Audio dealers/makers and their crack runner fans. There’s always a way to dig into the wallet deeper. 
I’ve got an electrician here today, too. But not for my audio, thank goodness.
In the end I “settled” for one Everest…
I understand now that the truncation of the first second or so of DSD tracks is a known problem at DCS and a fix is in the works. It apparently does not affect all DSD recordings, is not going to be easy to resolve, and there is a workaround to use a USB to play these tracks (not worth my effort at this point).
I’ll be patient.
David, as turns out, I actual do have Elton John’s “Goodbye Yellow Brick Road (30th Anniversary Deluxe Edition)” double disc SACD with the “Candle In the Wind (acoustic mix)” track that you mentioned in a previous post. So, I ripped the SACDs to see if I could replicate your issue.
Yes, I’m experiencing the exact problem you describe with the truncation. However, only under one single condition; when played back via Mosaic immediately after a PCM track.
I tried a PCM 24/96 track, and queue’ed up Candle in the Wind DSD64. During track transition, the Upsampler switched through 24/96 to 24/176.4 to 0/176.4 and finally settled on DSD, and the truncation occurred.
The problem does not occur under any other circumstance, say if it’s preceded by a DSD track from the same or a different album, or even if I playback via Roon and preceded by a PCM track - looks like the Roon/RAAT stack on the Upsampler stream board is working as it should, whereas the Mosaic/UPnP stack needs some fixing
ps: By the way, this issue does not affect my Bartok, there’s no truncation even when playing back a PCM track followed by this DSD track via Mosaic. So, it’s just the Vivaldi Upsampler-DAC combination.
Here’s the track if anyone’s interested;
Anupc,
Thanks a ton for checking, and assuring me I’m not crazy…I have the same issue if I play a hi-Rez recording from Quboz after a PCM 24/96 recording too. I have poured over the Vivaldi manuals and changed every setting I can find to resolve the issue…I thought the DAC buffer would be the solution but nope. I question what purpose the DAC buffer serves if it doesn’t fix this specific problem.
I had my collection of SACDs ripped to DSD when I got the Vivaldi stack and sold them off along with my Puccini SACD/CD player and UClock, assuming I could achieve the same sound quality. The SQ is there, but the truncation is annoying.
It is not the end of the world for me, since anything I have on my NAS as DSD is also available in PCM but I play guitar along with my stereo, and it is kinda jarring for a song to start on the third strumming down stroke, as is the case with my DSD version, or a hi-rez version from Quboz, of A Horse With No Name. It is hard enuff playing the song, without giving the band a head start.
I’ve been advised that a subscription to Roon will probably solve the issue, but I’m happy using Mosaic as my interface, and don’t need Roon for anything else.
Cheers,
Mitch
David, would you kindly clarify your problem as it doesn’t seem to be exactly the same as Anup’s or as the other David ( acousticsguru) . Both of their issues involve changing from PCM to DSD whereas I understand your issue involves only PCM recordings ( all Qobuz files are PCM ). Are you upsampling the Qobuz track to DSD ?
Have you tried changing tracks rather than playing through?
Recall the Rossini issue is seen when jumping between DSD tracks, but not when playing from one DSD track to another; I don’t know what happens if you play from a PCM track to a DCD track, I would expect the same loss of the first .84 second.
Recall also playing .dsf via USB only works for tracks after the first; the first is always truncated as when the datastream stops it assumes it is PCM and the delay in detecting and switching to DSD is what is the root cause here.
By comparison, a network connection defaults back to PCM between each track, so it will always truncate the beginning.
The problem (at least on the Rossini) is that it takes time to determine whether a bitstream is PCM or DSD and if DSD switch to the DSD decoder.
However, rather than cache the data that is incoming while the detection and switch occurs as virtually all other DACs do, at least the Rossini either just drops the DSD data received to that point or sends it to the PCM decoder assuming it was PCM.
Imagine if you had to switch flow coming in from a main to one of two pipes based upon whether the content was water or alcohol.
Rather than build in a reservoir to hold a second’s worth of liquid flow while you decided which destination pipe got the flow, you just let whatever was in the pipe spill onto the ground while you figured out which liquid it was and routed it appropriately, or sent one second’s worth of alcohol down the water pipe.
That’s what dCS is doing in the Rossini.
The only solution at the moment is to use playback software that can insert a second of silence before each new track - the equivalent of pouring one additional second’s worth of alcohol into the pipe to make up for the amount you would know get spilled out/misdirected on the receiving end.
