Sound Galleries Music Server

[EuroDriver]

They could pack this with the DSD Dac project from Jussi...that would be Faun, I mean, fun...

 

Guess what Jussi is carrying in his hand luggage back to Finland, my DAC 8 DSD :-)

 

Guess who follows Jussi's blog, Lothar Wiemann of T+A

 

No one is resting on their laurels :-)

[EuroDriver]

Hi Eurodriver,

 

Would the ABC PCB company in Switzerland be able to do this?

 

home - Electronic, Engineering and Consulting

 

there are so many great electronics companies here in Europe !

[audio.bill]

Haven't seen this posted yet, sorry if I missed it. Sound Galleries SGM Demonstration | AudioStream

[EuroDriver]

Haven't seen this posted yet, sorry if I missed it. Sound Galleries SGM Demonstration | AudioStream

 

 

This is the quote that sums up what we were trying to do to show the capabilities of DSD512

 

This system nailed many aspects of reproduction extremely well like dynamics, nuance, detail, clarity2, and more.

 

I like the Clarity Squared comment which is consistent with Matthias's comment about tracks sounding like Reel to Real tape

[Distinctive]

If a reasonably priced PCM or DSD DAC with I2S input emerges that takes the sound quality crown away from the DAC 8 DSD, then it would not take us long to put in one of the best performing I2S input cards.

 

Our good friends and OCXO clock board supplier, Pink Faun, are on the leading edge of I2S. The latest Pink Faun I2S board has an OCXO as well, a spin off from the work that Pink Faun did for us :-)

Thanks.

At least then it will be PS Audio DirectStream compatible.

The 'reasonable priced' DS DAC has been out for ages....

[tsaett]

Being back from Munich I would like to post my findings as well.

 

First a big thanks to the organizers, Geoffrey, Edward and Emile for your great hospitality. It sure must have been a lot of work for you but it paid off.

The sound was among my top 4 in Munich, which were the Living Voice room, Gryphon Audio, Zellaton and Yours. Considering the prices of the others, that's quite an achievement in my books.

Particulary interesting for me was the technical session after dinner, where we compared Redbook PCM with DSD128 and 512, each step was clearly audible with 512 being the clear winner. Especially the background noise of the Miles Davis track was far less distracting with 512 than with the others. I still would rate the shift from PCM to 128 more worthwhile, but diminishing returns have to kick in somewhere. Being upsampling to 128 for 2 years now, I am willing to try 512 myself in the future.

For me personally the SGM is out of my financial league, but I definitely rate it among the top digital sources available today.

 

It was great fun to meet some forum members and journalists, this sure was an illustrious crowd.

I even had the the chance to briefly chat with Chris Connacker who could answer some of my questions regarding my CAPS Pipeline and DSD512 capability.

Finally it was nice to meet Lukasz of Lampizator and Jussi Laako of HQPlayer, nice guys indeed.

 

One note to Matt: Greetings to your wife Claudia, we had a nice conversation but somehow we didn't come to talk, save that for the next time. Grüße nach Freiburg.

 

Thanks again SGM team

Christoph & Tanja (from Cologne)

[EuroDriver]

The sound was among my top 4 in Munich, which were the Living Voice room, Gryphon Audio, Zellaton and Yours. Considering the prices of the others, that's quite an achievement in my books

 

This is the music server thread, and this is Computer Audiophile, but thank you Christoph for bringing up the fact that the total system electronics cost was less than Euro 23,000 in Europe including VAT and US $ 24,000 in the USA.

 

 

The amazing sounding Taiko Kodo amps list for Euro 4,000 including VAT and US $ 4,000. They are probably the most expensive NCore 500 based amps, but in a completely different sound quality class IMO. The devil is in the implementation details and Taiko has put in the best. The design of the input stage with an input impedance so high that any DAC with volume control will sound great driving it directly. A frequency response of the input stage way up high, the way Goldmund and other top flight High End electronics companies do it. The quality of the internal wiring and hardware is the best of the best. This engineered by ear approach is expensive, but boy does it deliver results !

