Sorry for this basic question - I read the manual but it’s not making this totally clear:
If I make a preset with a sampled sound, will that sample be stored within the preset file itself? Meaning, if I wanted to make 20 different sound presets using the same recorded sample, I will end up with 20 independent copies of that sound?
Does this also apply to any preset you make with factory samples - they’re also stored inside of the sound preset file when exporting via Transfer?
The reason why I suspect that the sample is stored in the sound preset is because when I copy a sound preset from the DT2 to the computer, it creates a 3MB large file, meaning it surely includes the sample too. But is it managed differently on the +Drive maybe?
What if I make a sample pack with 20 Sounds all originating from the same one sample - does that mean others would import that as 20 different samples?
Correct, if you mean sound as an audible object as opposed to the elektron word “sound” from digitkt 1.
Yes, correct. They are new samples, just based on your original sample. I’d have to search for it but I’ve seen davemech break it down in the past and this is how they accomplish it, so that it’s stored in a manner which is nondestructive to the original sample.
I think that they’re stored as a DTSND file though which also stores the parameters etc. It’s a container file, of sorts.
Wouldn’t it be pretty quick to try it out ? Change a parameter , save as a new sound … do this 5 times
Copy to pc
Remove old sounds
Copy from pc back to digitakt
Done
Also , I had a look at an sfz file yesterday , loading it into audacity as a raw file brought up the waveforms
If the preset is big , try loading it into an audio app and you’ll likely see a waveform
Thanks @shigginpit. I just tried it now, I made one sample and referenced it from two sound presets. The .dt2pst files both contain a copy of the 2.9MB sample, but there’s still some magic wizardly happening because it seems that if you import a bunch of .dt2pst sounds from the computer, Transfer will ask you if it detects if a sample already exists on your DT2. So, although the preset pack ends up being unnecessarily large, the actual import to someone else’s DT2 is still efficient provided that you don’t click “Import all samples”.
If your pack is 20 dt2pst files, each referencing 3Mb sample, yes, the files will obviously total 60Mb. However, others importing those 20 presets to a single project should find they have 20 presets referencing a single shared file.
This wasn’t true in earlier DT2 firmware, but it was fixed later.
Nice, and jinx! 
Do you know if it looks for the exact file path or if it will just react to the file name itself when determining if it’s a dupe?
Snap
Another question: is Transfer smart enough to update all references to a sample file across projects and presets if you rename it and/or put it in a different file path? Or do you run the risk of breaking all your sound presets if you decide to organize your samples differently later?
Neither, it uses a ‘hash’ which is a unique number generated from the file content. I understand that a file is considered a match when both the file hash and the file length match on the two files. The file name and path are not part of the matching.
That makes total sense and explains the perceived “magic” around how it all just works. Thank you very much for your answers!
Now it’s time to sample the hell out of my synths without worrying too much about the perfect file structure.
As the samples use hash, it also means you can move samples on the digitak and sounds / projects are still ok , they don’t lose the reference . This is a long known thing on digitak 1 … likely on analog rytm etc etc too
Maybe not octatrack though
That should not happen. On loading a project, the project knows all the hashes of the samples, and the DT2 locates the files based on the hashes.
I think there’s some sort of numbering logic which gives the files a unique identity under the internal naming convention.
It may be stored within the container though.
Yep, sounds like it’s a hash based on the file size/contents, which makes sense as that means it’ll be unique yet persistent even though you rename the file.
