DSD to PCM (1 Viewer)

[Chris_C]

Kevin, I'm with you on the fence about the conversion to PCM. Why do time alignment for DVD-A but not SACD? From the other sources I've seen discuss this issue, the DAC processes DSD in a different path from PCM. I believe the manufacturer of Burr-Brown Dac's even have diagrams depicting the path of the bits and DSD is handled diferently from PCM. I'm not too concerned about time alignment, but maybe I'll be contacting outlaw for an ICBM.

 

[Lewis Besze]

I've heard this before too, but no official acknoledgment from any of the manufacturers,so till then I stick with Denon's "story".

 

[Chuck Kent]

I thought that the only absolute need for DSD to PCM conversion is the ability to set delay. I didn't think BM has any bearing on the need for conversion. (I suppose it might be easier to do BM in PCM but I'm no engineer.)

 

[JohnMW]

Please excuse my lack of knowledge, but since I have the Denon 2200 I'd like to have a true DSD signal when I'm listening to SACD's.

With this is mind do I simply go to the audio setup, turn the FILTER off, set channel level to default (0dB), and the delay times to default?

The only thing I've noticed with the filter being off is ALL speakers run as large, and in most cases people have a small center/surrounds, especially in my case (Klipsch RS3 v2, RC3 v2). I don't think I could run high volumes with all speakers set to large. :frowning:

 

[Lewis Besze]

John,
the "filter off" will set all speakers to large and sub "on".This is to be done to accomodate those rear recordings that has the "height" channel encoded normally into the "LFE channel".The manual blabbers about bass boost but it's not true.The boost happens when you set all speakers to "large" manually,if you select the filter to "off" after that, it won't boost the bass at all.In other word if you the BM[select all speakers to small] you will loose some bass output from the sub,which is the opposite effect what should happen.I hope I didn't confuse you even further.

 

[JohnMW]

Lewis, I appreciate the reply. However is it then true that if I set all of my (main) speakers to small & sub to yes, enable the +10dB subwoofer channel level, and set the delay times to the correct spacing of my speakers that I retain a pure DSD signal?

 

[Michael St. Clair]

There's always the possibility that these 'non-converting' chips actually convert regular 1-bit DSD to "DSD-Wide" 8-bit DSD.

"DSD-Wide" is actually a form of PCM.

It's really all about whether it is transparent or not. I'd suggest judging a player based on how it sounds, and don't trust semantics either way.

 

[FeisalK]

I appreciate the fact that people want pure DSD if they think they are paying for it, but the mere fact we are having this discussion says (to me, at least) that 1) the conversion to PCM and back is imperceptible under normal listening curcumstances and 2) any perceived degradation may be psychological as a result of reading this thread.

I read somewhere that music on a lossy compression method with proper BM and TA may even sound as good as DSD without proper BM and TA. Then it boils down to which is more important, the sound or signal purity.

So echoing Michael above let your ears be the judge.

 

[Lewis Besze]

The 2200 and the 2900 don't do time alligment on SACD.Other then that, you got it.

 

[BeatCrazy]

Originally posted by John Kotches:


John, while I know your background and have great respect for the work you do over at WSR, I think it's a bit bold to make this statement as a fact.

If I was to tell you that I disagree and provide you with internal Sony documentation that proves otherwise, would you change your views?

Sony's own chips have always been able to do bass management in the DSD domain, and starting with their DVP-NS999ES model, the can do time-alignment in the DSD domain as well.

 

[David Judah]

I think it is a matter of semantics as Michael mentioned, Sam. Some have even suggested that DSD, even before DSD Wide, is basically 1 bit PCM similiar to the 1 bit bitstream PCM systems of the past. In fact, Sony even used the same delta sigma modulators, initially.

As for Sony's chips, I've always read that the signal has to be converted to multibit to do any kind of DSP such as time alignment and BM. Is that not true anymore?

I suppose it depends on who you are talking to(IOW, who supports whichever format).

DJ

 

[Peter Ping]

DJ, if that is the case, we are beating a dead horse.

 

[John Kotches]

Sam,

So long as you provide the documentation without requiring an NDA. If it does indeed exist, why not present this at an AES convention?

The fact that the solution is DSD in and DSD out says zero about the internals of the ASIC.

It'll be interesting to see what happens at the March AES.

Cheers,

 

[John Kotches]

David,

I don't recall who originally said it, but the expression "highly degenerate PCM" has been used to describe DSD on more than one occasion.

There was some very interesting discussion about DSD, DSD-Wide and PCM a few months back on the Surround Sound mailing list....

Regards,

 

[David Judah]

Yes, I can understand why someone might say that, John. It's interesting that Sony/Philips went the route that was originally designed to make cheaper PCM converters.

Depending on who the author is, DSD goes from offering no advantage over 24/96+ PCM to being the best sounding and most analog sounding digital system to being inherently flawed. It's no wonder there is so much confusion amongst us non-technical folk. Like many things in audio, even the experts don't agree.

Do you have a link to the discussion you mentioned?

DJ

 

[BeatCrazy]

John Kotches,

At one time, you stated that Sony was using DSD DSP to accomplish the very things you say they can't do. See your quote here:

http://db.audioasylum.com/cgi/m.pl?f...+C555&session=

As for NDA, it depends on how detailed you want to go. Here is Sony information not distributed to the public:

 

[FeisalK]

Sam, that description is in the ES series manuals (for example, the CX777ES, page 23)

 

[John Kotches]

Sam,

I did report DSD-DSP, which says nothing about the internals. I've learned a heck of a lot in the two years since that post, but I'm glad you took the time to do the research.

BTW, I notice you haven't bothered to post anything about bass management, just some information about the buffering circuitry

Cheers,

 

[BeatCrazy]

FeisalK,


Bold highlights are my own.

 

[John Kotches]

Sam,

This is nice...



All this says is that the LSI does bass management. It says zero about how that bass management is accomplished. LSIs are very good about doing specific jobs, since it's hardware. Much faster than software, so doing the DSD --> DSD-Wide --> DSD work in the LSI is well within the realm of possibility.

Cheers,