I'm trying to record a spectrum with a memorywriter and then replay it while being able to manipulate in different ways.
First of all I want to play it back slower.
When I record with memorywriter and play it back I use SampleWithTimeIndex and a TimeIndex. At normal speed at works fine, but as I try to slow down, I of course get in trouble with my oscillatorbank expecting the same framesize but my SampleWithTimeIndex of course playing it slower, and probably interpolating. I thought of just ouputting frames of the same size, but looping each frame so to say. I wanted to do it with my TimeIndex looping around at 1 frame at a time, and then being offset 1 frame by another TimeIndex. This way I could do it sort of Phase Vocoder-style (as I've understood it).
Is there a way to do this at sample accuracy? Having a time index offsetting another timeindex.
If my recorded buffer is 100 * ?framelength, one timeindex would be scaled down 1/100 and the other would offset it of it's total path 1/100'th at a time. Maybe using SampleAndHold.
I've attached the sound I was trying to do it with, building on the FrameSync-example I found here in the q&a.
I think I understood FrameSync... But this is of course another aspect of keeping frames in sync, and I might be in over my head.
I don't know if this question is nonsense, as I'm venturing in to the low-level prototypes
Keeping frames in sync