Advice on a new 256 or 512 DSD DAC

[wisnon]

I’m looking to purchase a new DAC as the older Lite Audio DAC-83 has developed an issue and return / repair to China isn’t, I believe, merited. My thinking is to jump into the world of DSD – using HQ Player to up-sample PCM to 256 / 512 DSD. The set-up will involve a dedicated Roon / HQ Player server specified to support the conversion with a Windows-based NAA that will connect to the DAC via USB or, if supported, i2S. (Might re-image NAA as Linux if I’m bold.)

 

So, with advice from the HQ Player thread, I’ve identified the T+A DAC 8 DSD as an option for true discrete 1-bit chip-less conversion which is at the upper limit of the budget. The other options look to be Mytek Brookyn , exaSound E22 and Gustard X20U all of which seem to use SABRE DAC, along with upcoming iFi DSD DAC Pro. (I’m ruling out the more expensive options of Lampizator, Playback Designs etc.)

 

.

A DSD Euforia with upgrade to 256 should be about $2,500 or so...

 

Whatever you do, do discrete chipless architecture for the best SQ result.

[wisnon]

The last sentence needs a lot of clarification, it annoys me frankly. As far as DACs go, there's no best implementation of chip or chipless as yet to obtain the best SQ, if that could ever be defined to begin with. I'm not asking for evidence, or supported documentaries, the market is mature enough to cater for all tastes and at all different prices without one technology or method dominating all.

 

Agree with hifal to choose other than the Gustard if the T+A is a preferred choice.

Use your ears and it will become immediately apparent.

 

Chipless discreet fuxxes up the signal much much less than Dac chips.

 

For DSD, you need to get out of the way and let the HQ P Server processed goodness flow thru with the minimum of filtering.

[wisnon]

Has anyone compared the T+A Dac 8 against Rossi's LIO...slightly different beasts for sure, but am curious about the DAC section of the LIO against the DAC 8.

 

The Dac 8, like the Lampi, is 2 Dacs in one box. The PCM section is different from the DSD digital section.

 

If someone is asking for a 256 or 512 capable DSD Dac, it is HIGHLY likely that they want to employ Jussi's luscious upsampling algorithm to mate with a discrete chipless implementation. Going for a Dac chipped DSD solution IMHO would thus be folly.

[wisnon]

As I remember it EuroDriver compared the T+A DAC 8 DSD to a Lampi Big 7 and the T+A beat it doing DSD512 vs the Lampi doing...and this is were I can not remember...if it was DSD128 or DSD256. But I do know it was the highest DSD rate that the Big 7 could do.

 

He never did. He had a Big 7 in May 2015 (doing DSD128) and tested the T+A in 2016 doing DSD512. There was no T+A Dac 8 existent in May 2015 and the latest Lampis of 2016 do 256 and 512. There was no SG Server in May 2015 either...

[wisnon]

I never said it was a direct comparison, (and I said it may have been DSD128), Never said he used the "SG Server" in either case. I believe EuroDrive was still doing a lot of server testing when he first heard the T+A. Actually there was a T+A DAC 8 before that time period but that one did not do DSD, hence the T+A DAC 8 DSD (lol).

 

Never said that there are no Lampi that can do a higher rate. Just that at one time that the Big 7, at what it was capable of doing at that time, was heard and the same person also heard the T+A DAC 8 DSD and of the opinion that the T+A DAC 8 DSD was better.

 

Look, I know it involves Lampi and its like catnip for you (lol), it is just one persons opinion.

 

It is great to hear that Lampi can play DSD 512! I would assume that not all models do. Can you please let us know which ones do DSD51. Besides myself I am sure there are a lot of Audiophiles that would want to know.

I speak to Ed quite a lot and know exactly what is going on. I am not engaged in guesswork. No wild speculation here. If the SG Server is a big advancement as a transport and it has really only been tested on 1 or 2 Dacs and never a Lampi, I believe your assertion is disingenouus. Fact is SG Server and Lampi have never been paired and until that happens...neither you nor I know what the outcome will be.

 

If you dont know what Lampi offers, I suggest you contact the man himself, the US distributors or read the website.

[wisnon]

Always interesting to hear that kind of feedback for timbral accuracy and transient reproduction from people who actually play.

