Discrete output

Hello, first post here. Happy owner of Bartok. Loving Mosaic also.

I think I recall reading that the RCA output on the Bartok is via opamp, while balanced is fully discrete. Is this true? I may be new to dCS, but in my many years using higher-end components the consensus was that opamps were usually relegated to either portable or lower-grade equipment. Same with avoiding ribbon cables to conduct audio.

Main reason I ask is that I use Bartok to drive a single-ended First Watt amplifier directly, and with efficient speakers I am attaining great volume and terrific sound quality even with the audio output set to 0.2V and 0dB attenuation. Is there more to be gained via the XLR? If so, I will need to use a preamp that accepts balanced and outputs single-ended (which I do, but prefer not to add to the chain).

Second reason is that in researching this matter I’ve stumbled upon some threads on a certain headphone-oriented forum site where Chord fans were repeatedly pointing these out (opamp and ribbon cable internal interconnects). Chord products are utterly undesirable to me purely on account of their (to me) ghastly visuals, and I do take forum postings from competing products’ fans with tremendous amounts of salt, but nevertheless would like to hear the other side of the story.


Hello Nick,

Welcome to the community. You may find reading these FAQ topics informative:

Hello Erno,

Thank you for the info. I will give it a try with balanced cable terminated as advised and see what it sounds like. I could of course use my preamplifier but prefer to go direct if possible.


Well that’s confusing me: balanced output gives noticeably better sound quality and can drive an unbalanced input in which case they recommend unbalanced output😫. Does that mean I should use a XLR to RCA cable terminated correctly for optimal sound quality?

You should read it as this:

  1. The balanced outputs behave like coupling transformers so they can drive an unbalanced XLR input provided that XLR connector pin 3 is connected to pin 1 at the end of the cable farthest from the DAC. Used like this, you may find that muting clicks are more noticeable.

  2. We recommend using the unbalanced RCA outputs to drive an unbalanced RCA input.


Thank you. I am easily confused. I will just have to try to learn to live with the inferior sound quality of the unbalanced output. Is that what they really mean? Why didn’t they make the RCA outputs as good as the XLR?

Or do I have to buy a balanced amp?

It couldn’t hurt. You’re suffering from GAS after all :rofl:

Good point. I am trying to cut down though. And I have a back issue so I’ve found the suffering power amp demos bring can be really unpleasant.

I just can’t believe dCS would out a poorer quality RCA output stage in the Vivaldi DAC. Shirley some mistake!

Same here—I much prefer keeping the audio chain short and simple. I ordered a custom-terminated interconnect and should have it in a week or so, will give it a try then and see how it compares.

Why do you assume that the RCA output is worse ( rather than different) to the balanced one? Yes a balanced connection should provide lower noise figures than single ended but that is the inherent nature of the beast. If you need to use single ended then there is nothing that can be done in regard to that aspect. An XLR/RCA cable will not provide a balanced connection in any case even if it does provide electrical contact. So,in answer to an earlier question from you, that does not provide optimal sound quality. A balanced connection has to be balanced at both ends.

Your assumption seems based upon an audiophile urban myth that OP amps are always inferior to using discrete components. That was largely true in the last century but things move on and these days OP amps can offer excellent performance. It is a case of selecting the best tool for the job.

Knowing dCS’ ethic of making everything as good as possible I would be confident that the Vivaldi RCA output uses the finest suitable components. It so happens that, for this particular purpose, the dCS engineers have decided that using OP amps provides a better performance than could be achieved by other means. Why would they do otherwise?

I know you’re not aiming this at me, and I only have the Bartok, but I originated the question in this thread so allow me to offer my thoughts: a properly implemented opamp is better than a poorly-conceived discrete solution. I don’t think anyone is doubting the prowess of dCS engineers, but the essence of engineering is managing tradeoffs given constraints. Space, for one, is a constraint. Cost, not so much with dCS, as they don’t compete in the value segment. Perhaps there would not have been sufficient space for yet another discrete output circuit. Or perhaps that little extra power consumption would have put the whole unit close to needing a new power board. etc. etc. Additionally, perhaps dCS have surmised that people playing at this level are more likely to use a balanced audio chain and hence the unbalanced is provided almost as a “backwards compatibility” solution (albeit a terrific-sounding one).

My reason for starting the thread was to attempt and find out if there is a tangible (well, audible, rather) improvement using the balanced output on grounds of the circuitry involved (in addition to the intrinsic benefits offered by a balanced connection).

Thanks for your response. If you really get into the dCS frame of mind my ( reasonably well informed) guess is that if the there was a need to compromise the RCA output board they would either have not done it at all or would have changed the box to suit. After all this circuitry is more or less what they have been using for a long time with earlier models so when designing the third generation ( Vivaldi etc) they already knew what was going to be required.

Why do I think that compromise ( or, more accurately, too much compromise ) is not acceptable at dCS? Did you ever read the reason that they decided not to add a USB output on Network Bridge even though it was highly desired by their customers? This really indicates their mindset:

Thank you Pete, it is refreshing to read their candid response. So far I have thoroughly enjoyed my experience with dCS and its companion Mosaic app. Also the kind and respectful atmosphere in the dcs discussion community. A rare breed these days.


It came from the dCS FAQ quoted by Erno:

Am I interpreting that incorrectly as suggesting that there is more to the difference in the output stages than just the balanced nature of the circuit and that the balanced output stage sounds better than the unbalanced (RCA) for reasons other than just the balanced design?

Not at all. I don’t know what an OP amp is. It’s just the FAQ statement.

I agree. Which is why I find the FAQ so puzzling. It would be really good if dCS could answer Nick’s original question or reword the FAQ.

Yes the FAQ requires clarification. Are they really saying that these particular RCA outputs are inferior or that these RCA connections are the best achievable but that balanced connections are in any case inherently better due to e.g. improved noise rejection?

I wonder what the situation is.

Just from magazine tests of dCS players (mine is a P8i mkII): Better sound with fully (end-to-end) balanced, but the reason may be that the balanced had the top priority in the design process. This is a little different from putting emphasis only on RCA/XLR technical properties (even if these also point to balanced as the best). Other players may be made the other way around: designed as single-ended with an XLR connector as an extra option. I would try an XLR to RCA cable, it might be better on a dCS.

How did this comparison work out?

Wondering whether to use the dCS DAC’s…

  • RCA output for my single-ended headphone amp
  • XLR output for my speaker system

…or run both off an XLR splitter like the Dodocus boxes, with an XLR—RCA cable for the headphone amp.

The FAQ stating that the XLR output is higher quality. And contrary to others’ opinion I believe a good discrete Class A stage should better pretty well any opamp out there. Easy enough to prove: just substitute a Burson discrete output stage for one of the best opamps around and I’m sure you’d prefer the Burson (or Sparkos etc) option.

But why ask for opinions when it’s so easy to test? Buy some good quality adapters and try with the same phono leads. I’ve only had my (2nd hand) Bartok a matter of days and tried this briefly. With a quick test I could not definitively say one was better than the other, but I will try at greater length.

There’s plenty of ‘fun’ tests we can do. EG I compared the USB input with the Ethernet one. I played my Mac Mini through an Innuos Phoenix reclocker and compared to the Ethernet connection found little or no difference. Again, I will try at greater length later. Of course I’m playing the USB connection thru a £300 lead, and currently have a £2.50 Ethernet cable, so I’ll upgrade to a Blue Jeans cable and see if that makes any difference!

1 Like

I ask only because I don’t own an external headphone amp yet, and other people’s experiences help me work out what sort of direction to take there. Once I have something to test, I’ll set to :blush: