I’ll keep you informed 
The funny thing as an answer to the topic question is; nevertheless if the 2v Red-Book reference or 6 V is better, there is an pretty huge sound difference between 2v and 6v whereas 2 v sounds soft and laid back vs 6v the soundstage is much more pronounced and mor forwarded and focused.
And this behaviour is on all dCS dac i have tested connected to all different amps i have experienced.
I tend to like 6 v more if you like more attack and focus vs 2v.
1 Like
I have to say this pretty much flies in the face of the explanations dCS have given here about the differences between the voltage outputs. What would be the explanation for these SQ differences?
Please read this article, and ignore the Benchmark part:
Many high-end consumer products have balanced interfaces, but they operate at a maximum level of 4 Vrms which is + 14.2 dBu. This is 10 dB lower than the level used in professional interfaces. This means that the 13 dB advantage provided by a +24 dBu [ which is 12.28 Vrms] balanced interface is reduced to just 3 dB when operating at a maximum level of +14.2 dBu. Consumer-grade balanced interfaces are definitely a step better than unbalanced interfaces, but the signal levels are too low for use in very high perfomance systems.
This is only my experience since a 15 year dCS user , and this includes all my friends that owns dCS also.
It is extra obvious if you are using dCS as an pre amp, but you can here it with an pre amp connected also.
So try it by yours self and see what you think.
Thanks Erno, I read the entire interesting piece, and it all makes perfect sense, but it still does not address the points @James made in this thread. There is no reason to be hearing different sound stage, depth, etc., as described by Fredrik between the differing voltage output settings on the dCS equipment. I’ve been listening at .6, 2, and 6 over the last few months, with and without a passive pre. I might/may/maybe/possibly hear a difference that we would characterize as “blacker black” (lower noise floor) on the 6V setting. But it’s a tenuous perception at best. And as to other SQ differences, there are none. Doesn’t mean Fredrik isn’t hearing it; only that I am not.
Greg, your post inspired me to do a quick comparison test between the 2V and the 6V setting.
My Vivaldi stack output is usually set for 2V going direct to my Amp which has a sensitivity of 1.3V - allows me to listen mostly at “25dB” or better volume setting on the Vivaldi.
Using a Dayton OmniMic, I recorded my Speaker output from the listening position with the Vivaldi set to 2V compared to 6V; Speaker outputs level matched on pink-noise to within 0.2dB, and recording peaks at -9dBFS, tops.
It was tricky to perceive any tangible differences, but analysing the recordings via Adobe Audition, turns out the 2V setting recording has a 3dB better dynamic range than the 6V setting! 
I think all things being equal, clearly it’s best to follow dCS’ advise and set the DAC’s output voltage level to whichever is closest to the Amplifier sensitivity. 
3 Likes
Or to the preamp sensitivity, which could make a lot of positive difference.
So I agree with your (and dCS) advise (to set the DAC’s output voltage level to whichever is closest to the Amplifier sensitivity), but like to add that IF one would use a 6V capable preamp, one could have a positive effect of 6V out.
1 Like
Absolutely. Relying on subjective preference does not necessarily result in the best objective result. As I have mentioned in another thread, additional but limited distortion e.g . by mildly exceeding a preamp’s rated input sensitivity, can be interpreted by the ear as preferable with some musical genres. BTW I am not saying one should not prefer it if that is your thing but that it shouldn’t be mistaken for a bona fide improvement.
2 Likes
That is not the way to go. One could use a preamp with a 6V (or higher) input sensitivity.
[
But some seem to like it 
Audio distortion is a fascinating subject. One stimulating article on deliberately adding low order distortion to provide an improved subjective result by masking high order distortion was written by the recording engineer Bob Katz and published in the now defunct AudioStream. Worth a read:
2 Likes
If the ceiling of a device is 6 V (rms) and the noise floor is 10 µV (rms) then the dynamic range is 115.6 dB.
20 x log10 (V1 / V2) = 20 x 5.78 = 115.6 dB
For 2 V this is 106 dB.
So, theoretically, the dynamic range at 2 V should be - 9.6 dB lower, not +3 dB better. When you measure +3 dB, then the level matching did it (and the -9 dB peak), not the output V setting.
Besides, you recorded your Speaker output from the listening position. That means you recorded sound amplified through/ boosted by your room.
2 Likes
But I think that is precisely the correct thing to do. What matters is the total practical outcome, not only the theoretical construct. The theory and math calculations define the bounds, but not the actual outcome. It’s important to understand all the various pieces in play, and that would include the room.
Level matching may be the Gordian knot of the problem, though. We’ve been taught for decades to level match to remove the natural influence of loudness on the perception of “betterness.” But dynamic range includes loudness as “a feature, not a bug,” so to speak. Intriguing. If 6V indeed presents the possibility of greater DR, as the math indicates (and I believe noted elsewhere here in the forum), and it can be run into one’s amp, shouldn’t one prefer that overall? (That’s back where I started almost a year ago.)
This has become quite the interesting discussion. At bottom, we all seem to agree on two pragmatic outcomes:
- Match the DAC output selection as closely as possible to the preamp/amp sensitivity; and
- If you’re running the DAC direct in to the amp, set the output to the range that allows volume control by the DAC in the -20—0db range.
4 Likes
And this could be the third pragmatic outcome 
1 Like
I love this thread. I also wish it weren’t proving so interesting again 
My T+A integrated has a maximum input of 6V, so because of my past tests with separates I’ve been using the Bartók at 6V output without any further comparisons. It sounds wonderful.
Reading this on a Sunday morning has me planning a 2V/6V comparison, though. Or, maybe, for extra fun it should be 0.2V/6V. Uh oh.
I’m pretty sure that’s an inherent subset of the first outcome I suggested. If one can only use 2, use 2. If one can use 6, use 6. Same.
It is, but it emphasizes that one could be better off with a 6V or more preamp input.
I have flipped about between 2v and 6v, its very hard to really compare them, as matching volume is very difficult and the greater volume obviously sounds more alive.
Luckily my naim 552dr can take the 6v ok, but having a turntable as well, the 2v matches much better.
Also i feel that my naim 500dr likes producing a bit move volume, rather than the easier load at 6v.
But its all very system, room and yourself dependent i feel, for me right now its 2v, but its good to be able to switch.
2 Likes
Yes indeed. Mosaic makes this very easy.
You might have missed the point actually Erno 
Greg is spot on. This little exercise I did on a lazy Sunday afternoon wasn’t about the theoretical performance of the dCS stack… I actually wanted to know if I could perceive an actual audible difference between the 2V vs. 6V setting, when fully level matched out of my speakers.
Turns out it’s exactly as we already know; the closer one gets to “0.0dB” on the dCS volume knob, the better it will sound. Whichever way you get there; through the use of a Pre-Amp, or by closely matching the output voltage setting to your direct connect Amp’s sensitivity.
The discussion of dCS’ digital volume control usually revolves around “resolution”, I must admit I was a little surprised to see it so easily measured as changes in dynamic range (on hindsight, I wonder why Stereophile’s John Atkinson never bothered to measure this 
).
2 Likes