Hey,
I’ve noticed that in some Digitakt sample packs - usually with wavetables, single cycle waveforms etc. - there’s a disclaimer that these are only compatible with Digitakt 2. And from what I understand it is related to how the Digitakt 2’s Grid Machine works.
I have a D1 only, but I’m just curious - what’s the difference between D1 Slice and D2 Grid machine?
1 Like
Yeah people, what are the differences?
2 Likes
Edited. Are You happy now?
I don’t understand what you’re inferring from my post but I’m asking for an answer to the same question: What are the differences between the slice machine on DT1 and the grid machine on DT2?
I’m emphasizing the word are with italics as a manner of speech, not in a way that questions your choice of grammar or diction. If you made a syntactic error in your opening statement, I was too busy thinking about the question you asked to notice it so… if it makes you feel better to change your statement that’s cool, but it must have been something minor and didn’t really impact the relevance of your question.
So … yeah, awkward.
4 Likes
Ok then, so back to the original topic - I always assumed D2’s Grid Machine works just like D1’s Slice Machine, so I’m waiting for people to correct me 
It’s been discussed elsewhere but not as its own topic, and I can’t find it at the moment but I think it was confirmed by someone on the Elektron team:
They’re the same, except that, under the hood DT1 splits the slices at the nearest zero crossings only, while DT2 splits samples evenly at all times. On DT1 wavetables will sometimes not work because a zero crossing in a wavetable will force it out of tune slightly—the DT2 doesn’t have that issue with wavetables, though with both they work like the original PPG wavetables and do not interpolate between frames.
7 Likes
Thank You!
Ok, that makes sense for wavetables and why some sample packs are “D2 only”. Although I wonder If the zero crossing isn’t better for drum loops. Would be nice to have both options, I guess thats for the feature request topic.
3 Likes
Do you know if for sampling, does DT2’s zero crossing algorithm split the difference between the nearest L and R zero crossings? Meaning that if there is no common zero crossing, will it truncate to the mean?
1 Like
Ohhh, I just realized: the stereo problem (that there’s not always a zero for L and R simultaneously) is probably the technical reason the DT2 doesn’t split grid mode samples at the nearest zero crossing…
And I don’t think it does zero crossings when sampling either, but I’m less sure about that…
4 Likes
I think it does and that’s the reason that people are complaining that they’re losing between 1-2 ms from the beginning of fixed length gated sample recordings, but I don’t know, I just seem to remember some mention of DT2 having different zero crossing detection that DT1.
Will have to wait for something official or a more clinical style testing otherwise it’s just guessing. Could be a bug I guess, but a bug is usually bad code / conflicting routines / whatever, and thus behaves as such.
1 Like
Is the main difference one is pretty useless in mono and the other is pretty useless in stereo?
2 Likes
Huh, this has happened to me a few times and I was going mad why this happened after I placed all the drum hits at the exact trick that should line up to the grid. Is there any thread or post where this issue (and possible solutions) are discussed?
1 Like
Definitely identified as a problem here: Digitakt II 1.02/1.03 : Bug reports - #34 by zhelnerovichda
Could possibly be happening here too, but no definitive report yet: 128 steps samples are not exactly 128 steps, bug?
2 Likes
Imagine you could set your own start and end points per slice to avoid these issues? /s
Is there a bug report about this? It will help a lot of this is fixed 
Not from me (I’ve not investigated this one yet) … maybe ask @zhelnerovichda who provided the screenshots … or maybe just submit a bug report yourself.
Where?
The first post in the DT2 bugs thread has the details.
1 Like
Sounds like the bug is fixed in firmware 1.10
The length of the recording could be slightly incorrect when using the R.LEN parameter in the SAMPLING menu to record a set number of steps at the current BPM.