Ripping SACDs via SPDIF at 176.4 with a PS3 - Page 2

Wavoman, I might help you with this one... I just realized that the Lavry only takes up to 96KHz through USB, but can easily handle up to 192KHz through SPDIF inputs (either optical or coax). Since I recently upgraded to a Macbook, I have now the possibility to play with TOSlink Optical and I am not limited to USB only, which means it could make sense for me ripping my SACD at 176.4KHz as well.

As per the file size limit, I am afraid the only solution is to rip only half on the SACD at a time, save the resulting file, and then rip the other half.... Makes the editing process a little more complicated after, but might well be worth the hassle.

I am unsure if I will ear differences between 88.2KHz and 176.4KHz...

Oups, I didn't realize the optical output seems to be limited to 24/96 on the Macbook as well ! I am afraid I will give up on this one and stick to 24/88.2...

Shamu did you DIY the Vanity board ? I had a look at the site and it just figures they crossed the Denon 3910 off their list. I just picked up a used 3910 for just such a purpose...tap that DSD stream and convert it to LPCM to feed my reference DAC.

I wonder if a master clock module upgrade would bring the 3910 into compliance ? I'll email Audiopraise for some details. Thanks go out to the OP and you Shamu for shedding some much needed light on this subject for me

Peete.

Sorry, it wasnt DIY... I had to send the player (ouch, 15 kilos both ways) to them so they could install the Vanity board... They did a perfect customized job.

They have been very responsive with all my inquiries... You should have no problem dealing with them.

By the way, look at what they say on the Nu Force web page presenting their inhouse upgraded OPPO 83SE version regarding jitter through HDMI:

The HDMI signal interleaves video and audio data which, upon decoding, produces an inherent jitter as high as 7nanoseconds (For comparison, a good CD player’s jitter is in the picosecond range)...

There is actually a white paper written by Dan Lavry that seems to conclude that above 24/96 is not recommended for recordings and playback... I am unsure of the conclusions myself as I am not a technician, but I would definitely trust this source.

I confirm I can hear differences as well between 44.1 (hybrid Redbook layer) and 88.2 (SACD layer)from the same SACD... Not night and day, but still a significant improvement (smoother, better 3D, highs extended and more defined...).

Thanks, I'll contact AP about the 3910 and see what I can finagle.

I'm not surprised HDMI is a bad idea for overall SQ. I was always dubious of packing everything into one massive "snake" cable anyway (analog/video/digital all together like that is crazy in my books).

Peete.

I tried this with a hdmi - > SPDIF splitter I got locally and it indeed appears to work although I don't have a DAC that can confirm I'm getting 176.4. One strange thing is that the SPDIF output doesn't work properly (scratchy background sound) with my Buffalo DAC but works with perfectly on my OPUS dac and my receiver. I've had similar things happen when I used my Buffalo DAC with cheap transports. I have to use my Audio-gd reference 3 to downsample to 96Khz, then it works with my Buffalo DAC.

I do not manage to get anything else than 44.1 resp. 48 via HDMI.
This is not limited to SACD, with everything I play.

I use the Porta splitter and a PS3 with 3.15 firmware and yes I did the right settings for the sound output. If I use the SPDIF output I get the frequency I chose (88.2/96/176.4/192) for everything (except SACD of cause).

I see 2-3 possible sources.
1. The firmware (Sony disabled something), for those who were successful, which firmware do you use?
2. My monitor (a HP LP2475w) has a SPDIF out. Perhaps it talks back to the PS3 and says it can not handle anything higher than 48?? I did not find any specifications on the SPDIF-out of the monitor, and if I connect the DAC to the monitor, the dac usually reports only 32kHz for the higher settings and produces strange sound (somehow contradicting the hypothesis of talking back). On the other hand If I leave the monitor switched off and play the music blindly via the splitter I get the same behavior.
3. The Porta splitter changed and does no longer support HD formats.

Any Ideas? ... other things I can try, other reasons, ...

