posted 12 July 2013 15:11            

WaveSynth.kym

The Sound in question is attached. Actually, there are two sounds in that .kym file, but I'm only concerned with the one called "TEGPicoWaveSeq". If you want to try it out, you will need my wavetable. it can be downloaded from here: http://www.theelectronicgarden.com/Scot/TEGPPGWavetable.aif (probably have to right-click)

What I have is a sequencer-driven wavetable sound based on the analog sequencer prototype. The synth sound is a single-oscillator wavetable job with a custom wavetable I made using my PPG Waveterm and Wave 2.2. It has 128 different cycles of 4096 each. Works very well in a "standalone" context, as you can see by the WaveSynth Sound. This was all loosely based on the WaveMachine sound from this very forum.

What I want to do is to add a row of potentiometers beneath the eight sequencer steps. I want these potentiometers to control the wavetable index - wavesequencing in other words - a new value for each step. Currently, I have the index field reading " !WaveIndex + ( L * 128)", and a stepped fader allowing me to choose the wave from the VCS. I still want to retain the use of the WaveIndex fader. In the context of the sequencer patch, the envelope modulating the Index isn't really much use, so I thought to replace that with a Multisegment Envelope to do my wave sequencing, ala the Wavesequencer example from the forum. But I can't for the life of me figure out what my Oscillator's Index field should read. I've tried variations on the one in the Wavesequencer sound, but so far, no good.

I can't help but think I am missing something obvious. Maybe I shouldn't have given up caffeine this week!

posted 13 July 2013 11:18            

WaveSynth2.kym

I think the attached may be what you were talking about.

The index on each note adds an offset to the overall index value so that each note can have it's own wave.

I found that some where there is some internal smoothing going on in the modules. I think it is the index of the oscillator which is not jumping to the index but taking approx 20ms to get there (sweeping through the waves), or it's could be in the extra values in the sequencer smoothing the value.

So to get round this I've delayed all the other triggers (added overall latency) to make the sweep happen 20ms before the note is triggered. There is a control on the surface that you can play with and tweak to what you think is best.

To hear the problem move just one notes index to say +10 and and put the 20 ms compensation to zero. you will hear a sweep noise at the beginning of that note and also here the revers sweep at the beginning of the note two notes later. Two notes is the next note on the same poly channel.

Also there is an oddity in that the change of velocity (scale in the ADSR) makes a blip in the output even after the the release time has finished. I don't know why this is, but I've put a sample and hold in there to make that change 20 ms later as well.

I hope it makes sense and I hope it was what you were talking about?

Thanks