Nope, it was all immediate (using midi cables to and from cirklon)
Your experience with all tracks as master clock doesnât seem to align with others. Unless your machine is a lemon, something is up.
1 Like
Individual sounds triggered from MIDI notes.
Layered Gens = 6msec latency + 8msec latency for samples.
All Tracks = 26msec latency. Samples in sync.
When compensating for either itâs tight both with VSTi and other hardware
However playing notes by hand at 26msec feels sloppy. No difference in discernible jitter in either mode. Currently Iâm using Layered Gens for analog & acb and my Z4 is doing sample duties.
How Roland did not come up with a memory architecture from the get go to prevent this ugly issue and still have a song mode riddles me.
Seems like a core design requirement, being able to seamlessly transitions patterns/kits. This is not an amateur device. Itâs the flagship of flagships for decades to come.
Also I think Roland should not answer this through support and leave the community guessing. They have some explaining to do.
6 Likes
Yes. Design defect.
Whatâs yaâll take on this?
Nope, itâs exactly how some others have measured it. I now changed to Internal clock + Layered Gens and the jitter improved. My Cirklon and Mulitclock are fluctuating in 0.4 BPM ranges now. Itâs a lot better than Master Clock All tracks that was up to 3 BPM or even more. But it still is not good enough for me for clocking my whole Studio. My Cirklon Donât even show 0.1 fluctuations when getting clocked from the Multiclock.
And in this case (Layered Gens) I can than better use it as a Midi Slave. Because its responsive enough in that mode. Just the Sample Playback is than late , so this mode is basically only good to use if you donât use any samples. But that is one of the points why I bought the machine.
I found this in another forum and it aligns with my experience 100%
All tracks:
Internal sync:
Tight sync between tracks BUT up to 20ms of jitter
External sync: 50ms latency
In both modes, no flamming when same samples placed on non layer tracks.
Layered Gens:
Internal sync:
Only sync between layer tracks, samples on every track fall 8ms behind non sample (including PCM) tracks. About 4-5ms jitter.
External sync: 8ms.
In both modes, flamming when same samples placed on non layer track.
The jitter on all tracks and 8ms samples only delay most certainly should be raised to Roland as bugs since it contradicts the manual and statements from support.
Yeah I wrote that.
3 Likes
However, since then, and further in that thread and elsewhere, people claim that they have 0.x ms of jitter in all tracks mode. Check out the gearspace thread. It doesnât seem to be a consistent behavior across users. Which indicates that there are variables in the configurations and tests.
1 Like
thanks. that helps a lot. so its gonna be master or not at all for me.
I can attest to variables in some respect. Not specific to latency, but how sequencing internally using an Oxi One differs from the Cirklon, which may even differ from a KeyStep, which I believe they used for testing this hardware.
I wonât go into it, but I think this is a very fair assessment.
I would even go as far to assume that they may not be testing this machine with a lot of different pro audio midi devices in general.
That would be a good sign, but I doubt it. Donât want to question other users findings. But I canât imagine there are different machines out there.
Many moons ago when I started with hardware I went for months trying to solve the pain of midi clock sync of my DAW to hardware sequencers . It didnât want go in my head that itâs so hard to sync stuff. I gathered a lot of experience in that regard. Solved it with the multiclock back than.
When I read about super low jitter when synced via DAW clock and usb midi⌠been there. I have a different experience.
For me it looks like a architectural design mistake and I am not sure this can be solved to an acceptable degree.
At the moment the machine is only usable in full force as a standalone device (master clock + all tracks sync mode). Than itâs responsive and Sample Playback is in sync. Itâs even mentioned by Roland to use it in this mode. But it doesnât provide a stable clock this way to sync other devices.
As a midi clock receiver in all tracks mode the device gets sluggish with unacceptable latency of over 50ms. Not cool for a flagship Drummachine. I mean we are not a bit off here. You could never sell an audio interface with that latency and we talking here about a hardware Drummachine. And one of the reasons to buy hardware is to avoid latency challenges.
Layered Gens is acceptable if you donât use samples. I didnât had time to dive deeper into it how the process is to fix this with micro timing or something else. But You would always have in mind to shift things around. Also here in this case : not cool for a flagship Drummachine.
I put it now back as MIDI Slave (all tracks) and see how it goes.
Some Information from Roland about this and what is to expect would be nice. I would be happy if they could improve jitter to <5ms and latency <10ms in Midi Slave + All Tracks Mode. It would still not be stellar in comparison to other devices like the Rytm in this regard, but acceptable. Just at the moment we are far far off from this numbers.
Come Roland. You are so close !
Itâs a shame, because beside the sync catastrophe this thing rocks so hard !
2 Likes
But that is what I am trying to say. Others have reported 0.x ms jitter when used as master.
Can you find the statements ? I mean if some users have a different machine and some batch only have such timing problems than it would be good. Doesnât sound realistic and we have already statements from Roland that there is a challenge. But I take any hope I can get because again : the machine is amazing in all other parts.
Than we could compare testing results.
From the post right above yours. TR master syncing Cirklon.
My guy, Iâm working and at this point itâs thousands of posts spread across three different forums. Iâve done my best to try and organize all the issues but itâs getting difficult to maintain and it does seem that some people have quite contradictory experience than others.
For example: TR master all tracks clocking external sequencers and synths. Itâs not working for you, itâs working for other. Some reporting up to 20ms of jitter, some reporting 0.x ms. There is no control. We donât know the patterns, the gens, etc.
6 Likes
Yeah, maybe it sounds for him ok. Not everybody has such requirements for tightness.
Itâs stays in sync with my Cirklon too. But I feel the âwonkâ. Like I mentioned I went through this topic extensively for many weeks and I feel and hear such high jitter. Kills the groove for me at this numbers. It would be interesting if he could look at BPM numbers on the Cirklon and report in which range it fluctuates.
I think this is unaccptable behaviour and a disrespect towards consumers, I would return itâŚ
Thomann???
Yeah. Itâs a mess. I can only speak from my setup and experience. And a Cirklon and a Multiclock each connected direct with midi cables are a setup that hardly can be a better combination. And I had 20+ similar devices like many Elektrons and also Roland devices. The TR-1000 is miles away with their midi clock performance.
Need to think if I send it back for now and wait till the improvements come or find a way to live with it and hope for improvement. Itâs also important for potential buyers to inform them about the struggle some of us have if they have similar requirements of such device. Itâs tricky. thank you for your investigation and findings too !
5 Likes
I think you need to be open to the idea that youâve got a malfunctioning unit or thereâs some other lurking variable in your setup that is contributing (cables? power? etc)
In addition to lots of testing on my own unit, Iâve been following the threads on here and ModWiggler very closely. No one else, as far as Iâm aware, has reported jitter / BPM issues on the scale that youâre reporting here.
Best of luck getting it sorted.
2 Likes