Mistakedly, I made a double-post, see below...

Edited by mg2c - 5/12/10 at 5:33am

Hi
I tried to digitize a SACD using this very method  to output PCM via HDMI, then feed it to an HDMI->SPDIF adapter and record it from there using professional hard-&software.
I was using a professional AD/DA: RME FireFace 800, that can accept SPDIF-signals up to 192k/24bit, and professional recording software Samplitude V8 (sampling frequency matching the PS3's output, 32bit floating).
Analyzing the recorded signal (I can do that INSIDE the software, before even storing it as wav, so I am still inside the 32float-realm, with no additional errors...), I do get very strange, constant-levelled frequencies up to 88k (with 176.4 sampling rate) and up to 44k (when limiting PS3's SPDIF-output to 88.2k). These frequencies are not content-dependent. No matter, what music / SACD I feed, they are there at constant levels, the are up to -40dB in the region of 80kHz!
Attached graphs do show this.

I know, that when converting DSD to PCM, you need to apply some sort of dithering and noise shaping, but I am not sure, whether this in my eyes huge amount of "out of band" noise isn't problematic musically and sonically. Also, imho these graphs do not look like other noise-shaping graphs I saw on the web. Any idea, how I can avoid that and what is happening there?

The spectrum analysis of the output at 88.2k sampling rate:
 

 

The spectrum analysis of the output at 176.4k sampling rate:

 

 

Edited by mg2c - 5/12/10 at 5:37am

Hi mg2c! The noise which you are refering to - and which is clearly visible in your FFT graphs - IS indeed the DSD (noise-shaped) noise inherent in every DSD recording (the level may be vary a bit, depending on how loud the DSD master was transfered). So, this is nothing unusual, and that is also the point why I prefer transfers with "only" 88,2khz samplingrate - to avoid massive noise being send thru the DAC - AMP and speakers) This is (may) be the reason why Audiopraise (with their Vanity board) rely on filtering above appr. 32 khz. I've looked (tested so to speak) a lot of SACDs now, and there are only a few which have musical content (at reasonalbe level) above 25 - 30 khz. So, I don't care about the filtering anymore ... ;) But, may I aks you what kind of breakout-box (HDMI splitter) you are using? Cheers Harald

Hi

 

thanks for your answer. I am not so much worried, that I might cut off high frequencies when "ripping" the SACDs, but more, that this "loud" noise would lead to intermodulations and distortions, that would become audible. Just bear in mind, that indeed we do have out-of-band noise up to -30dB below peak level, and this is the signal that I will feed to my streaming client from my server later on. I don't know, how it will manage this huge amount of out-of-band noise...

As far as high frequencies are concerned: I don't want to go into the debate about audibility of high frequencies etc. - there are so many discussions and different opinions on that, that we should just exclude that in this thread ;-)  

But I do know, that measuring frequency range of SACD-players shows pretty flat response up to 80kHz at some players at the analog output, and there, they also can measire dithering noise, but this noise is between -100dB and about - 80dB, depending on the frequency. What I have gotten by the conversion DSD->PCM in my case is something completely different, hence my supscion, that something might be wrong, that the PS3 is doing something, that it should not or that the measured output would degrade sound-quality...

 

On the following graph, you can see measured Noise Shaping (I would feel much more comfortable, if my graphs would look like this )

 

 

I bought the breakout box on ebay, my seller's URL has expired, but this one here looks just the same:

http://cgi.ebay.co.uk/HDMI-4-Port-Way-HDMI-Audio-Breakout-Switch-Coax-SPDIF_W0QQitemZ230473540879QQcategoryZ109015QQcmdZViewItemQQ_trksidZp4340.m263QQ_trkparmsZalgo%3DDLSL%252BSIC%26its%3DI%26itu%3DUCI%252BIA%252BUA%252BIEW%252BFICS%252BUFI%26otn%3D20%26po%3D%26ps%3D63%26clkid%3D8586851555882451719