BRYSTON and DSD

0 Members and 1 Guest are viewing this topic. Read 6193 times.

James Tanner

  • Facilitator
  • Posts: 20477
  • The Demo is Everything!
    • http://www.bryston.com
BRYSTON and DSD
« on: 12 Jan 2014, 01:55 pm »
DSD

Hi folks

I have been asked a number of times by our customers our position on DSD so please find below our thoughts.
 
DSD currently is available in a number of variations. Assuming you want to produce a product that can grow with time and market forces both hardware and software should to be considered.

DOP 64, 128, 256, 512 (last 2 are theoretical at this point) and DSD Native are the current options.

Currently DSD is possible relatively easily with current software and hardware utilizing DOP 64 on current DAC’s and to some degree DOP 128

For example:
DOP-64 requires 176.4 capable DAC’s
DOP-128 requires 352.8 capable DAC’s
DSD Native 64 x 44.1 (standard 2.8224 MHz) (which is equivalent to 16 BIT 176.4 PCM) DAC’s.

Currently the majority of DSD capable products today are using DOP-64 architecture and some DOP-128 and DSD Native. There are also some manufacturers that have proprietary hardware and software which are capable of DOP-64, DOP-128 and Native DSD but proprietary software is not compatible with USB 2 drivers etc. .

The other issue we are investigating has to do with the sample rate converters in DAC’s.  In order to activate DSD the sample rate converter on the DAC inut stage would have to be a custom piece because current sample rate converters are not equipped to handle DSD to the best of our knowledge?

So that means in a standard PCM converter like our BDA-2 you would have to bypass the sample rate converter which may cause more jitter because our current sample converters reduce jitter on the input.  So in Bryston’s case we definitely want the sample rate converter in the circuit for 44.1 to 192 PCM signals. Also some DAC’s we have looked at that are DSD capable convert all the incoming sample rates (44,48,88,96,176,192Hz) to a very high single sample rate ‘Asynchronously’ whereas our preference with our DAC’s is to maintain the ‘Native’ incoming sample rate throughout the conversion process with all PCM signals.  If we do up-sample we do it in a synchronous manner so (44.1 becomes 176.4 and 48 becomes 192) not asynchronous.

So before Bryston goes down the DSD road I want to make sure we are providing our customers with the full story and allowing for possible advancements in both hardware and software development as much as possible. Also please offer any info or input if you feel we are mistaken or misinformed.

james tanner


Note- DOP – Stands for DSD over PCM

« Last Edit: 12 Jan 2014, 06:35 pm by James Tanner »

R. Daneel

  • Full Member
  • Posts: 1088
Re: BRYSTON and DSD
« Reply #1 on: 12 Jan 2014, 07:56 pm »
Hi James!

This is highly interesting!

I think DSD is something of a sales tactics more than anything alse for manufacturers at this point. Considering how much DSD material there is out there, that certainly seems to be the case. Change of situation where music would be mastered for DSD instead of PCM is unlikely and as long as this is the case, DSD cannot outweigh PCM in terms of sales. Also, SACDs often don't sound nearly as good as they should and that is the consequence of industry's own unpreparedness. DSD files are smaller than equivalent PCM files so that might be an advantage in this day any age where everything is downloaded from the internet. Still, internet bandwidth and disk space are no longer a problem so even that advantage is no longer crucial. To conclude, having a DSD-capable DAC is therefore more of a compatibility thing than anything else - making sure the customer is able to play the file.

Based on that conclusion, your BDP range is a powerful one. It is a computer and I have no doubt it is capable of processing DSD in such a way where the signal output through it's S/PDIF interface would be PCM. That would seem to be the most effective way of preserving compatibility.

Now, from everything I have gathered from Chris this last couple of months, DSD is possible only via USB connection since Sony/Philips protocol doesn't support it. In that case, DSD output through the BDP's USB interface is processed by the main CPU and software decoders of your BDP. Computers are a tricky business so hardware decoding is usually more reliable than software decoding. In other words, your Envy24-equipped sound card is possibly more reliable than using software decoders powered by your BDP's main CPU.

This means the following - you should not exclude the possibility of having DSD processed to PCM and then output via standard S/PDIF interface in favour of having direct DSD output through the USB output. DAC still has to process that digital signal and I am not convinced that USB input sounds better than a S/PDIF input. Not even on some ridiculously expensive DACs.

