so many questions , when i use overbridge it’s USB, i think i made the set up correclty because years ago i wasn’t able to do this , but since i read overbridge manual ( yes i did not the first time ) i changed daw because cubase was a headache , now i 'm on windows and ableton live
Ok I misunderstood the issue, others have confirmed it so you can point Elektron support to this thread. If DT1 is unaffected then it ‘should’ be fixable.
ok, Can you also show the important part? recording in a short transient heavy sample on 16ths, slice it at 16ths and place one below the other like I did in my post further up (helps to increase the clip volume to see the start points clearly) and post that.
Edit: and can anyone who says theirs works correctly please post screenshots showing it via one of the methods outline upthread. “Trust me bro” does not mean anything when I can see with my own eyes that there are issues here.
1 Like
sorry i repacked DT2 ( i said i don’t use anymore) i’m done for today maybe tomorow . i sincerely think it’s the resampling not the clock that shows this jitter
Well, ran a quick test at lunch because y’all got me curious about this.
I tested in the most stable way I know which is to record over USB class compliant mode from my DT1 (don’t have a DT2) into Live without any specific sync. Then line up the first hit and measure the deviation from the downbeats in the DAW to the recorded notes.
I recorded several minutes just to ensure that there is not any drift (based on my assumptions about how this system works there shouldn’t be, but might as well be careful).
Oh, also, I filled every track with trigs but then turned all the tracks down. Except for track one which was playing a kick drum on the downbeats. This is just to give the CPU something to do in case that affects timing accuracy. 
My results were really good. I scrolled around and measured about 10 or 20 spots with the Live cursor. All were within 1 ms. Most downbeats were within 5 samples of perfect. Some spots were slightly ahead of time (by 10–16 samples) and it always seemed to follow a pattern (beats 2 and 4 were often slightly early, which might point to details about how DT timing works, such as processing audio in blocks as mentioned above).
I can’t test more now, but for a quick test (which deliberately eliminates all the common “gotchas” w/r/t running a test like this that I am aware of) I think this is pretty good!
Not disagreeing with the OP at all, just wanted to give one other data point.
2 Likes
By the way, if you want to multitrack audio with stable timing, don’t use the Overbridge plugin and definitely don’t solo one track at a time and hope they line up later. Use the Overbridge app and it’s recorderbox feature, which can multitrack tempo-accurate WAV files without needing to sync to a DAW at all. You then just have to trim carefully.
4 Likes
Can you say more about what this picture is showing? I see “Main” in all the recordings which looks to me like you are trying to line up multiple takes. I think it’s not super surprising that each time a Digitakt plays through a song, the timing might be ever so slightly different.
Folks in the other thread were concerned about phase-aligning bass and drums and stuff like that. If one records a single take via the Overbridge recorder in multi-track mode, that is a way to get multiple tracks that are phase aligned (assuming that DT clock jitter affects all tracks equally, which itself may not be a perfect assumption).
What I’m doing is
There is no warp enabled. Using the same BPM in the DAW as in the DT2
For completeness sake because some seem to think that jitter is introduced during recording
here is a my daw playing a sample every 16th going out via the interface (not overbridge or USB audio, all analog) and into one of the DT2 analog ins.
Then the recorded file transferred out of the DT2 and analyzed in a DAW.
Zero jitter.
2 Likes
Useful to test this for sure but IIRC the OG DT is running on a different platform to the DT2. I remember doing a similar test to this with my OG DT when I owned one and the clock was super tight with no perceptible jitter. I don’t remember seeing anything like the results from the DT2 in this thread.
ah that incorrect statement is rearing it’s ugly head again. see here for discussion on that topic: The new elektron platform(s) - DT2/DN2 vs Tonverk - #58 by sdkboi
TL;DR the DT2 is using the same CPU as the DT1 (and DN1), along with an additional DSP chip. the DN2 doubled up on the CPU used in the DN1 neither of which appear in the TV and the TV is using a completely different OS, Linux rather than bare metal so the statements about the DT2/DN2 having parts of ‘the new platform’ appear to be incorrect.
2 Likes
Have you reported to Elektron yet?
Ah, OK. Thanks for correction.
No I’m still gathering my thoughts I want to fully document this in a way that is clear as possible.
As the confusion in the thread has shown people completely misunderstood what the nature of the problem is, so I want to hash it out and make sure it’s 100% clear what the issue is before writing out a report that I’ll send.
To that end I have updated the OP with images from the device itself and much clearer wording.
If there are more misapprehensions posted I can do my best to correct them. I don’t want to be told an answer to X when the problem being described is Y.
3 Likes
That looks like a very thorough job to me. I didn’t get it first time around.
4 Likes
I don’t think this is a bug, it’s a hardware limitation.
I just recorded my DN2 directly into my DAW and it too has the exact same jitter as the DT2.
The Digitone/takt 2 share a CPU, the same CPU as in the DT1 + DN1 and people had the same issue with the DT1 and jitter as seen from the threads I linked before.
it looks like this is just how the hardware is and you need to keep it in mind when doing exports.
4 Likes
Re-skimmed the OP. I see differences in the screenshots you posted, for sure. But it’s hard to tell how big they are. What sort of a range of jitter are you seeing, in ms?
As I posted above I tested a DT1 and saw “jitter” (I sort of disagree with how you’ve formulated this thread, but using your wording here) of under 1 ms on average.
If the DT2 was worse that would definitely be interesting! But I just can’t tell from what you’ve posted.
1 Like