Event Tables, Mouse Area and Morphing

Hi all,

I'm working with event tables. I've used the mouse area module overtop, plus the snapshot module so I can save the table data with snapshots. I've included an ensemble, macro and instrument versions of a simple sequencer. They're all the same, just included them all for ease of testing.

References:
https://www.native-instruments.com/forum/threads/event-tables-and-snapshots.29437/
https://www.native-instruments.com/forum/threads/event-tables-and-snapshots-revisited.227373/

The issue I'm having is with morphing using the built-in "Morph + Random" controls shown at bottom of the snapshots. If I select the snapshots individually, I can recall the values of the event table as expected. But if I try to morph, the table seems to get filled with 0.5 for every index.

Does anyone have a solution to this problem, or is familiar with it? I'd really rather not do morphing using the table's y-index. And I don't really want to have to use the multi-display, but if I have to, I would. Is there a fix?

Thank you for your help!

 

instead of or in addition to an event table, use the selector module with value modules as latches to make a ring buffer

something like this:

 

I'm interested in this approach. Why not use the sequencer module for that type of implementation??

 

it was just something i dreamed up to have a ring buffer for a markov chain thing i was doing. i think you very well could, i think the sequencer works pretty similarly just make sure to trigger the value macro every time you want to change the parameters. also whatever is pictured might not be the best way to do it necessarily, just a general idea

 

BTW, why is there very few (I think only one if I did not missed something) projects using Markov chains ? Long time Mac user could remember the glorious M and Jam Factory programs. I used them a lot in the end of the 80's, I think I never saw so fun and efficient programs to play with music, even these days..
For people who did know these programs, some links I quickly found :
https://www.jstor.org/stable/3680237?seq=1#page_scan_tab_contents
http://www.muzines.co.uk/articles/intelligent-music-jam-factory/255

So, what prevent Reaktor builders to use Markov chains ? Choose your answer :
- nobody cares
- do nor fit well in Reaktor environment
- they're not in the current trends

PS : I finally got back my M licence from Cycling, but unfortunately, there are graphical problems (parts of window not updating, undrawn etc)

 

you've seen mine right?
https://www.native-instruments.com/en/reaktor-community/reaktor-user-library/entry/show/11685/

i got the idea to do one from the forum here, just from seeing people talk about it.

i'm pretty sure that reaktors limitations relative to other options makes things like markov chains more difficult than would be in other applications. reaktor doesn't like midi very much. /

but it is certainly possible. so would a basic neural network as long as it didn't get too fancy. i plan to build one eventually for something not too complicated like parameter optimization

 

In the mid 80s I've own one for Yamaha CX 5M II. Don't know the name.

 

Yes, it’s the one I was talking about. But unfortunately I could not really make something with it. I’ll have to give it a new try.[/QUOTE]

[/QUOTE][/QUOTE]
Ok, we will hope something good come up.