While I have nothing to observe regarding the transparency of the digital volume control in then Vivaldi I can observe that the line driver circuit in the Vivaldi is not as good as it could be. Having compared single ended and balanced outputs using identical cabling I found the single ended to be substantially more transparent.
I run my Vivaldi into an ARC Ref 40 and this is a great combination, the RF then drives 10M of balanced cable to my amps - something the Vivaldi would not be comfortable with at all
I also found marked differences between the output level settings with .6V sounding much better than 2V, this may however be a function of interaction with the preferred range on my preamps volume control
|
42k for the VTL MB450III, but not sure why that’s relevant as I’ve never tried driving them direct from the DCS ... |
The Ref 40 input impedance is 60k SE, 120k balanced. The Vivaldi spec recommends 10k-100k ohm |
It's surprising how definitive people can be based on measurements of components they do not own, and may never have actually heard? Anyway I'm happy to connect my Vivaldi directly and report back. As I prefaced in my original post this is not something I've ever tried given my setup and my past experience with direct connected digital components which always sounded worse but I'll have an open mind 😏 But perhaps you have all forgotten about this long discussion ... https://forum.audiogon.com/discussions/the-jadis-ja200mk2-review-or-slam |
And wherein perhaps lies Solomon’s answer ... for if the dCS output is truly native SE then my experience that SE sounds better than balanced via a preamp (same cable, same length) then it could be the case that dCS->VTL balanced sounds worse than dCS->ARC (SE)->VTL (BAL) especially as the VTL itself is a balanced design ... so lets see
ps the Vivaldi has seperate output stages balanced and SE, and both can be driven at the same time, question is which is closer to the Ring DAC output ... which is ostensibly a balanced design but not clear what they have implemented in analog circuitry after the DAC -- clearly there’s quite a bit there at least to switch output levels and so on (btw interesting side note is that the BAL and SE output levels cannot be set independently ...) |
The input sensitivity of the VTL MB 450-iii is a function of damping factor but at the setting I use (minimum damping) is a low 775mV hence one needs to be very careful in setting the level on the Vivaldi. Luckily dCS provide a 0.6V (and a 0.2V) setting as well as 2V and 6V.
I currently use 0.6V with my active pre-amp and find that this works well with the volumes I like to listen at (i.e. operating the amp at close to unity gain in effect) so my presumption is this would be the best place to start with a direct connection so as to minimize the amount of cut I need to take in the DAC volume control and more importantly avoid the danger of over driving the speakers with catastrophic effect (this has happened to me in the past when I had a cable short and it blew the woofers in a set of Magico V3s)
|
One quick addendum (albeit from audio memory)
I previously noted a difference between the balanced and single ended outputs on the Vivaldi and my preference for single ended. The difference manifest as somewhat of the same sense of compression and homogenization noted in the direct so perhaps this is largely a function of how dCS have implemented the balanced output (perhaps another amplification stage).
I did not see any difference in soundstage however between the two connections when going via the amp -- so as @shadorne suggests perhaps that is an (artifact) of the pre-amp 😏 |
Time to report back on my experiment with comparing the Vivaldi DAC direct vs via my ARC Ref 40 preamp. The full description of my setup can be found in my virtual system but the key things today is that we were comparing two different analog outputs from the DAC to my VTL MB450III Path A was my usual setup - DAC to ARC via 2M Single Ended WEL Signature
- ARC to VTL via 10M Balanced WEL Signature
And Path B was direct - DAC to VTL via 10M Balanced WEL Signature
The DAC is set at 0.6V output and the VTL has 0.775V input sensitivity All levels were matched using pink noise (Stereophile test CD2 track 15) and incidentally my listening levels for the tracks were between 9.5dB and 14dB cut on the DAC Four tracks were used for comparison, all CDs 16/44.1 upsampled to DSDx2 - "Bye Bye Blackbird" by Nancy Harms from "In the Indigo"
- "Old Coyote" by The Weepies from "Hideaway"
- "Jardin d'hiver" by Stacey Kent from "Raconte-moi"
- "Wohl mir, dass ich Jesum habe" track 1 from "Bach in Jazz" by Stephan-Konig Trio
All are tracks I know intimately Overall it was an interesting experience with as some have suggested two different presentations. If I use Path A (via ARC) as my reference the direct route was in comparison definitely more analytical, arguably clearer but with some important caveats On the plus side - Preserved full soundstage width, possibly sounding even a fraction wider?
- Very analytical ie. ability to distinguish artificial reverb on the Harms track vocal
- Greater clarity on individual instrumental lines e.g. different elements in a percussion kit could be followed very well
- Seems louder than the route via the pre-amp (note they were level matched)
But a number of significant negatives - Overall homogeneity to the sound, the instruments all tended to sound like one another
- A sense of compression (matches with the perceived louder)
- Tensing up on peaks - especially male vocal (e.g. the Bach track) which sounded forced
- Massive collapse in soundstage depth (maybe 20% of the perceived depth conveyed via the pre-amp) -- especially evident on simple miked track like the Bach where its key to keep the scale and relative position of each instrument distinct -- direct they all seemed to be on top of each other
- Some tailoring at both frequency extremes -- bass seemed less deep, very high treble (supertweeter range) seemed to be missing
So if I was in a studio and doing a mix I might prefer the direct but for recreating the sound of real musicians in space the pre-amp is definitely adding something -- I don't know if its coloration but if it is its a very beneficial one. I think the massive power supply and ability of this pre-amp to swing peaks while still preserving the lowest level details (i.e. the soundstage information) is actually what's making a difference here I can understand that others may prefer the direct path, arguably my Magicos sounded more like classic Magicos going this way that's fine, and also bear in mind that my entire system is built and optimized around including the pre-amp |
Just to add to this discussion on space and warmth. I cited the example of the Nancy Harms track and the reverb. Via the direct connection it was clearly artificial reverb in a halo around the voice. The striking double bass was clear and direct but with no sense of body or presence in space.
