DSD-2-PCM -- proof of concept, test sample and source code here Options

[SebastianG]

The ZIP contains 4 files:
- dsd2pcm.jar (the converter written in Java, outputs raw 24/88 PCM, intel byte order)
- info.txt (contains some infos)
- test2822k.dsd (14 seconds, mono DSD, 5 megs)
- test44k.mp3 (conversion result)

have fun,
Sebi

Edit-2011: For the latest C/C++ source code see https://code.google.com/p/dsd2pcm/ It is released under the new BSD license. The attatchment to this post is still the first Java release.



This post has been edited by SebastianG: Dec 4 2011, 17:45

Attached File(s)

 dsd2pcm.zip ( 3.56MB ) Number of downloads: 2486

 

[spoon]

The final amusing aspect to all this is that Scarlet Book spec recommends that SACD players incorporate some low-pass filtering (either 50 kHz or 100kHz) to lessen the amount of ultrahigh frequency content the downstream gear would have to (perhaps badly) deal with.

Makes interesting reading:

http://grad.uprm.edu/tesis/latorrenavarro.pdf

In addition, Sony claims to have developed a decimation procedure, named Super Bit Mapping Direct Down conversion, which uses a “super-power one-stage” FIR digital filter/noise shaper with 32,639 taps to down-convert the DSD signal into a CD quality signal in a single stage [1]. This allows for a CD with the highest possible audio quality.

According to manufacturers, DSD can offer a frequency response from DC to 100 kHz and a residual noise power of -120 dB thanks to its 5th-order noise shaping filters [1]. However, given this noise shaping technique, the residual noise power is not constant throughout the signal bandwidth. By using noise-shaping filters of higher order, it is possible to increase the resolution in the audio band at the expense of resolution at exceedingly (inaudible) high frequencies. These results are valid however only for static, non-transient signals. Transient signals will have poor resolution in a one-bit system. If the signal does not endure for a long enough time, the noise shaper of the one-bit system will not minimize the error.

As indicated by its proponents, DSD offers the advantage of eliminating the decimation and interpolation processes necessary in PCM modulation, thus preventing additional-unnecessary re-quantization noise from polluting the signal. In the SACD "Scarlet Book" Sony and Philips mandate the use of a 100 kHz low-pass filter in SACD mastering so that when the playback volume at standard DSD level is equivalent to 100W, the noise component "outside the audible sound spectrum" is 1W or less. In addition, SACD players must low-pass-filter their analog output above 50 kHz.

[krabapple]

The final amusing aspect to all this is that Scarlet Book spec recommends that SACD players incorporate some low-pass filtering (either 50 kHz or 100kHz) to lessen the amount of ultrahigh frequency content the downstream gear would have to (perhaps badly) deal with.

Makes interesting reading:

http://grad.uprm.edu/tesis/latorrenavarro.pdf

Interesting listening tests, though also a bit curious. Not sure how someone managed to get 81.5% correct in a 16-trial listening test of 24/88.2-->redbook-->24/88.2 vs original 24/88.2, as his Table 1 shows -- not because I expect it should be hard to do (though I do) but because the math doesn't work:
13/16 = 81.25% correct (p = 0.011)
14/16 = 87.50% correct (p = 0.002)

So how does one get 81.5% correct?

His threshold for calling a non-null result seems to be p<0.01, by which criterion one, not two as he says, listeners could tell the difference. Or did I miss something? I also assume (because the author used Audition), but do not know for sure (since he didn't say so), that proper dithering was performed during downconversion. And as always a later retest of the 'best' listeners would have been nice, to see if their results were a fluke.

Finally, my experience with audiophile naysayers suggests they would argue from the other side: namely that using headphones limited to 22 kHz for recordings with a 44 kHz bandwidth cripples the test.