Creating a long and complex waveform with a wavetable
Hi, for a specific project, i would like to create a time table with a thousand of numbers going from -10 to +10. I need to enter all of this precise values in a table that will allows me to shape a complex waveform. Then I want to turn it into sounds, to control it and make it evolved.
I'm a beginner in Max. Merely i am looking for some documentations and a line of thinking. Objects like [buffer~] and [table~] seems to be a first step to explore. What do you think about it?
Kind regards
you can successfully use an audio buffer to store and read data, but it does not really work the other way round.
as for the "evolving", this is quite difficult to implement as soon you think about writing arbitrary changes of individual samples into the wavetable while audio is beeing played from it. this required quite some experience and a form of windowing technique for the wavetable generator.
if you can, concentrate on things like linear fades between different wavetables, which can be reachd by simple post processing (here: crossfading)
Hi Roman,
Thats a great idea to create something like a big morphing between several wavetables,
I'm afraid of getting a sound that would be difficult to control in its entirety and immediately, because of its long and multiple nature -- in the sense that you'd have to control the different wavetable parameters of 30 to 40 wavetables at the same time.
Some software packages allow us to draw waveforms directly. Novation Component's Wavetable editor lets you create waveforms that can then be imported into Novation synthesizers such as the Peak or Summit. The precision of the grid is around 120 steps on the abscissa axis.
I'm trying to draw a waveform with a precision of around 800 to 1000 steps.
If you know someone who's good at this sort of thing, I'd be interested in talking to them about it and potentially collaborating on the project.
Another solution would be to write the waveform manually in mathematical software and then have it read by a Picture to Sound algorithm, then use the sound generated by the AI as a new audio source that I would import into Synplant 2, for example, in order to control its characteristics, its 'DNA'.
to generate a waveform at init or by pressing a button you need something which counts, and then run these numbers into a formula.
[peek~] will then write the values into a buffer.
for example cos(n) will create a tone, cos(2%n) the first overtone, n>= 0.5 a square wave and so on.
tri is more complicated... but most useful wavetables contain only sinewaves anyway.