However via the preamp the bass was palpably in its own space and had scale and heft, almost three dimensional, and the reverb on the vocal dropped back behind in space.
So ok maybe this was all studio trickery but via the preamp it sounded more like a group playing together in one space while direct it was just like a bunch of tracks on a mixer
Of course for recordings that were more honest like the jazz trio Bach the spatial cues and scale were all present via the preamp but much less distinct and appreciable in the direct connection
|
@jareko if you look at my virtual system you will see the (somewhat extensive) steps I have gone to to optimize around the Vivaldi. As you note such details as footers, power cords, cabling, clocking, damping all matter and are very audible.
However even with all of the above I for one would never choose to listen to it direct when I have the option of a pre-amp like the Ref 40.
This of course is in my system, built up to suit my preferences, silver cable and all.
The continued observations of "experts" who based only on specs and their biases know what other users systems sound like and which is obviously "best" is one of the mysteries of on line forums such as AudioGon
My rule is to take note of input based on personal experience only and discount all others |
@grannyring asking me for folk recommendations is opening rather a can of worms ... Firstly I will note in passing that I am not much of a current US, so-called folk/Pickathon fan -- all rather arch and affected to my taste My preferences are more towards the stalwarts of the 80s and 90s. It is great to see Cry, Cry, Cry back up and touring again and of course any of the work of their three members ( Dar Williams, Richard Shindell and last but in my opinion best Lucy Kaplansky) is worth looking up I’m also on a big Jonatha Brooke jag at present -- a very solid artist with a long history Among new artists I’m really liking The HarmaleighsMy real passion however is for the folk tradition of the British Isles, which is where I am from. What makes this community distinct is how the historic tradition is constantly reinvented and revitalized by each generation. I love hearing the sons and daughters of the past generation taking the music and making it their own. Great places to start are the artists on The Elizabethan Session album, or any of the work of Kathryn Roberts and Sean Lakeman, or of course Kate Rusby -- I’m really into her 20th anniversary album where the cross over with guest artists (including many Americana names you will recognize) gives a different take on her songs Finally the Scottish tradition is very much alive - check out Salt House and my current number one album and something you must listen to (it’s a concept piece so ideally you listen all the way through) -- A Pocket of Wind Resistance by Karine Polwart |
|
Why match? To avoid having to remember to add or cut 20-30 steps on the volume control or risk playing way too loud. As it stands the two inputs use the same range which is what I want. Also we want to avoid at all costs using any gain at all in the pre-amp and reducing to 0.2V would mean needing to use almost the full gain stage (12dB) to get 0.7V rated output (not that I of course drive the amps that high but peaks may) while 600mV is much closer to the desired 0.7B output sans gain ... make sense?
In other words the volume control setting and cuts I describe represent the average, the peaks may be close to unity
|
Fair point @seigen but the other consideration is level matching with my phono stage. With the current settings the two need the same rough preamp volume settings and so avoid the danger of playing too loud. Gain matching across the entire amplification cascade is an interesting topic and one I’ve previously commented on (albeit in the context of phono) |
Plus I should add that the test tone I used for level setting is -20dBFS so if you like the peak you are actually hearing with 14dB cut is +6dB
This all gets confusing which is why I always keep my gain cascade spreadsheet handy and check all the levels so that I get as close as possible to 0dBFS at power amp input sensitivity at the level of preamp gain I prefer
|
Will you guys please get of your hobby horses and do some simple maths.
My amps rated sensitivity is 750mV. The 600mV output of the (0dBFS) of the DAC preserves as much of the potential dynamic range in the signal without needing either attenuation (acceptable but undesirable) or amplification (always and in every case undesirable)
When using a test tone at -20dBFS my preferred volume settings were in the range of -9 to -14dB -- now no one listens to test tones, we listen to music with dynamic range -- in the case of the dynamic range database the peak to median level for the Stacey Kent track I use is 14dB (which by the way I play at the -9dB setting) -- if we assume (which is wrong) that the test tone represents the median (it doesn’t as test tones tend to need to be played lower than music) then the setup I have corresponds to peaks almost exactly hitting unity gain
I know you people believe that in all circumstances a preamp is bad but none of you have offered any explanation as to why the direct connection appeared to limit dynamic range and cause the peaks in the tracks (especially vocals) to harden?
ps interesting that only now do you talk about bit stripping .. of course by using the dCS at 0dB this problem is avoided, plus I understand that the implementation of the Vivaldi DAC avoids this issue ... |
|
Fine -- and when I change my phono stage and drop from 45dB gain to 40dB the 0.2V may also be a better match than the 0.6mV
But still no explanation of why the direct connection does such a relatively worse job of handling dynamic range (if we put all the soundstage stuff down to coloration, which btw I wholeheartedly do not) -- and no it’s nothing to do with the -9dB cut as the dCS volume control is digital and non bit stripping |
@georgehifi -- I'm very happy with the verdant, lush, wild and exciting primal woodland panorama my tube based active pre-amp gives me thank you very much and will not be spending a moment more worrying about fussing with going back to a direct connection
Having said all that spinning vinyl has turned out to be so much more rewarding these last few days -- loving the Weitblick Peinemann WDR releases for example |
|
