material recorded before the late-80s/early-90s takeover of sigma-delta ADCs and do so only through an R2R DAC. That's it. So when John Siau and Mark Waldrep talk about the sound of DSD, they are describing an unavoidable stage of every 24/96 recording Mark Waldrep ever made and everything John ever heard from his sigma-delta DACs. If they say they prefer the 'sound of PCM,' then what they must prefer is what is being lost during the lossy conversions to and from DSD format.”

    So I sent this description of the process of AD and DA conversion to John for his thoughts. He replied below:


    "Modern PCM sigma-delta converters produce much lower error signals than 1-bit sigma-delta DSD converters. The errors in the DSD system are due to the 1-bit quantization that occurs in 1-bit sigma-delta DSD converters. Multi-bit PCM sigma-delta converters can be fully dithered and do not suffer from this un-dithered truncation. Every added bit reduces the noise signal by 6 dB. A 4-bit sigma delta converter is 24 dB quieter than a 1-bit sigma-delta DSD converter. Right from the start, 1-bit DSD signals have higher losses than multi-bit PCM signals.

    Conversion from 1-bit DSD to multi-bit PCM is a lossless process inside the audio band. The only thing that is removed is the out-of-band noise above the Nyquist limit of the PCM system. Nothing else is lost. <strong>Don't believe the DSD marketing hype</strong>.

    Conversion from multi-bit PCM to 1-bit DSD is always a lossy process. The loss is due to the 1-bit truncation. This truncation introduces a very large ultrasonic error signal that makes the ultrasonic region unusable for audio.

    Processing a 1-bit signal to create a 1-bit signal is also always a lossy process. A volume control is one of the simplest processes in a multi-bit PCM system, but it creates a large error signal when applied in a 1-bit DSD system. The same is true for any other 1-bit to 1-bit DSP process. The lossy part of these DSP processes is the quantization back to 1-bit. Cascaded 1-bit truncation processes can rapidly degrade the audio quality. For this reason, DSD is always processed as multi-bit PCM.

    Any DSP process applied to a 1-bit signal produces a multi-bit signal. No loss of information occurs until this is quantized back to a 1-bit signal. Why incur the loss by going back to a 1-bit signal after the processing chain?

    All practical DSD systems require some sort of DSP processing (gain control, mixing, filtering, etc.) and all of these processes produce multi-bit PCM results. Taking these lossless multi-bit results and adding loss by truncating them back to a 1-bit DSD signal makes absolutely no sense. DSD complicates the processing and adds unnecessary losses to the signal path. DSD does not simplify the signal path. There is absolutely no truth to the marketing hype that claims that 1-bit DSD is a simpler system than multi-bit PCM. The exact opposite is true.

    John Siau, Benchmark Digital"

    I'll parse through John's response in a future post but the gist of the argument is that the CA comment conflates 1-bit DSD with PCM sigma delta conversion. Yes, modern converters make use of internal sigma delta conversion at very high sample rates BUT they are not 1-bit DSD processes, they use multiple bits making them very different than DSD.

    Thanks John. It's really nice to have really smart friends.