samples disappear from slots on digitakt 1 after using elk-herd sample manager

dimashenme commented

2025-03-06 18:14:03 +01:00

(Migrated from github.com)

I have encountered this issue after moving samples around directories of +drive using elk-herd sample manager. Some of the samples disappear from project slots, and one sees a single tick (') instead of sample name. Interestingly, when one browses the file system (by going into samples -> view +drive) the samples appear as if assigned to a slot, but they do not play in fact (see pictures attached). I was unable to find out which samples in particular were affected. Definitely there were many samples that were moved using elk-herd which were not affected. I am using firmware 1.52 and elk-herd 3.2.1 on chrome 133.0.6943.143, windows 10 machine

I have encountered this issue after moving samples around directories of +drive using elk-herd sample manager. Some of the samples disappear from project slots, and one sees a single tick (') instead of sample name. Interestingly, when one browses the file system (by going into samples -> view +drive) the samples appear as if assigned to a slot, but they do not play in fact (see pictures attached). I was unable to find out which samples in particular were affected. Definitely there were many samples that were moved using elk-herd which were not affected. I am using firmware 1.52 and elk-herd 3.2.1 on chrome 133.0.6943.143, windows 10 machine ![Image](https://github.com/user-attachments/assets/7c15e599-3943-4a81-bf39-e8cbab7b879a) ![Image](https://github.com/user-attachments/assets/b5a8f8db-76fe-478e-86aa-6f3d51ca9bd5)

mzero commented

2025-03-06 19:24:09 +01:00

(Migrated from github.com)

Did you only use elk-herd to move samples around in the +Drive tree? Or did you also load a project into elk-herd and save it back?

Am I correct in seeing that the +Drive tree on the unit thinks the correct sample is assigned to the project slot (cl_C assigned to 5 in this case), but the project's sample pool list shows only a tick mark, and doesn't play it? And can you confirm if the sample should be named cl_C and that the size of 0.01MB seems right?

Did you only use elk-herd to move samples around in the +Drive tree? Or did you also load a project into elk-herd and save it back? Am I correct in seeing that the +Drive tree on the unit thinks the correct sample is assigned to the project slot (cl_C assigned to 5 in this case), but the project's sample pool list shows only a tick mark, and doesn't play it? And can you confirm if the sample should be named cl_C and that the size of 0.01MB seems right?

dimashenme commented

2025-03-07 12:23:03 +01:00

(Migrated from github.com)

I only used elk-herd to move samples around the directory tree, I didn't load projects into elk-herd.

yes, device thinks that cl_C is assigned to slot 5, but slot 5 doesn't contain any sample, displays a tick mark, nor does it play. the size is right, the sample is the right one (I checked against a backup I had).

I only used elk-herd to move samples around the directory tree, I didn't load projects into elk-herd. yes, device thinks that cl_C is assigned to slot 5, but slot 5 doesn't contain any sample, displays a tick mark, nor does it play. the size is right, the sample is the right one (I checked against a backup I had).

mzero commented

2025-03-08 01:16:23 +01:00

(Migrated from github.com)

Turns out I have that sample set!

I tried to reproduce, and failed - that is, it always worked fine.

SO - could look over the detailed steps of how I tried to reproduce and see if anything stands out as different than what you did when you had the problem:

On DT: Create a new project in the project manager, and load it.
Open elk-herd, connect
Using Elk-Herd:
Create a new top-level folder in the +Drive called "tanzbar"
Drag all 55 samples from my hard disk onto this folder
watch them upload
On DT: Using the Samples menu, navigate to the tansbar directory, then Select All, then Load to Proj
On DT: Using Pattern A1, set the 8 tracks to use 8 samples from now filled sample pool: bd1_C, cl_C, hh_C, oh_C, ma_C, cb_C, sd_C, bd2_C. Make a little pattern with 'em. Save the Proj.
In elk-herd: switch to Project tab, load the project, verify samples listed, and that pattern A1 references correct samples
In elk-herd: switch back to Samples tab, make new directories, moves some samples from tanzbar to the new directory (one for basses, one for snares, etc...) each time checking that the sample pool (Samples menu > View RAM) was correct. Make new directories, move those samples again in different ways.

I did step 7 about a dozen times. The samples remain named correctly in the sample pool, and play correctly in the pattern.

Also, long shot question: Is there any chance you have two copies of these samples on your +Drive? (This isn't easy to do, actually as the DT tries hard to de-duplicate samples, but I have seen it!)

Turns out I have that sample set! I tried to reproduce, and failed - that is, it always worked fine. SO - could look over the detailed steps of how I tried to reproduce and see if anything stands out as different than what you did when you had the problem: 1) On DT: Create a new project in the project manager, and load it. 2) Open elk-herd, connect 3) Using Elk-Herd: Create a new top-level folder in the +Drive called "tanzbar" Drag all 55 samples from my hard disk onto this folder watch them upload 4) On DT: Using the Samples menu, navigate to the tansbar directory, then Select All, then Load to Proj 5) On DT: Using Pattern A1, set the 8 tracks to use 8 samples from now filled sample pool: bd1_C, cl_C, hh_C, oh_C, ma_C, cb_C, sd_C, bd2_C. Make a little pattern with 'em. Save the Proj. 6) In elk-herd: switch to Project tab, load the project, verify samples listed, and that pattern A1 references correct samples 7) In elk-herd: switch back to Samples tab, make new directories, moves some samples from tanzbar to the new directory (one for basses, one for snares, etc...) each time checking that the sample pool (Samples menu > View RAM) was correct. Make new directories, move those samples again in different ways. I did step 7 about a dozen times. The samples remain named correctly in the sample pool, and play correctly in the pattern. Also, long shot question: Is there any chance you have two copies of these samples on your +Drive? (This isn't easy to do, actually as the DT tries hard to de-duplicate samples, but I have seen it!)

mzero commented

2025-03-09 19:24:02 +01:00

(Migrated from github.com)

By any chance... did you have this sample already on your machine, then using elk-herd, move it to trash, then empty the trash, then re-upload the sample again?

dimashenme commented

2025-03-11 11:40:03 +01:00

(Migrated from github.com)

I understand that in order to tackle it, you need to reproduce it, but I am struggling to reproduce it in a controlled way as well. I will try to recall all the things I did prior to the issue occurring. I had moved a lot of different samples across +drive tree, they were parts of several projects, and only maybe 10% of moved files were affected by the issue. I did delete some samples by moving them to trash. Another thing: as I was moving files, sometimes there would be several identical files that I moved to the same directory. They were renamed by elk-herd, so eg when I copied a sample "cl_C" to a directory that already contained a sample named "cl_C", it would create "cl_C 1", "cl_C 2" and so on. I deleted these duplicates. I will try to do something like this, maybe this is what is causing the issue

Rows
Columns

samples disappear from slots on digitakt 1 after using elk-herd sample manager #2