A lot of clicks after sampling [fixed* for next release]

[avantronica]

@krunchr

In every 2nd or 3rd sample I get nasty clicks at the beginning

I have been dabbling with DN inputs and wanted to sample for some exploration … what was throwing me was that the captured samples all seemed to have an artefact close to the beginning, but at what appears to be a uniform position … irrespective of the RLEN i chose - it’s like a short echo length following the capture start point

My simplification was to drone a simple waveform, filtered to low harmonics and i sustained it - it sustains without any change or any modulation and it’s fed (microfreak mono left) into DT which has no usb or midi connections

Afaict every sample i capture has this artefact not long after Fn+ Yes is used to sample

The preview makes it clear, zooming doesn’t quite show it consistently, but there is a small pixel clue in the samples waveform on DT - anyway, exporting it reveals that it is dropping two samples and this is the click that i experience (easier to hear on a low harmonic sustaining tone)

Can anyone compare this method to see if it’s a bug

this is time stamped at the place it keeps appearing when i sample with RLEN at 16 or longer 120bpm

i haven’t counted, but it could be spurious additional samples or just zeroed samples which were absolutely not present at the source

[avantronica]

I have reported this now … I’m fairly confident this is a bug - there are other users discussing clicks and i tried my steps on a new project

It’s not every time, but the majority and always in the same location afaict

somewhere just under 176 ms … it is not extra samples, it is nulled samples in two positions

[krunchr]

Thanks for reporting. I have a feeling that the clicking issue has something to do with a bug I found (and reported) shortly after 1.50 was released. The bug is bpm-related and does only occur when the RLEN parameter is in use.

Sampling length issues at specific BPM values

I resampled 64 steps of a 4/4 kick pattern (threshold setting = 0) using the new RLEN parameter. As you can see on the screenshot, the end of the resulting sample isn’t exactly the end of the pattern. (Note: This is not a graphical problem, you can hear that the resulting sample doesn’t loop correctly) [IMG_6247] EDIT: Issue seems to be related to BPM value and RLEN parameter (see the postings below). Steps to recreate the issue here

[garf]

click here too, at same position of 0.176

[avantronica]

yeah - it’s already been fixed, it was related to the new ux settings which i speculatively called out only a minute before it was confirmed on the DT bugs topic, just a few minutes ago btw

[golddust]

Awesome detective work all - so it’s been fixed but the firmware update has not yet been released?

[avantronica]

no idea of a date as of yet, but if they are not expecting too many bugs from a minimal update they might put it out sooner than normal, or may test other areas potentially affected - either way, the issue is fixed but there is no guarantee of a quick patch - with it affecting the core purpose of the device one might expect it could get released sooner rather than later tho

[orionka]

I would like to sample a few things directly to my DT and there they are, those clicks.
Is there any method you detected where this bug does not occur and one could sample just fine?

[krunchr]

Unfortunately not. Sometimes you hear it, sometimes not. I think it depends on the audio material you’re sampling.

[garf]

Start sampling, then start your source audio a half-second later. It means you have to forget about using the threshold function and do it the old way, but your sample should be clear of the click (always appears at the same time, see posts above). Then you can adjust the sample start, past the click, to your sample starting point.

[orionka]

Ok, makes sense. Thank you!
Edit: actually one could prepare a click a half second before a sample if using the threshold function is important

[prydal]

So incoming update confirmed?? Wohoooo

Hope it also brings the euclidean sewuencer update to the digis…

[MuttBeats]

Resurrecting this thread as I’m also running into the same problem in January 2024. Running latest DT firmware.

The issue arises when sampling from any external source (external L&R or USB from iPad).

I set the sample length to RLEN of 16,32,64 etc instead of max, arm to sample and then play whatever source I’m sampling. About half a second or so into the recorded sample there is a click.

Please note:

1. This only happens when I select RLEN 16,32,64 etc. It does not happen if I set the sample length to max.

2. I don’t think it is a zero-crossing point issue which could happen as the sample source crosses the DT sample recording threshold. The click/drop out does NOT happen at the start of the sample (which is what you’d expect with a zero crossing problem), it happens AFTER the sample has begun playing, roughly half a second to just under a second later. That is not a zero crossing pop.

3. It is not a clipping issue from a too hot input. I’ve tried this with several different sources, all of which have been low volume and well below clipping range.

4. It is 100% repeatable every time whenever a specific RLEN (ie 16,32 etc as opposed to max) is used.

5. This is with the latest firmware that’s available as of January 2024.

[ptomaselli]

@MuttBeats

Running latest DT firmware.

Are you sure you have the latest? For me, 1.51A fixed this issue. (As far as I remember, I am not at my desk right now )

[krunchr]

I installed the 1.51A update a few days ago (because I only recently found out that the A update is available), have sampled quite a lot since then and have not gotten any more clicks.

But the other (tempo-related) sampling bug is still there. I hoped that both sampling bugs were somehow related. But that doesn’t seem to be the case.

[MuttBeats]

Yes, 100% sure. Updated as soon as it came out. I wasn’t doing much sampling in November and December so only just ran into this now. It’s reproducible each time but only when I select a RLEN of 16,32,64 etc. Does not happen with RLEN set to max.

Really odd if this has apparently been fixed in the latest firmware because it seems to be the same bug people were reporting earlier.

[ptomaselli]

Just to be super clear, 1.51A (note the “A”) was released at the end of November and as far as I know, was specifically targeted at fixing this issue. It was released without fanfare so I didn’t even realize it was out until much later.

I’m not an expert, but definitely double check that you have 1.51A at boot up!

[MuttBeats]

OK so my bad. I did NOT have 1.51A installed, I thought I did but it was sitting on my Mac download folder waiting to be installed. Duh!

I have now installed and am delighted to say… all good. RLEN at 16,32 etc is working properly with no clicks.

VERY good news as I sample loops from other gear all the time, it’s a big part of my workflow now.

All we need now is stereo sampling on the Digitakt and my world will rock. Yeah yeah I know… Octatrack

[d4ydream]

@MuttBeats

All we need now is stereo sampling on the Digitakt and my world will rock

You can stereo sample. It’s just summed to mono.

[MuttBeats]

LOL yes… true!