1 Like
You’re right Bill; if I force a track change mid-track when playing back a DSD album over the network on my Vivaldi Upsampler, the truncation problem does occur. The Upsampler reverts to PCM before switching back to DSD. Same thing on my Bartok as well. So, the bug is across the family.
However, it does not occur when the album is allowed to play track-after-track by itself, nor when tracks are queued with “Add next”. No reverting to PCM between tracks then.
Likewise, no problem at all when using Roon to playback DSD tracks either, even jumping between tracks, the Upsampler never reverts to PCM (even with Resync Delay set to 0ms just to be sure).
I’m sure our dCS friends are well aware and working to resolve this with a Mosaic firmware update. 
2 Likes
Now that there’s been one, is your experience any different?
Similar issue as above. Vivaldi Dac and upsampler, no clock, but with clock cable connecting the two.
Using dcs’ own qobuz playlist: http://open.qobuz.com/playlist/2744707
Using Mosaic Play track two “Overnight”, then after a few seconds go back to track one “An evening I won’t forget.” The dac buffer doesn’t seem to help.
https://www.dropbox.com/s/klg3jdgpf5rsk13/Video%20May%2014%2C%205%2020%2046%20PM.mov?dl=0 Shows me starting the playlist from nothing and I lose a full second. Jump to track two which is ok enough, back to track one and I lose half a second.
Mosaic 1.1.3 seems to have resolved the issue on my Bartok, I couldn’t reproduce the problem when I tried it briefly. On my Vivaldi stack though, same as Chuck, no, it doesn’t seem to be resolved yet.
Still having this issue myself. I was told dCS is aware of the problem, but the fix is gonna be complicated, and I think that it is not too high on their priority list of things to resolve. The “get a Roon subscription and insert some dead space before song starts” solution does not appeal to me, and it only addresses the symptom really.
In the larger scheme of things, the stellar performance of my Vivaldi stack by far outweighs the inconvenience of missing the first seconds of a DSD song when the previous song is not DSD, as the stack changes configurations. It does kinda irk me though, when I listen to music on my modest studio system (all of which cost less to me than a single component out of the Vivaldi stack) and I don’t have this problem.
I’ve had the Vivaldi about a week, and it is a certified pre-owned piece, so on one hand yes I got a deal. On the other hand the ex-software developer in me is bewildered that this got out the door in the first place, let alone unfixed after 7 years of updates. “We committed to playing music beautifully, but playing all of it wasn’t in the spec” is something my least favorite program manager would probably say. As you can see above going from silence to playing music you should do that flawlessly without error. Play one bitrate, play another bitrate is also a pretty high priority scenario. Ignoring an out of box scenario for this long makes me wonder where else corners were cut. I feel I may be overreacting, but then no, it should just work. I’m sure there are unknown (to me) technical reasons why this isn’t fixed yet but “While clock not locked, buffer” seems easy enough. And I do not like the “just use roon” solution. Most on here say mosaic sounds better, and so far I agree. If I was happy with a compromised solution I would have kept my old dac.
2 Likes
As an aside, that studio / office setup is pretty sweet. I’m a big fan of Sonus Faber from the Franco Serblin days and have Stradivaris now, with Guarneris stuck in storage until I have more room for an office system. My dad got me hooked with ~2000 era Amati Homage.
As to the truncation issue, I agree it is a real issue, arguably a serious one, and should not be a flaw in a shipping product of this expense. However, it is, to borrow David’s words, in fact largely a “symptom issue,” no other flaw preventing operation having been identified. So solving it with a Roon “band-aid” hardly seems like the worst solution in the world. Since anyone can try Roon for free, and at least verify if it solves their truncation issue, it’s a simple test. [No one here has said it, so this is not a pointed comment, but I do find amusing the people who pay tens of thousands of dollars for equipment while objecting to Roon’s piddly annual or lifetime fees.]
As to Roon v. Mosaic/Minim, this has been discussed at length, perhaps excessively, but there is simply zero reason why Mosaic and Roon should sound any different. And I don’t think it is “most on here” who feel that way. It seems to be, in the words of dCS, a fairly even split, and we know what that suggests. We can talk it to death—and I will never argue that you or someone else doesn’t hear it—but as there is no rational, engineering explanation for it, the most likely explanation is some form of psychoacoustic experience in the realm of confirmation or expectation bias. I’ve tried it both ways using Minim on Lumin, NADAC, TotalDAC, DirectStream, MSB Select II, and Vivaldi. Headphones and speakers. My wife, eyes rolled and all, but who can hear the difference clocking makes, hears no difference. Piano-playing daughter, with much younger ears than mine, and who can identify pianos by ear, hears no difference. They are not audiophiles, but they have gently suggested at times my hobby has gotten the better of the music. I would describe that as a symptom of audiophoolishness.
