jitter matrix / fmat ftm library
Hi,
I have been browsing Graham Wakefield waaa~ library to generate envelopes. These envelopes are generated from a [patcherargs], [jit.expr] and [jit.buffer~] objects. So far, I have been using gabor objects from the ftm library, and stored my envelopes in [fmat] objects, which I can interpolate with [ftm.interp]. I'm trying to associate both approaches, but I might be wrong.
I apologize for the number of questions, and for joining the library help patcher that I have modified, but I am currently only exploring objects and am completely new to jitter.
I have several issues :
I am trying to copy the [jit.buffer~] contents into [fmat] matrices.
Changes in the envelopes parameters are not very smooth using [jit.buffer~]. At least, it seems to be the case when displaying the envelope in the waveform~ inspector. I assume I wouldn't have this problem using [fmat]s, but I am not even sure about this.
I am not certain on how to convert the jitter expression in an expression suitable for an [fmat]. I am mostly confused about finding an equivalent parameter to the normalization parameter ( norm[0] ).
Last I don't know if interpolating two envelopes of the same type with different scalars amounts to changing the scalar, since I haven't found out about changing scalars in real time with matrices. I still could use [function], but it is not always the most convenient object in my context.
Any help at any level will be appreciated.
Thanks!
Hi Coralie,
I am not sure if I fully understand your questions but this might help.
There is an object available with ftm to exchange information with jitter. [ftm.jitter]
I had a quick test and it works well only that there is some reformatting needed as the output from [jit.buffer~] is a matrix of one row and n-columns and FTM usually assumes one-dimensional matrixes to be one column and n-rows. The patch below shows this and should get your envelopes into ftm.
I can't test it as U have not installed the waaa objects in my computer but will test later on. One note though is that you can normalize your envelope once you have it in an fmat. Also, if you have the formulas for each window that you need you can generate your fmats directly using [ftm.mess] without the need for jitter.
Hope this is helpful :)
Hi Zancudo,
Thanks a lot. You have quite understood my problem. I do have the formulas and have used them to generate matrices without jitter objects, only, a few of them are a little tricky because they were written for jitter and I have some trouble translating them. I will certainly use the conversion object you mentioned, which I didn't know of. Interpolation with ftm works fine, so I don't think I will need to modify the matrices themselves.
Thanks again.
Yes, the solution in your patch is exactly what I need. Great!
Hi Coralie,
breaking down the expressions into manageable parts can help with the translation. So instead of putting the full expression you take smaller parts and test if the results are the same with both objects. Of course, if you already have that in jitter you only need to pass it to ftm once and you can save your fmats rather than calculate them each time.
:)
Yes, I had noted this too. Well both solutions do work. I'll try to find which one is the handier.
Thanks again :-)