 

There's a new USB card from dyinhk which apparently can do DSD512 and I2S. I don't know if anyone has tried it with the Gustard yet.

 

I think the Gustard and the possible tweaks to it make it one of the best value propositions for high-rate DSD currently.

 

The "level 1" tweaks by Ric which were implemented by Quadman himself on his own are rather easy to do if you're used to DIY. Some additional ones may be a little more intricate but I wouldn't classify either as "a lot of major tweaking" as hifial seems to imply.

 

For more info, including a lot of good end-user DIY feedback by Quadman, look for the relevant Gustard Tweaks thread over at WBF. It's quite cool he shares his experience doing the mods and his impressions of the results. Quite cool of Ric to share his info as well.

Isnt the Gustard a Sabre Dac?

 

If so, then who knows what they do to the DSD signal fed to it.

 

I have not heard it and so cant comment definitively, but IMHO, that would disqualify it as a "primo" Dac for high rate DSD playback. High rate DSD is not a sweepspot for a Sabre chip Dac.

[wisnon]

With all due respect if you have not heard it then you should not comment on it at all. Look...I am not taking anything away from the Lampizator GG or the T+A DAC8. I will not comment on those as I have not heard them. However, I HAVE heard the Gustard X20U and to be perfectly frank, I could give a rat's behind what chip is under the hood. After 35 years of playing in the Academy of Music in Philadelphia, Carnegie Hall in New York, Severance Hall in Cleveland and other venues, when I hear the accurate yet beautiful tone and timbre of acoustical instruments, I really don't care if I pop the hood to see two rubber bands and some popsicle sticks.

 

I am waiting with bated breath to see how quadman's new DSD512 card performs.

 

BTW, I also have heard the Fore Audio DaisY1 at Ted_b's place and it sounds very very good indeed!

 

My apologies to the original poster.

 

Regards,

 

Randy

 

Randy, I speak of the philosophy/technology of what goes on inside the Sabre chip with DSD. There is a reason why DSD is handled with kids gloves and a chip is primarily for PCM with DSD as a workaround. A saber chip would not be my first choice for dedicated DSD playback...though it does not mean that it can't sound good....as it clearly can. Just don't think it can be the best it can be. I have heard plenty of chipped DSD and FPGA DSD and none please me like chipless. If you are not DSD crazy like me, its not really a big deal.

 

I wait to see how the new saber chips treat DSD, the 9018 and 9038. They have very different tech again.

[wisnon]

Randy

 

I speak of the INNER workings of the saber chip wrt DSD. Its not a knock on the Gustard, which may be as well implemented as any Sabre Dac out there, rather the limitations of the chip itself. I am pretty sure DSD is not processed natively inside the ESS chip in 1 bit format. They do some kind of state variable manipulation AFAIK and the rest is a black box.

 

So yes, it can sound good, but for someone like the OP chasing SoTA DSD, it would not be a natural first choice.That and that alone was my point. I hope I made myself clear now.

[wisnon]

SoTA DSD? He is asking for a good sounding DSD256 and/or 512 DAC, and finds the T+A slightly outside his budget. So we began discussing compromises, as that is what you do when discussing within someone's budget (i.e lower price than T+A). If the guy can afford the T+A this thread should be closed, as we all know that end of the budget is clearly T+A. ?? What am I missing here?

Dépends on one's interpretation:

"I’m looking to purchase a new DAC as the older Lite Audio DAC-83 has developed an issue and return / repair to China isn’t, I believe, merited. My thinking is to jump into the world of DSD – using HQ Player to up-sample PCM to 256 / 512 DSD. The set-up will involve a dedicated Roon / HQ Player server specified to support the conversion with a Windows-based NAA that will connect to the DAC via USB or, if supported, i2S. (Might re-image NAA as Linux if I’m bold.)

 

So, with advice from the HQ Player thread, I’ve identified the T+A DAC 8 DSD as an option for true discrete 1-bit chip-less conversion which is at the upper limit of the budget."

 

If you really want to fully experiment with DSD at the limits, to my mind it must be a 1 bit solution...

Just having DSD playback is not the same as experiencing the best that DSD has to offer.