• Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

PCM vs. DSD: Encoding Method Accuracy, SINAD, and What Actually Limits Modern DACs

DavidEdwinAston said:
Anybody who vacuums the carpet, while listening to music, should be burned at the stake! :)
Click to expand...
Even Freddie?
urbs-freddy-urbs.gif
 
DavidEdwinAston said:
Anybody who vacuums the carpet, while listening to music, should be burned at the stake! :)
Click to expand...
And they probably bang on endlessly about "loudness wars" or the price of fish....
 
AUDACC said:
Do modern DACs treat the two formats equally.
  • Modern DACs do not process PCM and DSD with equal accuracy in practice. Most popular delta-sigma DAC chips (ESS Sabre, etc.) run an internal multibit PCM-like path; DSD is converted/expanded internally and generally measures a few dB worse than the chip's best PCM SINAD on the same hardware.
Click to expand...

The generalization is tricky.
With SDM DACs PCM (input) ALWAYS gets upsampled and goes through the SDModulator before the final reconstruction filter for conversion to analog.
Unlike DSD (input) in conjunction with DACs that provide a direct DSD path (TI, ROHM, AKM, depending on implementation) - here the source is passed through untouched and only the final reconstruction filter is applied before conversion to analog. So for these DACs you would get different results.

See for instance here for the audio path of the ROHM DAC: https://www.audiosciencereview.com/forum/index.php?threads/smsl-d200-dac-review.65764/post-2520980
 
AUDACC said:
  • DSD can sound different from PCM of the same recording due to several measurable DAC-level mechanisms: differing reconstruction/low-pass filter characteristics (typically gentler in DSD mode), the character of ultrasonic DSD noise interacting with amplifiers and speakers, and the choice of delta-sigma modulator oversampling rate. These are real, measurable, hardware-level differences — not mystical properties of the 1-bit format.
Click to expand...
This conclusion is suspect. The AI has integrated audiophoolery into its analysis.
 
thon said:
The generalization is tricky.
With SDM DACs PCM (input) ALWAYS gets upsampled and goes through the SDModulator before the final reconstruction filter for conversion to analog.
Unlike DSD (input) in conjunction with DACs that provide a direct DSD path (TI, ROHM, AKM, depending on implementation) - here the source is passed through untouched and only the final reconstruction filter is applied before conversion to analog. So for these DACs you would get different results.

See for instance here for the audio path of the ROHM DAC: https://www.audiosciencereview.com/forum/index.php?threads/smsl-d200-dac-review.65764/post-2520980
Click to expand...
Good catch. It would be more accurate for the original post to say that “Many modern DACs using the ESS chips have a greater SINAD for PCM than DSD especially if the bandwidth for noise measurement is 90Kkhz.” This is because the DSD slow roll off output filter allows ultrasonic noise to bleed through and be measured with the analog audio signal.
 
AUDACC said:
Please be specific about what you think is audiophoolery in this thread.
Click to expand...
That’s easy—the entire paragraph that I quoted promulgates myths believed by audiophools. For example, the statement “DSD can sound different than PCM” is an impossibility. Or the myth that “ultrasonic DSD noise interacting with amplifiers and speakers” is audible. Or that different reconstruction/low-pass filters produce audible results, unless they are very poorly implemented. Everything in the paragraph I quoted is audiophoolery, not backed up by even a rudimentary understanding of human hearing.

But that’s what you get when you rely on an AI/LLM to make rash generalizations and conclusions. It spits out whatever audiophile nonsense it has been trained on—and there is a lot of it on the web. Garbage in, garbage out.
 
Last edited:
Interesting. I remember using DSD technology for the STL (studio to transmitter link) for FM radio back in the 80s when we were still trying to figure out digital audio. It worked, and sounded good enough. My recollection is that it was a dbx product. I had no idea that it was still a thing.

Instead of taking a sample of the voltage, it just measured whether the voltage went up or down compared to the previous sample and stored a one or a zero accordingly. Seemed to be about a wash as far as bitrate was concerned.

The obvious problem with DSD is that you can't edit it. And at the time, the product that was all the rage was a digital tape recorder that allowed you to razor blade edit the tape, just like analog.

Ahhh, the good old days...
 
Billy Budapest said:
That’s easy—the entire paragraph that I quoted promulgates myths believed by audiophools. For example, the statement “DSD can sound different than PCM” is an impossibility. Or the myth that “ultrasonic DSD noise interacting with amplifiers and speakers” is audible. Or that different reconstruction/low-pass filters produce audible results, unless they are very poorly implemented. Everything in the paragraph I quoted is audiophoolery, not backed up by even a rudimentary understanding of human hearing.

But that’s what you get when you rely on an AI/LLM to make rash generalizations and conclusions. It spits out whatever audiophile nonsense it has been trained on—and there is a lot of it on the web. Garbage in, garbage out.
Click to expand...
The phrase “DSD can sound different from PCM” should be more precisely stated “To some listeners playback of a recording that was delivered in a PCM file can sound different than the same recording delivered in a DSD file”. Intermodulation and harmonic distortion of ultrasonic frequencies passed through the gentler DSD reconstruction filters to amplifiers and speakers can fold back into the audible frequencies. Whether its audible or not will depend on the listeners hearing and the magnitude and frequency of that distortion.

Anecdotally some listeners claim to hear a difference between PCM or DSD playback of the same recording and anecdotally some claim no difference. Can you offer some proof that a difference can’t be heard?
 
Recorded music is not the same when delivered as DSD and PCM. In any case one format needs to be transferred in to the other one. And how well ist this transformation process? So it may be that there is an audible difference between the formats whatever the reason for this is.
 
What has dsd got to do with anything?
 
You must log in or register to reply here.

Similar threads

D
Questions About DSD and Audio Sample Rates
Replies
5
Views
2K
danadam
danadam
Magos Vox
The Concordance - where a DAC's marketing meets the truth of its silicon
Replies
3
Views
525
Magos Vox
Magos Vox
M
SMSL DO100 Pro Advertises DSD512 and 384 kHz — But No ASIO Driver Exists
2
Replies
31
Views
1K
Ruediger
R
ManniX
foo_dsd_trellis: foobar2000 DSP plugin for DSD and PCM
Replies
0
Views
599
ManniX
ManniX
A
Is Bit Perfect DSD possible through Foobar2000 to Topping D50iii
2 3 4
Replies
68
Views
5K
AUDACC
A
Back
Top Bottom