I think the idea is really good and I have also requested something similar in the survey.
At the moment, the sounds are hard linked to a pattern and I totally get the idea of staying in a pattern, improvising on it and then changing a pattern for something completely new.
However, I find it difficult to realize longer (live) arrangements. At first I thought that having the option to switch the pattern while keeping the sounds invariant would do the trick, but soon realized that this also may be too limiting.
For instance, maybe one would like to keep a bassdrum sequence on track 1 going and would only like to switch the bass sequence on track 2.
Ideally, each track in a pattern would have multiple, e.g. 8 instances of sequencer information that could be freely mixed and matched during a live performance, for instance:
Pattern 1
Track_1 Track_2 Track_3 ... Track_6
TS11 TS12 TS13 ... TS16
... ... ... TSij ...
TS81 TS82 TS83 ... TS86
TSij is basically the sequencer information with parameter locks and everything, just as it is now. In addition, one could use the same track sound with different sequences and sound locks would still be functional also. With this setup, one could mix and match the sequences, for instance:
Performance
Track_1 Track_2 Track_3 ... Track_6
TS11 TS12 Mute ... Mute
TS11 TS22 Mute ... TS16
...
Here, track 1 is played by the information in sequence TS11 twice, track 2 is played by TS12 followed by TS22, and so on. But, how to deal with polymeters? For instance, TS11 could be 64 steps and TS12 and TS22 could be 18 steps long. Personally, I would just chain them and let the sequences play out completely, but things could get complicated quite quickly, which may be desired, by the way (phasing sequences). Another question is how to handle pattern switching in case of polymetric patterns?
A “song mode” could be realized by simply having a table that specifies the order in which the sequences are played within a pattern.
However, looking at the concept, I can understand why Elektron is a bit reluctant with implementing something like this.
- In the above example, the sequencer information would be multiplied by a factor of 8 per track, adding to the fundamental structure indeed.
- It is not clear how an ergonomic live switching of sequences (per track) and patterns would work with the current user interface.
- Keeping the back compatibility with the current state is important, so that the current tracks of the users are still playable.
- People seem to like the sequencer as it is. For this reason it may be a bit risky to alter the workflow.
- …
At first sight, an easy solution could indeed be to use an external sequencer, however, one would lose all the goodies of the parameter and sound locks.
After all, I still think that having an option to be more creative with the sequences is a really good idea, potentially opening a lot of new performance and compositional options. However, I would also make sure that the current workflow is still functional. In the above concept this could be the case, because one possible configuration is simply
Pattern 1
Track_1 Track_2 Track_3 ... Track_6
TS11 TS12 TS13 ... TS16