First time here? Check out the FAQ!
x

Recording to and playing back from disk sound portions in MultiGrid/Timeline during performance

0 votes
169 views
Is there a practical way to record during a performance arbitrary portions of sound to disk and using them afterwards as input to processing from within a MultiGrid or Timeline? I mean not recording a complete session, but something like sort of several extended loopings.

Using DiskRecorder and DiskPlayer with Gates in MultiGrid tracks doesn't work immediately as if I would have expected (for example: when compiling the MultiGrid there is a count down, even if I don't want to start the recording at that moment). Any hints are highly appreciated.
asked Oct 31, 2017 in Sound Design by explodingtickets (Practitioner) (740 points)

3 Answers

+1 vote
 
Best answer
In the current version, it's not possible to record to and play back from the same disk file within a compiled Sound (where compiled Sound includes Timelines and Multigrids). You could, however, do this with a Tool or a CompiledSoundGrid.
answered Nov 1, 2017 by ssc (Savant) (79,620 points)
selected Nov 3, 2017 by explodingtickets
understood. Interesting idea to use a CompiledSoundGrid ...
0 votes
Have you tried using the Memory Writer prototype? It writes to the Paca(rana) RAM and not your disk but other than that it might do just what you're looking for. You can place several instances of it anywhere you want in a timeline, specify the length you want to record and access those recordings from other sounds for further processing.
answered Oct 31, 2017 by will-klingenmeier (Practitioner) (490 points)
Thanks for your answer, but unfortunately the solution is more difficult. I intentionally stated *to disk* and don't mean a looper machine; the sound portions I'm thinking about may last from one to ten minutes: definitely too long for RAM (and of course I know and already use extensively MemoryWriters ...). Furthermore excessive usage of MemoryWriters in MultiGrids with multiple tracks of different processing sounds leads often to out-of-realtime or scheduling problems. After recording the disk file portions should act later as pure inputs to the tracks/processing sounds within the same interactive performance (somehow like transformed memory echoes within the same piece).
0 votes
Hi!

I'm not sure if I get your description right. But could you try to do this by using a mixer between your computer and kyma. Record the live performance on disk with Kyma, and then replay it with another program. Route it to the input of Kyma and add Live processing through Audio Input prototype?  

If musical timing is very critical then this would not help.
answered Nov 2, 2017 by anssi-laiho (Practitioner) (690 points)
Did you try this? Sounds complicated, causing increased efforts ... Nevertheless I guess one same problem remains: you must finish the Kyma sound (or Timeline or MultiGrid), which is used for recording, in order to get the recording file closed, before you can open it in another program (compared to this the proposal with CompiledSoundGrid seems to be easier ...). Roughly technically spoken: for my intended usage we would need something like a pipe instead of a file ...
...