Roon RAAT approved DAC recommendations

I’m considering the purchase of a Roon RAAT approved DAC for my main listening room, and coupling that with Bluesound for multiple rooms. (I separately posted regarding Bluesound, and assume that any RAAT approved DAC can be grouped with Bluesound.)

I’m considering the dCS Bartok Headphone DAC or the T&A DAC 8 DSD, and welcome any comments on those DACs, along with any other recommendations that are RAAT approved. The dCS unit is at the high end of the price range I can consider, however.

One question that I have is I was told that the an AES connection is limited to DSD 64 single rate, and the ethernet connection on the Bartok is therefore an advantage for DSD 128. Is that correct?

Many thanks in advance for any advice and recommendations!

Can only recomend Pro-Ject Pre S2 DAC…
I use ROON ROCK>DAC…Upscaling everything from Tidal to DSD512 and play MQA native…for the money…sounds awesome

1 Like

I don’t have any good answer, but would like to hop on this train anyway… As my issue is related.

Livingroom setup I have is. Hifiberry DAC+ -> Conrad-Jonson PV10 AL -> Adcom GFA-535 power amplifier -> TDL Electronics RTL 4 / with active Proel EIKON10S subwoofer.

Looking to reasonably replace Hifiberry to make use of my Tidal HiFi subscription. To put it otherwise What DAC that can do MQA is sensible purchase for this setup?

I use the Teac NT-505. I am conscious that it may well sit in between the two examples here in terms of price positioning but it does what both people need in an exemplary manner. Roon and MQA capability is done via the streaming input. The streaming engine is compatible with Lumin and is the same as Esoteric use in their streamer (an alternative option if you have more to spend). The DAC section makes use of the expertise Teac have built up over multiple generations of their modern DACs. I use mine straight into NCore power amplification so there really is nowhere to hide from poor quality, clicks, pops and other potential gremlins. These have been taken care of and the device is deathly silent and with a bit of care taken with the Ethernet network it is truly transparent in a way USB never really was for me.

I use a Rossini with Roon, so expect me to be biased.
dCS have done a wonderful job with the streaming engine in their DACs. Via ethernet high res to 24/384 and DSD128 can be streamed. Via SPDIF its 24/192 and DSD64. The beauty of the ethernet input is that you don’t need to worry about compatibility issues with an external streamer. Forget high end USB cables, USB decrapifiers. Tidal, Qobuz, Roon and Spotify and (dare I use the unword) MQA are built into the dCS Mosaic app of the DAC. There is a Mosaic control app for Android and iOS. The MQA decoder has been built by dCS to their own high specs.
I find the sound quality of the Rossini superb. The Bartok should be identical in sound to Rossini with FW 1.0. FW 2.0 improved the Rossini to the level of the FW 1.0 Vivaldi.
So my recommendation for a fuss free setup that will likely last you 10 years plus is the Bartok.

Many thanks to everyone for their responses, and many thanks again for responding to my questions below, and educating me on these issues.

The dCs specs confirm Rudi’s report, namely: “Network interface on an RJ45 connector . . decoding all major lossless formats including FLAC, WAV & AIFF at up to 24 bit 384kS/s native sample rate, plus DSD/64 & DSD/128 in DFF/DSF format.”

However, by contrast, the T+A DAC 8 DSD specs state that over USB it can “accept digital PCM stereo signals with sampling rates in the range 44.1 to 384 kHz, and digital DSD stereo signals from 2,8 /3,1 MHz (DSD64) to 22,6 / 24,6 MHz (DSD512)”

I’m not an expert on DSD and am a bit confused by the different formats. Acoustic Sounds sells downloads that are “Quad Rate or 11.2 MHz/256fs.” A review on the AudioStream web site said that “the buzz surrounding the T+A DAC 8 DSD is all about octuple-rate DSD; DSD512 (512 times that of CD)/22.5792 MHz.”

What does it mean that the T+A DAC can accept DSD 512; whereas the dCS units can only handle 128?

Is this a difference that matters?

Is it an important reason to consider one DAC over another?

1 Like

No. My DAC will do 512 via USB but I haven’t bothered to try it. All of my owned DSD is ripped 64 stuff. The rest are free samples and test downloads. Those 512 files are massive and sonic benefits, while there don’t justify the compromise of having to use USB in my view.

due to a persistent/permanent firmware issue for the usb receiver, the t+a dac8 dsd can only play dsd256 and dsd512 when used with a windows OS based PC or streamer. consequently, those with linux or mac systems are limited to a max of dsd128 via DoP – this includes users of the roon nucleus as well as any streamer running roon ROCK

also, the dac8 dsd does not have an rj45 ethernet port which precludes network connectivity.

i have the dac8 dsd and am quite happy with its performance – the SQ when playing PCM is exceedingly good and equal to that of dsd128 in my experience. it also offers a huge bang-for-the-buck and has incredible build quality.

do not know much about the bartok; however, in that price range there are quite a few superb options which i think are worth investigating. examples include lumin x1, msb discrete or premier, chord dave… just to name a few

1 Like

In principle this is correct.