 

Who would have thought a year ago that Sound Galleries of Monaco and newcomer Taiko Audio, manufacturing music server and amps in Holland, with the help of the German T+A DAC 8 DSD, could achieve this level of performance at a fraction of the cost of the big boys.

 

As Geoffrey Armstrong said in his opening remarks, the catalyst for this co-operation which has delivered this price break through system, is none other than Computer Audiophile.

 

 

 

 

BTW Lawrence Dickie was with us during the afternoon, listening intently with eyes closed catching as much of the music before he had to leave for his next appointment. The German distributor for Vivid told us the next day that they had to drag Lawrence away, he just did not want to leave :-)

[Elberoth]

I'm finally back at home, after having spent 3 full days at this massive show.

 

First of all, big thanks to Geoffrey, Edward and Emile for pulling off such a great and well attended event.

 

I've been a big proponent of DSD upsampling for the past 2 years (ever since I bought my first DSD only DAC from Lampi) so the sound quality you got should not come as surprise, but it was. It was easily my #2 sound of the show, after the Magico M-Project/Soulution demo (on Saturday only, when they were using the 701 monos). I didn't found it lacking in color saturation department - and that comes from Lampizator Golden Gate user.

 

What was astonishing to me, was that you managed to get the sound of this callibre from a relatively inexpensive set of electronics - $3000 DAC and $5000 amps ! Many exhibitors at the show, could not get sound nowhere near as good having $200k-$300k worth of electronics at their disposal. I'm pretty sure your Sound Galleries server, with seamless Roon/Tidal/HQ Player integration and DSD 512 upsampling, played a great role here.

 

Wow, just WOW.

[Priaptor]

I'm finally back at home, after having spent 3 full days at this massive show.

 

First of all, big thanks to Geoffrey, Edward and Emile for pulling off such a great and well attended event.

 

I've been a big proponent of DSD upsampling for the past 2 years (ever since I bought my first DSD only DAC from Lampi) so the sound quality you got should not come as surprise, but it was. It was easily my #2 sound of the show, after the Magico M-Project/Soulution demo (on Saturday only, when they were using the 701 monos). I didn't found it lacking in color saturation department - and that comes from Lampizator Golden Gate user.

 

What was astonishing to me, was that you managed to get the sound of this callibre from a relatively inexpensive set of electronics - $3000 DAC and $5000 amps ! Many exhibitors at the show, could not get sound nowhere near as good having $200k-$300k worth of electronics at their disposal. I'm pretty sure your Sound Galleries server, with seamless Roon/Tidal/HQ Player integration and DSD 512 upsampling, played a great role here.

 

Wow, just WOW.

 

While I can't comment on the server I have been playing around with the T&A up sampled and am very impressed and more so when price is taken into account.

[EuroDriver]

Hi Adam,

 

It was easily my #2 sound of the show, after the Magico M-Project/Soulution demo (on Saturday only, when they were using the 701 monos).

 

I did not realize that the 701 monoblocs list for $ 155,000. I think Emile / Taiko Audio has nothing to be embarrassed about being bettered by monoblocks costing 38 times more LoL

 

I didn't found it lacking in color saturation department - and that comes from Lampizator Golden Gate user

 

Many thanks for this observation and comment. We spent an awful lot of time voicing the system on Thursday, and thought we got it pretty much right, it makes us very happy that you thought so too

[EuroDriver]

While I can't comment on the server I have been playing around with the T&A up sampled and am very impressed and more so when price is taken into account.

 

I know that you have had in house the best of the best in DAC's so I take your comment as a significant data point.

 

We very much look forward to you having a chance to hear the DAC 8 DSD being fed by the SGM in the not too distant future

[hifial]

I didn't found it lacking in color saturation department -

 

Wow, just WOW.

 

Ed, I think this is very telling!

[marslo]

I would like to join my congratulations to Geoffrey and his team for the achievement. We visited the Sound Gallery together with my wife last year and spoke to Geoffrey during very kind and pleasing presentation of Duo Omega driven by Audiopax and Lampi DAC fed from HQ player installed on Mac mini.