So you would have to weight everything and decide according to criteria of reliability and audio quality.

If you would consider converting DSD to PCM on the BDP first, then one of the advantages is certainly compatibility for the days to come. Since BDP is essentially an open architecture system which functionality is dependent on software, perhaps it would be less costly to do so than invest time in research into new D/A hardware when it becomes available or design modular DAC systems which by definition, are never as trouble-free or as affordable as manufacturers are promising. This approach would also leave more signal processing on BDP's shoulders rather than having to have your BDA process USB stream.

At least, these are my thoughts.

Cheers!
Antun

Mag

Re: BRYSTON and DSD
« Reply #2 on: 12 Jan 2014, 08:45 pm »
I don't think I'm up to speed or understand the issue. But what I want from DSD assuming in becomes a more viable format is to maintain the multi-channel input/outputs. IMO the multi-channel format is the reason for sacd. And I've heard nothing on my system that sounds as good as multi-channel sacd.

So then the 5.1 outputs would go to a dac that can upsample the dsd then send it out to the 5.1 bypass on the pre-pro. Otherwise you lose the multi-channel capability of sacd by going usb, if I understand correctly.

Personally I'm not interested in 2 channel sacd. :smoke:

wisnon

Re: BRYSTON and DSD
« Reply #3 on: 13 Jan 2014, 10:41 pm »
DSD does not require a Dac chip…simply a low pass filter.

Mag

Re: BRYSTON and DSD
« Reply #4 on: 14 Jan 2014, 12:30 am »
I gave my sacd's a spin yesterday and as good as it sounds. I still can detect some distortion on playback. So it needs to be cleaned up to be top notch whether it be a filter or dac correction. :smoke:

R. Daneel

  • Full Member
  • Posts: 1088
Re: BRYSTON and DSD
« Reply #5 on: 19 Jan 2014, 02:08 pm »
Hi James!

I just read some information from Berkeley audio design about their upcoming D/A converter. What is interesting is the way they implemented DSD support. The DAC itself  features all the standard digital inputs but does not support DSD itself. Instead, the user is provided with a PC-based application that converts DSD to PCM and the signal is then sent to the DAC as PCM. According to Berkeley, PCs and computers in general are better suited for accurate and reliable conversion than DACs due to superior processing power and it is generally a good idea to distance the DAC circuitry from the rest. They also mention the very thing I mentioned in my first post on this page - the importance of supporting future DSD versions that might appear along the way. PCs are dependent on software so any future format can be supported just by upgrading the application responsible for DSD-to-PCM conversion. Berkeley mentions there is no loss of audio quality with DSD-to-PCM conversion.

I think this might be the way for Bryston as well. However, there is one distinct difference and an advantage on your side. You already have a highly advanced computer system in the form of BDP-1/2 digital player and so the user would not be dependent on a classic computer. You could design an application for the BDP that would convert any DSD stream to PCM and then outputs it via the standard digital outputs on the back. By doing this, you would ensure future compatibility and also backwards compatibility with DACs that have no USB input and/or DSD support.

I think this would add value to your current line of products. Since they are already well received and regarded, I don't see how it could be a bad thing.

That, however, does not mean you should abandon the premise of having the BDA-2 converting DSD on it's own just for those that do not have one of your wonderful BDP players.

How does this sound to you?

Cheers!
Antun

James Tanner

  • Facilitator
  • Posts: 20477
  • The Demo is Everything!
    • http://www.bryston.com
Re: BRYSTON and DSD
« Reply #6 on: 19 Jan 2014, 04:17 pm »
Hi James!

I just read some information from Berkeley audio design about their upcoming D/A converter. What is interesting is the way they implemented DSD support. The DAC itself  features all the standard digital inputs but does not support DSD itself. Instead, the user is provided with a PC-based application that converts DSD to PCM and the signal is then sent to the DAC as PCM. According to Berkeley, PCs and computers in general are better suited for accurate and reliable conversion than DACs due to superior processing power and it is generally a good idea to distance the DAC circuitry from the rest. They also mention the very thing I mentioned in my first post on this page - the importance of supporting future DSD versions that might appear along the way. PCs are dependent on software so any future format can be supported just by upgrading the application responsible for DSD-to-PCM conversion. Berkeley mentions there is no loss of audio quality with DSD-to-PCM conversion.