However: There is NO DSD512 source material available. The DSD512 option of the T+A is used only in conjunction with upsampling source material to DSD512 (e.g. using HQPlayer or another upsampling engine). Some people claim that in this use case, the T+A sounds superb. I have not tried this.
I have compared the T+A in a “normal” usecase, not upsampling to DSD via an external upsampler and in that case I found it clearly inferior to the Rossini. DSD less so than PCM, but still inferior.

@AMP Yes the dCS Bartok currently does not play DSD256 natively. I expect dCS to deliver that functionality at some point in time in the future via a firmware upgrade. I have very few DSD256 albums that Roon currently resamples to DSD128 for me.

I responded in a similar topic so I’ll include the same here for the record.

Our products currently do not support DSD rates higher than DSDx2 or DSD128. This is not a technical limitation and is only related to what we have elected to implement in our software at this time. We may support rates higher than double-rate DSD in the future, but there is a fair amount of work to be done in filter development since we use bespoke filters for each incoming sample rate. It’s not as simple as just flipping a switch and having a new format supported.

We base our feature content on technical evaluation as well as overall market demand. In the case of DSD256 and above there is very little market demand. Furthermore, there is very little truly native content available in that format and what is available is primarily sourced from one (very small) provider. While we fully support the work that this provider is doing and truly wish them all the best if they were to disappear from the market then there would be virtually no content available in these higher-rate formats. From a business standpoint there simply isn’t enough critical mass here to support the development effort. If that changes then we will reevaluate.

The reason that this feature is available on a number of other products has little to do with the technical benefits of the format and more to do with the fact that once the design is etched into silicon it cannot be changed. That encourages chip manufacturers to include every possible feature that they can think of so as to maintain the market longevity of their chips (and the products that use them). They have to include this support since if it weren’t there then a major market shift could prematurely obsolete their designs and the products in which they’re used. The issue here is that when you attempt to design something to do absolutely everything there’s a high probability that it won’t offer truly cutting-edge performance in anything. That’s not meant to denigrate any particular chip or chip manufacturer, it’s just an engineering fact of life.

We aren’t limited in this way and can easily reconfigure the FPGA to do whatever we want (within the limits of available processing capacity) and we can do that whenever we want. The only requirement is that there be processing capacity available. We design our products with a significant amount of headroom so as to be able to freely evolve our offerings over time.

1 Like

I’m the guilty party for posting twice. My first post (in this thread) was done when I was attempting to decide among several DACs, and also attempting to get better educated about the issues. I subsequently posted in the dCS thread on the Bartok when I was better educated and considering that unit, specifically directing more detailed questions about that unit to Andrew. So apologies to Andrew and others for feeling that they must double post. In any case, I greatly appreciate the detailed responses from Andrew and others on this set of issues.

In any case, I’m attempting to accomplish several goals. One is to improve two channel sound quality, and the questions about DACs and the Bartok are aimed at that objective. The second is to intregrate a DAC in my main listening room with sound in other rooms of my house, and be able to group RAAT Roon-approved devices together. I’m receiving several Bluesound items today. (And yes, I separately posted in the Bluesound thread about that.) If there is a weak link in my plan, it is Bluesound, given the number of negative posts about that product. Hopefully Bluesound will operate properly in my house. It appears to be a hit-or-miss proposition, and Bluesound has a reputation for blaming Roon for any problems, based on reports in this forum, which is bogus IMHO.

Again, many thanks to all for the excellent and informative responses. I hope that others in the future will benefit from these responses as much as I have!

I think, if anything that link confirms the point. 16 discs worth of plinky plonky audiophile music. :rofl:
And I know the Jacintha CD wasn’t recorded as 512, I wonder how many of the rest were? I have the Girl From Bosa Nova as hybrid SACD.

Details are already there in the ‘tech specs’ for each album…

What do you see as the issue with using USB?

I think USB is for convenience based on the generation of DACs which have been around for the last few years. Quality of implementation can vary a lot. So one setup can happily be connected with a short but otherwise standard USB 2 class cable while others might demand cleansing and boutique cable to be at their best. In my particular case I didn’t want or need to go down that rabbit hole again. Especially for the sake of a dozen or so files bigger that DSD128.

OK, I was just curious.

Optical/SPDIF have their own different issues as well that can vary from one DAC implementation to another according to how they go about synchronising or adapting received data to the DAC - ie dealing multiple clocks vs a single clock.

Neither is perfect :slight_smile:

See the thread on the dCS Bartok, where that issue is discussed at great length by Andrew, the manufacturer’s representative. He discusses the case for using Ethernet, as well as the issues posed by USB.

Thanks. I was more interested in the reasons behind Henry’s view/choice which he has kindly explained here and coincidently to a greater extent in another thread.
With that said, I always welcome deeper insights from those with greater knowledge, so thanks for the reference - I do agree with most of the comments from Andrew of dCS in his various posts. Indeed they should be compulsory reading for many people considering esoteric digital interconnects and other strange gadgets :slight_smile:

You might want to consider the MSB Technology DACs. I have the now outdated Analog DAC. It has a Roon Ready Network Renderer to accept ethernet. It is a superb piece of equipment.
These have been replaced by the improved Discrete DAC. These come with a range of different input choices.

1 Like