But even if there is a difference, let’s just say there is, to refuse to use Roon to solve this problem sounds a tad short-sighted inasmuch as dCS have not yet solved it (as you note, seven years on). A two-week Roon trial (for those who don’t already have it) or simply living with Roon for a while to remove the annoyance seems like a worthwhile way to test one’s expectations. And enjoy the music. I could not do what David suggests, putting up with the truncation just to avoid using a solution that “doesn’t appeal to me.” Sitting across the table from each other in a pub, I would raise my glass, tap yours and say “you’re bonkers”! 
1 Like
Yup, we all have our preferences and what we’re willing to compromise on. I do have a Roon nucleus, and love the app layout and ease of use compared to Mosaic. It works flawlessly on my iPad but crashes all the time on my phone, over three iterations of iPhone and many more app deletes and reinstalls. As for the Vivaldi, it’s sold as a standalone product, and should perform “play music” without a gap, regardless of price bracket. Imagine Spotify doing it even on their free service. As for roon having no reason to sound a little worse, that might boil down to the “bits are bits” digital argument where Qobuz to Nucleus to Vivaldi shouldn’t deteriorate the Qobuz to Vivaldi path. Which is similar to iMac vs Nucleus, or usb vs Ethernet or the entire digital cable industry. If there’s a chance noise is being introduced, and I have no need for it, snip snip.
Also I was hoping to gift the nucleus solution to my Dad before he rewires his house for whole house music and drag him kicking and screaming into the streaming age. The group feature is really handy.
1 Like
Chuck,
Thanks, it has an Atoll integrated with built in streamer at its core, and a pair of SVS 300 subs along with the SF speakers. I just kinda stumbled on this stuff as I was looking for components for my main rig - bargains that were too good to pass up…
Slainte,
Mitch (aka David)
Greg,
I had around 50 SACDs ripped to DSD and loaded on my NAS and sold them all off, and don’t have a disc spinner in my system. I was using a Puccini SACD/CD player and a UClock for that, but sold them off too. There are only a handful of songs for which I default to the DSDs on my NAS, vs thru Quboz so in terms of frequency of occurrence, this is a small issue for me. I built a pretty comprehensive playlist of DSD songs from my NAS, so once I get thru the first truncated one everything seems to work okay.
On principle, I take exception to the Roon solution—if everyone did that, then the root problem becomes less of an issue for dCS to fix. Yeah, I could afford it, but in reality, if that is what DCS proposes as a solution, then they owe me $14.99 a month or whatever a Roon subscription costs.
I’m perfectly happy with Mosaic, and after having learned most of its nuances, I really don’t wanna have to figure Roon out too. This old dog doesn’t need to learn any new tricks. I thought I’d found the solution in the DAC buffer function available on Mosaic, but that did not work. If it isn’t there to solve this issue then I have no idea what is supposed to do.
It’d be nice if someone from DCS could come up on this thread to shed some light on the solution, or if there is one on the horizon.
Slainte,
Mitch
3 Likes
I hear you David, and certainly agree with you on that principle point. My only notion there was not to let dCS off the hook, but rather to not fail to enjoy the music unimpaired. (I have hundreds of ripped SACDs, and thousands of downloaded DSD tracks. This situation would infuriate me if Mosaic were my only “solution.”)
Had I known about this issue before I purchased, I readily admit I would have waited before making my purchase.
I don’t regret my purchase, but I would have waited to see if a solution was forthcoming, as it’s likely to be fixed in whatever follow-on products are released and as we know when they introduce a new product, dCS stops issuing updates for the old ones.
Here’s the issue with dismissing it as a nit:
At least with the Rossini, as each song that is streamed ends, the RP reverts back to PCM before noting the next is also DSD. That means you can never play a gapless album from the network, it has to be via USB.
The jumping to a new track issue is big, the not being able to play gapless is IMHO even bigger, and yet I don’t know of a single other DAC manufacturer that has this same issue; even my Oppo players can jump between PCM and DSD without truncation.
No one from dCS has commented, but as I’ve mentioned before, it’s almost certainly what I previously stated:
When looking for whether an incoming stream is PCM or DSD, the DAC needs to buffer the incoming data until it knows what format it is and then send the data starting from the start of the buffer on to the DAC, now switched to the proper mode.
I very highly suspect the dCS players have no easy way to buffer that data, which is why the problem cannot be easily solved.