We talked about the server and then I followed the threads here and on WBF.

I cannot comment the price tag but I relay on Adam' s opinion about the sound quality.

We bought our lampi dsd only dacs from Lukasz in the same time, probably being one of first customers in Poland.

I am still very happy with my Lampi Big 7 SE with Duelund caps , R2R engine and DSD 128 board which is close to my analog rig .

This year I could not be in Munich but I am very keen to compare SG server to my W20 one day.

[matthias]

 

One note to Matt: Greetings to your wife Claudia, we had a nice conversation but somehow we didn't come to talk, save that for the next time. Grüße nach Freiburg.

 

Thanks again SGM team

Christoph & Tanja (from Cologne)

 

Christoph,

 

I am sorry about that, yes we save that for the next time in Cologne or for the DSD1024 May Fest next year(???).

 

It was a pleasure to meet a great family of DSD lovers.

 

At the moment I am listening to Glenn Gould, Remastered, the complete Columbia Album Collection, 81 CDs, all DSD mastered from the analogue tapes. Terrific sound quality, waiting to be upsampled to DSD512 sometime.

 

Thanks again for this great event and

 

greeting to you and Tanja

 

Matthias and Claudia

[hifial]

FYI

 

Mono and Stereo High-End Audio Magazine: Sound Galleries SGM 2015 music server exclusive presentation at BMW Welt

[matthias]

Chris comments:

 

"I sat through a lengthy demonstration of the Sound Galleries Music Server at BMW World Friday evening. The server was resampling all audio to DSD512 for output to a T+A DAC8 DSD. I was very impressed by what I heard from this system. So much so, that I must have spent an hour talking about it with CA reader joelha when I ran into him in the lobby of the hotel at which we were both staying. I kept thinking to myself, how can we get something from nothing (Redbook CD at 44.1 -> Octuple-rate DSD at 22.5792 MHz). Fortunately I spent some time with another well respected engineer in the industry on Saturday and asked him about the whole concept and what I heard the previous night. He said it's a very valid concept, but should be looked at a bit differently than simply resampling to DSD. The DSD really isn't what makes the difference. He said it's more filtering outside the DAC and the fact that one can bypass much of the chip in the DAC that makes the difference. I'm sure I butchered his explanation, but I did my best. It's one of those things that I understand better than I can explain :~)

 

I plan to get the T+A DAC 8 DSD in for review and to spend time with it resampling audio to DSD512. I want to really spend time testing what many members of the CA Community have been saying for years, that resampling to DSD 128/256/512 can be a huge sonic benefit. I can't wait to dig in, much more than my previous cursory listening sessions at home."

 

Matt

[wisnon]

Yrs, it appears to me that the real magic is not the T+A DAC, WHICH NO DOUBT IS VERY GOOD for the money and pushes above its weight, but rather the SG transport and the advanced HQP convoy engine.

 

I suspect that the Dac8 all be a lot less impressive with an i7 laptop up sampling with Jriver to dsd512, for example and that any pure DSD Dac will get a commensurate boost from the SGM/hqp front end feeder.

 

With a very quiet and refined Transport..all you need is the simplicity of an elegant DSD playback engine and boom..there come the magic.

[EuroDriver]

Yrs, it appears to me that the real magic is not the T+A DAC, WHICH NO DOUBT IS VERY GOOD for the money and pushes above its weight, but rather the SG transport and the advanced HQP convoy engine.

 

I suspect that the Dac8 all be a lot less impressive with an i7 laptop up sampling with Jriver to dsd512, for example and that any pure DSD Dac will get a commensurate boost from the SGM/hqp front end feeder.

 

With a very quiet and refined Transport..all you need is the simplicity of an elegant DSD playback engine and boom..there come the magic.

