Ripping SACD's for Separate CD and DVD Tracks

That was acknowledged from the beginning but, to be clear, any BD player that can rip the DSD layer will play the DSD layer.

HDMI is digital output and is standard on all BD players.

:+1:

1 Like

If playing an SACD to a DAC that supports DoP (DSD over PCM), there are little boxes on Ebay that will take the DSD stream off of the HDMI signal and produce a DoP stream for a DAC to use.

Those little boxes would be breaking HDCP though.

These guys are still in business:

Nope. They work fine extracting the DSD stream, I have one.

I’m sure it works (still), but since S/PDIF’s copy protection is very basic, it can be bypassed easily without the possibility of revocation, so that basically enables you to rip SACDs through HDMI.

S/PDIF has NO copy protection. Yes, you’re effectively breaching the copy protection in HDMI. This is a way to PLAY SACDs directly to a DAC that admits DSD over DoP.

Ripping the SACD to DSF files is entirely different - there’s no special box, just an appropriate Blu-Ray player and software. Yes, it decrypts the encrypted SACD data thereby breaking its copy protection as well.

https://absolutelybaching.com/music-articles/how-to-rip-an-sacd/

I bought a used Sony BDP-S490 for $35 and have successfully ripped 40 SACD’s to ISO and converted these to 2 channel DSF for Roon.

1 Like

S/PDIF uses SCMS copy protection scheme, but that’s just a copy protection bit in a control word. You can of course reset the bit if you capture and retransmit the stream with another box.

But since you can capture S/PDIF, you can also rip.

Once you capture the S/PDIF stream, you can extract DSD from DoP and save as DSF.

In order to play the disc, the player has to decrypt, right? The breach is not in the decrypting of SACD stream inside the player, it’s in making it available unprotected, outside HDMI, by the injected software. That’s also what that little box does, in a slightly more legal way.

Given there’s a method to capture the DSD stream directly, that would be insane to do.

If you need only stereo for Roon, than it’s much easier and better, to extract the DSD stereo files as DSF and not coverting something:

I’m using always stereo und multichannel files in Roon, because my exaSound e68 DAC can handle DSD multichannel through USB.

1 Like

If you want to add DSP (room correction) to 5.0 / 5.1 DSD64 contents I’d suggest rip it as DSF files and playback from HQPlayer (HQPlayer can apply DSP on DSD stream directly without conversion).

1 Like

Well, there is some conversion happening even in HQPlayer. Before applying the actual DSP, samples are widened from 1 bit to 64 bits, then high-frequency quantization noise is filtered out. After the DSP, samples are reduced from 64 bits back to 1 bit through a delta-sigma modulation process. The only thing that HQPlayer doesn’t do compared to Roon is a down-sampling and then an up-sampling of the sample rate, before and after DSP, respectively. Whatever you use, if you want to do DSP during playback, DSD sources are not ideal.

1 Like

So far I know HQPlayer won’t convert DSD to PCM for applying DSP. @jussi_laako would you explain more about this?

It is not any different from PCM. PCM is not any more ideal than DSD in this respect. Only difference is that PCM is much lighter to process since the data bandwidths are so tiny.

PCM needs requantization through dither/noise-shaper which is equivalent of remodulation with SDM.

Roon DSP as well has the Enable Native DSD Processing (i.e. 2.8224 MHz, 5.6448 MHz, etc., floating point sample rate) option if that is your thing and your CPU is up to the task.

AJ

Thanks, good to know. I wasn’t sure if Roon supported that, and I didn’t bother to check. I think enabling native processing is useful for very simple DSP like gain. For convolutions, processing at megahertz rates is as waste.

That’s an important difference.

Given that the un-dithered quantization noise for 24-bit PCM is around -144dB, dithering and/or noise shaping at that bit depth is just a gimmick. Nobody can hear that, even if correlated, and no downstream analog components can come near that kind of performance.

How? Only if you are short of processing power.

And for DSD it is less than -180 dB, much lower than 24-bit PCM in any case… Not hard to exceed what 32-bit PCM is capable of.

But that’s only after dithering and noise shaping, which you’re comparing to unshaped, un-dithered PCM. That’s not a fair comparison.