First time here? Check out the FAQ!

How to transfer Spectral Data inside a Grid using the FrameSync?

0 votes

Hi there,

I'm stuck here.. I want to transfer spectral control data using a MemoryWriter/Sample pair in the multigrid using the FrameSync. I've attached a little test setup I've made. The problem is I can't get my control signals to be in sync no matter if I trigger the recording and/or playback or not (use the Active toggles to turn on/off triggering).

I want to end up with the possibility to switch between different control signals using a single MemoryWriter/Sample connection without the need to trigger recording & playback. I can't use submixes because I'm already using them to receive the spectral data (output of SpectralAnalysis) which I want to process. Kyma tells me that I'm writing the WaveTable more than once which is basically not true because I'm only writing it for the Sound I've selected, but maybe that's the problem?




asked Mar 23, 2015 in Using Kyma by kymaguy (Virtuoso) (10,560 points)
edited Mar 24, 2015 by kymaguy
Could you route the sample-sources track to another track containing a single MemoryWriter?  And then use the memory (written by the MemoryWriter) in the other tracks?
I just tried, seems like that doesn't help... Or maybe I'm using the FrameSync wrong again? I've edited the question and added another Grid with the MemoryWriter fed by a submix. I hope this will work, I'm currently working on a SpectralGrid which will totally change the way I've manipulated the spectrum until now but for this it is essential that I can send frame-based data in sync...
From these Multigrids it is difficult to tell how you intend to use the spectral data, so it is difficult to give advice.

Would it be possible to switch audio signals in one track, send the audio via a submix to a track that does the spectral processing? In this way, you would not be switching the spectrum and you would not have to be concerned with alignment.

Alternatively, if you are using pre-analyzed audio, would it be possible to use one of the Morph1dSpectrum Sounds to select the spectrum you want to use and have that go directly to the spectral processing?
It's not about the input it's about frame-based control data. The simplified setup is the following:

Track1: Spectral Analysis of the input, only one Sound which is always active
Submix to
Track2: Different Spectral Processing Sounds
Submix to
Track3: Resynthesis, OscillatorBank,...

That already works fine, I can switch between different spectral processes, no problem.

Now I want to have
Track 4: SpectralControl, e.g. a fullramp at framerate (e.g. 256 samp)
Submix to
Track 5: writing Track 4 to Memory which gets read by a Sample module in Track 2

The problem is now that I want to switch between different control signals in Track 4 but they get out of sync...

to say it with Pete's words: I hope that makes sense :)

I tried something different today: Track 1 (Which contains my frame-based control Sounds) gets algorithmically selected using (!Select_Trk1 hasChangedInLast: 768 samp) alignWith: (768 samp tick) for the trigger. 768 is the least common multiple of 48 (1 kHz Capyrate translates to 48 samp using 48kHz samplerate) and 256 (framerate). Seemed like a good idea, but it doesn't work...
Let me know if you come up with something... Thanks!

Please log in or register to answer this question.