 

Hi Norman,

 

Your post above is very much in line with our thoughts. We are eagerly waiting to get our hands on one of Lucasz's new DSD512 DAC's

 

Two observations,

- the implementation of the 1-bit converter in the T+A DAC 8 DSD is Lothar's second effort, the first being the PDP-3000HV which was finalized in its engineering about 2 years ago. The sound quality advances made in the DAC 8 DSD's 1-bit converter are substantial

 

- there is quite a bit of very clever design in the DAC 8 DSD 1-bit converter both to get a very tight control over the jitter which is even more difficult at DSD512 than at DSD 256, and to minimize voltage / current generated noise

 

There are several people we know of who are working on DSD512, very exciting for us as the steps that we took to reduce noise, and improve timing are magnified at DSD 512

 

An interesting takeaway. In Monaco we had a SGM prototype which was built in October 2015. After Munich we brought down 2 production models and hooked them up to the system which was SGM > Corning optical USB > DAC 8 DSD > Audiopax Pre > Audiopax tube mono blocs > Avantgarde Duo + Trio bass horns. A monster system where you hear absolutely everything, no hiding place !

 

We were expecting a nice sound quality improvement, but Geoffrey and I were absolutely floored by the difference. And this is with the Corning optical USB in the chain !

 

What this suggests to us is that RFI control when doing intensive computing, has even bigger on effect sound quality than even we had anticipated.

[EuroDriver]

There was a question in the T+A thread about mobo clocks, standard vs TCXO and OCXO.

 

We have done a lot of testing of clocks both on Haswell 4970 mobo's and Skylake i7 6700K with clocks from PPA, Russia, and the US

 

With the transformer coupled PPA OCXO, we definitely had a consistent improvement but we had mobo stability problems, and we looked for alternative approaches. In July last year we met up with Pink Faun, and they decided to engineer a PCB, which could easily mount different clock modules easily, and have good on PCB power regulation / stabilization.

 

Going into to this part of the project, we did not know the relative importance of low frequency phase noise, frequency stability, and the influence of how good the power supply is. Well, after testing 7 different clocks from 4 suppliers, we are confident that all three factors come into play.

 

About 2 months ago we were comparing a Octo core Haswell processor with no OCXO against the quad core Skylake with OCXO. Absolutely no completion, the Octo Haswell was in a different sonic space, and no way it was ever going to catch up with the SGM.

 

The high frequency power supply that we use for the OCXO PCB comes from Sweden, and is custom manufactured for us

[lmitche]

Hi Norman,

 

Your post above is very much in line with our thoughts. We are eagerly waiting to get our hands on one of Lucasz's new DSD512 DAC's

 

Two observations,

- the implementation of the 1-bit converter in the T+A DAC 8 DSD is Lothar's second effort, the first being the PDP-3000HV which was finalized in its engineering about 2 years ago. The sound quality advances made in the DAC 8 DSD's 1-bit converter are substantial

 

- there is quite a bit of very clever design in the DAC 8 DSD 1-bit converter both to get a very tight control over the jitter which is even more difficult at DSD512 than at DSD 256, and to minimize voltage / current generated noise

 

There are several people we know of who are working on DSD512, very exciting for us as the steps that we took to reduce noise, and improve timing are magnified at DSD 512

 

An interesting takeaway. In Monaco we had a SGM prototype which was built in October 2015. After Munich we brought down 2 production models and hooked them up to the system which was SGM > Corning optical USB > DAC 8 DSD > Audiopax Pre > Audiopax tube mono blocs > Avantgarde Duo + Trio bass horns. A monster system where you hear absolutely everything, no hiding place !

 

We were expecting a nice sound quality improvement, but Geoffrey and I were absolutely floored by the difference. And this is with the Corning optical USB in the chain !

 

What this suggests to us is that RFI control when doing intensive computing, has even bigger on effect sound quality than even we had anticipated.

Hi Ed,

 

Just to be clear, are you suggesting that the Corning cable added value to the SQ here? What motivated you to try the Corning in the first place? I thought we had moved pass the Corning.

 

Larry

[EuroDriver]

Hi Ed,

 

Just to be clear, are you suggesting that the Corning cable added value to the SQ here? What motivated you to try the Corning in the first place? I thought we had moved pass the Corning.