I think this might be the way for Bryston as well. However, there is one distinct difference and an advantage on your side. You already have a highly advanced computer system in the form of BDP-1/2 digital player and so the user would not be dependent on a classic computer. You could design an application for the BDP that would convert any DSD stream to PCM and then outputs it via the standard digital outputs on the back. By doing this, you would ensure future compatibility and also backwards compatibility with DACs that have no USB input and/or DSD support.

I think this would add value to your current line of products. Since they are already well received and regarded, I don't see how it could be a bad thing.

That, however, does not mean you should abandon the premise of having the BDA-2 converting DSD on it's own just for those that do not have one of your wonderful BDP players.

How does this sound to you?

Cheers!
Antun

Hi

Yes we have looked at that option as well and was part of my concern that changing the BDA-2 to handle DSD would compromise PCM quality.

james

R. Daneel

  • Full Member
  • Posts: 1088
Re: BRYSTON and DSD
« Reply #7 on: 19 Jan 2014, 05:56 pm »
Hi

Yes we have looked at that option as well and was part of my concern that changing the BDA-2 to handle DSD would compromise PCM quality.

james

Hi James!

Yes, I read your opening post.

You could, of course, process PCM and DSD with separate circuits, one having nothing to do with the other, but I think perhaps that would increase the cost substantially.

DSD recordings are rare, unless you are ripping SACDs (which is impractical to say the least), and it seems that designing a proper DSD/PCM circuit is a costly procedure and the outcome a bit uncertain. Sorting it out in the BDP software so that it does the conversion could be cheaper.

Sorry! I just keep talking!!  Canadians must be very patient people!

Cheers!
Antun

R. Daneel

  • Full Member
  • Posts: 1088
Re: BRYSTON and DSD
« Reply #8 on: 12 Mar 2014, 05:16 pm »
Hi

Yes we have looked at that option as well and was part of my concern that changing the BDA-2 to handle DSD would compromise PCM quality.

james

Hi James!

Are there any updates you can give us on the DSD support matter?

Your concern is a valid one and personally, I would not want to have DSD compatibiliy at the cost of PCM audio quality.

My concern is the USB interface. I am not convinced USB input is better than AES input. It is what it amounts to in the end - if you're going to have DSD support, then it will inevitably be done through USB. So my question to you is this:

given the fact AES generally offers better audio quality than USB with PCM material, would it not be prudent to see what works best:
a) running DSD signal directly into the  DAC via USB and doing the D/A right in the DAC or
b) having the DSD converted to PCM externally (in the BDP perpahs) and then running the PCM signal into the DAC via AES and doing a standard D/A conversion in the DAC?

I see few benefits in this:
1) no necessary updates for your BDA DACs or any other DACs
2) support for future DSD iterations since the limiting factor would only be the conversion software (or the firmware in your BDP)
3) use of chip-based audio device like the Envy24 in your BDP and proven AES interface rather than software controllers and USB interface

Other benefits come to mind but these are the ones I would mention.

Cheers!
Antun

R. Daneel

  • Full Member
  • Posts: 1088
Re: BRYSTON and DSD
« Reply #9 on: 21 Mar 2014, 06:48 pm »
How about it James?

James Tanner

  • Facilitator
  • Posts: 20477
  • The Demo is Everything!
    • http://www.bryston.com
Re: BRYSTON and DSD
« Reply #10 on: 21 Mar 2014, 08:08 pm »
How about it James?

Hi

We are still looking at it but it keeps getting more muddy the deeper we look.  We are definitely in no hurry at this point.

james

bk12

  • Jr. Member
  • Posts: 68
Re: BRYSTON and DSD
« Reply #11 on: 23 Mar 2014, 03:07 pm »
As far as the BDP DoP DSD options are planned, will the DoP output be available on both the spdif and the USB outputs, or just the USB output?

James Tanner

  • Facilitator
  • Posts: 20477
  • The Demo is Everything!
    • http://www.bryston.com
Re: BRYSTON and DSD
« Reply #12 on: 23 Mar 2014, 03:35 pm »
As far as the BDP DoP DSD options are planned, will the DoP output be available on both the spdif and the USB outputs, or just the USB output?

