Currently Vuo has many different data types, and there are many converters already available within Vuo. However there is no way to convert between VuoReal and VuoAudioSamples.
This node would simply fill the audio buffer with the value of VuoReal. This would simplify future nodes that use Audio Samples, as we wouldn’t need special nodes for maths functions between Audio Buffers and Real numbers.
Considering that the current audio buffer is quite small at 512 samples- I do not see this incurring too much overhead.
One issue that may need to be solved is the maximum refresh rate of such a node, and also if it would be beneficial or even possible to have a node that converts from VuoAudioSamples → VuoReal?
There would obviously have to be a written warning about utilising numbers above 1, but there are many instances where this could be needed anyway.
I’ve opened this feature request for community voting.
What should happen if you give this node a list with fewer or more than the required number of samples (currently 512)? I’m guessing just fill in zero values, or truncate, respectively? (That would sound really bad, but I haven’t thought of any better options.)
possible to have a node that converts from VuoAudioSamples → VuoReal?
If we have a mode port on the node can it still be used as a type converter?
No, type converters are 1-in/1-out, so adding a mode port would preclude it from being a type converter.
So maybe there should be 2 different nodes:
a simple 1-in/1-out VuoReal->VuoAudioSamples type converter — produces a sample buffer of exactly the size of the input list, leaving downstream nodes to cope with any discrepancies
a VuoAudioSamples->VuoAudioSamples scaling node — lets you specify the desired output buffer size and conversion type
I think deep down I really just want it to be a wave table synth. (Which is really exactly this with a frequency input).
Maybe type converter is just too simple a node for this feature.
Be great to have it function as a wave table with variable table size. That way you could feed data sets into it like we do with “make parametric mesh”. Especially good for scanning synthesis or hand drawing waveforms! I am guessing this would be quite fun to work on! ;-)
The other side of this was simply to output a VuoReal as an audiobuffer with Linear Interpolation. This would allow other nodes that use audiobuffers to accept VuoReal values.