 

Larry

 

Hi Larry,

 

We have gone on from the Corning cable in our development and evolution of the SGM, but we still use it in Monaco for convenience because it is so long. The reason why we dropped the Corning optical cable were lack of reliability, hand shaking problems, and the noise ignature of the Optical PHY

 

In the back of my mind I have been a bit concerned that we were ditching a cost effective but unreliable solution.

 

What this A/B comparison between October 2015 protype and the production SGM shows is that the RF isolation which the Corning cable delivers is a lot less than what is needed between DAC and PC, and in a SGM playback system, it does not bring much if anything audible to the table. It could well be noticeably negative due to PHY noise

[lmitche]

OK, we are in the same place regarding the Corning Cable. It is of course convenient for long runs, so I anticipated that was your motivation.

 

Thanks for the explanation.

[EdmontonCanuck]

I've been following this thread with interest, however I admit to being confused about a basic point, so please forgive my nooby question. I totally understand the benefits of performing the lion's share of digital processing on a powerful compute platform, and sending the final digital signal to the DAC which, hopefully, does the least amount of processing required to generate the final analog signal. I also understand that any generated RF noise leaking through the interface between the compute platform and DAC can impair SQ, and the more powerful the digital compute platform the greater RF noise is likely to be generated.

 

My nooby question is this: could a person not implement a fairly powerful digital compute platform that is very noisy but because the interface between it and the DAC is very well isolated (say, via a lightweight ethernet NAA device) and still get equal SQ? I'm coming from the point of view of cost here, ultimately. I don't really care if my PC is sitting in a different room from my listening room, and it's doing all the heavy lifting and sending a digital signal to an NAA device that has a pretty good isolated interface to my DAC. I could build a really powerful PC for quite a bit less than what the SGM costs.

 

The SGM looks like a mighty fine piece of kit that looks and sounds fantastic and is able to sit in my audio rack, but it's way out of my price league, and I don't mind at all if the PC sits down in my basement out of the way (both looks-wise and fan-noise-wise).

 

Or are you saying that it's really impossible to get a well-isolated interface to the DAC without spending gobs of cash?

[EuroDriver]

I've been following this thread with interest, however I admit to being confused about a basic point, so please forgive my nooby question. I totally understand the benefits of performing the lion's share of digital processing on a powerful compute platform, and sending the final digital signal to the DAC which, hopefully, does the least amount of processing required to generate the final analog signal. I also understand that any generated RF noise leaking through the interface between the compute platform and DAC can impair SQ, and the more powerful the digital compute platform the greater RF noise is likely to be generated.

 

My nooby question is this: could a person not implement a fairly powerful digital compute platform that is very noisy but because the interface between it and the DAC is very well isolated (say, via a lightweight ethernet NAA device) and still get equal SQ? I'm coming from the point of view of cost here, ultimately. I don't really care if my PC is sitting in a different room from my listening room, and it's doing all the heavy lifting and sending a digital signal to an NAA device that has a pretty good isolated interface to my DAC. I could build a really powerful PC for quite a bit less than what the SGM costs.

 

The SGM looks like a mighty fine piece of kit that looks and sounds fantastic and is able to sit in my audio rack, but it's way out of my price league, and I don't mind at all if the PC sits down in my basement out of the way (both looks-wise and fan-noise-wise).

 

Or are you saying that it's really impossible to get a well-isolated interface to the DAC without spending gobs of cash?

 

10 months ago, we were getting very good results with NAA, but when we moved from a MOSFET power supply to a dual choke power supply for the motherboard, the performance of a NAA arrangement was no longer better than a one box solution, but actually inferior to a one box solution.

 

Our take is that with the lower noise floor which we achieve with the dual choke supply, the noise that comes with passing data over Ethernet becomes audible. Ethernet is a well isolated solution, but it does come with its data overhead, and the Ethernet receiver is not noise free either

[tranz]

[emoji651]️ This is big news. Up until now NAA was touted as the way to separate the noisy computing environment from the delicate DAC side.