Hi

As far as I know just the USB - I believe that is the only way we get legal licensing .

james

bk12

  • Jr. Member
  • Posts: 68
Re: BRYSTON and DSD
« Reply #13 on: 23 Mar 2014, 03:37 pm »
Thanks James!  The support and communication that you and Chris provide in these enthusiast forums is of incredible value to this customer.  Truly appreciated. 

R. Daneel

  • Full Member
  • Posts: 1088
Re: BRYSTON and DSD
« Reply #14 on: 24 Mar 2014, 01:19 pm »
Thanks James!  The support and communication that you and Chris provide in these enthusiast forums is of incredible value to this customer.  Truly appreciated.

I couldn't have said it better myself. It is one of the reasons why Bryston is on the top of my list of audio manufacturers.

Cheers!
Antun

R. Daneel

  • Full Member
  • Posts: 1088
Re: BRYSTON and DSD
« Reply #15 on: 24 Mar 2014, 01:29 pm »
Hi

We are still looking at it but it keeps getting more muddy the deeper we look.  We are definitely in no hurry at this point.

james

Hi James!

Well, I know you'll succeed in the end.

I used to think DSD has every advantage over PCM but now I see that is not really true since DSD files are larger and therefore take more space on the disk and take longer to download.

The fact that numerous small manufacturers are now supporting DSD doesn't worry me since it is a selling point for them. When manufacturers of professional converters universally adopt DSD, that will be the moment when DSD will be worthy of consideration I think.

I am much more interested in the future of BOT-1 but I will leave my comments in the appropriate thread.

Cheers!
Antun

werd

Re: BRYSTON and DSD
« Reply #16 on: 25 Mar 2014, 05:38 pm »
To me, it looks like they are Looking for a way to get rid of of these single bit sabre chips. (who ever makes them). They're using DSD as a hype for marketing. Not really impressed with them compared to dual-dac 24 bit.

To get a decent DSD dac its in the order of $5k and it ends up being a dual configure 24bit dac with DSD output.

R. Daneel

  • Full Member
  • Posts: 1088
Re: BRYSTON and DSD
« Reply #17 on: 16 Oct 2014, 04:05 pm »
Hi folks!

This is an old topic but I have found something interesting. Bryston's approach to D/A conversion has always been very similar to that of Berkeley audio design (founded by Pacific microsonics' engineers). This means pefecting the existing technology rather than using different forms of data manipulation like the PCM-to-PWM conversion.

According to Bryston, you cannot have DSD support in the BDA-2 because it would essentially mean bypassing the internal sampling-rate converter. SRC is an integral part of the digital engine and helps to lower jitter. Bypassing would compromise the DAC's PCM performance. So you can either have a perfect PCM decoding or have suboptimal PCM decoding with DSD capability.

Berkeley recently released the Reference series DAC and this is how they solved the problem. They didn't! There is no direct DSD support on the DAC. The machine is supplied with JRiver software which can convert DSD to PCM in real-time. This means that the DAC's hardware can be focused on one thing only, PCM decoding. It means less components and data manipulation and ultimately, higher audio quality.

Bryston already has a digital player to take over the computer's role in audio playback. It would perhaps be possible to implement a DSD-to-PCM conversion in the BDP so that DSD becomes available for users that don't have a DSD-capable DAC.

Here's the link on the Berkeley reference series:
http://www.berkeleyaudiodesign.com/reference/

What would be even more cool is if Bryston would make a DSD-only DAC. Highly unlikely considering DSD is in it's diapers and there is a chance it will disappear with much less bitterness than was the case with magnetic tape.

I just thought this might be interesting!
Antun

James Tanner

  • Facilitator
  • Posts: 20477
  • The Demo is Everything!
    • http://www.bryston.com
Re: BRYSTON and DSD
« Reply #18 on: 16 Oct 2014, 04:10 pm »
Hi Antun

I think our approach will be a DAC that has two different circuit paths - one optimized for PCM and one for DSD.

james

Rod_S

  • Full Member
  • Posts: 1070
Re: BRYSTON and DSD
« Reply #19 on: 16 Oct 2014, 08:20 pm »
I would certainly purchase that unless whatever SSP I end up with down the road has the capability to deal with DSD and it's multiple formats internally.