events vs control rate

43 views
Skip to first unread message

Evgeniy Chernyh

unread,
Feb 26, 2015, 1:01:43 PM2/26/15
to axolo...@googlegroups.com
Can you, please, explain this proplem more detailed.
don't understand, why need use of dynamic memory, and why parsing
doesn't it just function calls with one parameter?
so, main problems are feedback loops?

Johannes Taelman

unread,
Feb 26, 2015, 4:36:25 PM2/26/15
to Evgeniy Chernyh, axoloti-dev
I believe lfo's, envelopes etc are good use cases for control rate. If those would be events-at-control rate, you get "hot" and "cold" inlet design philosophy like PD or Max.

One advantage of events is that it avoids re-computation at k-rate of (pure) functions whose input didn't change. However in real-time it is the worst case scenario that causes buffer underruns. Saving control-rate re-computations could optimize the best-case but create higher peaks in dsp load, when simultaneous events happen.

There is some sort of event functions in Axoloti though! Parameters have a "pfunction", this (pure) function can pre-process a (user friendly) parameter value into what is needed in the dsp computation. "pfunctions" aren't exploited at maximum yet.

"pfunctions" are triggered by manual parameter changes, modulation sources, MIDI-mapped parameter changes, preset recalls, or "scripted" parameter changes.

I believe this execution model is very similar as used in commercial fx processors, or synth engines. Though I'm not 100% percent sure, I haven't disassembled any :)

I believe the variety of patches shows this execution model is quite flexible and friendly.

I'm open for discussion, but please be a bit more clear - if you're reacting to some piece of documentation, please quote or reference.
Are you speaking from theory or practice? Are you a PD addict looking for trigger, spigot and moses objects? Or other software/hardware? What are you trying to do?


--
You received this message because you are subscribed to the Google Groups "axoloti-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to axoloti-dev...@googlegroups.com.
To post to this group, send email to axolo...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/axoloti-dev/641d7e7f-f6ca-43c3-8f6c-f94bd9b0c804%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Evgeniy Chernyh

unread,
Feb 27, 2015, 3:09:25 PM2/27/15
to axolo...@googlegroups.com, man1...@gmail.com
i'm just stm-synth-geek) it's interesting for me how it's made, and why it is done this way. i'm not criticizing any of your decisions, sorry if there any misunderstanding, my english far from perfect.
i'm trying to write synth, but those problems of choosing realization methods always stops me.. head explodes, when i think, which one is better. and have not enough time))

пятница, 27 февраля 2015 г., 2:36:25 UTC+5 пользователь Johannes Taelman написал:
Reply all
Reply to author
Forward
0 new messages