Audio File to Buffer

[sunep]

I am trying to do some shuffling of an audio file, it seems like there is no elegant way of doing this, so I would propose that a wave file to buffer node of some sort is made.

It would be even cooler if it is possible to shuffle these buffers around, treating the samples as a spread.

I might have missed this one if it is already there.

[sunep]

To clarify a bit what I want to do, I would like to take a video and shuffle it around according to different parameters, it could be image analysis or audio analysis so I would like to take each frame of the video with the sound for that frame and play it at arbitrary times together.

currently I do it by rendering the video as an image sequence, that works great with the Player node and the sound with the old wave player node where I spread DoSeek and SeekPosition pins to jump to these times. It does however seem like it is not working optimally and is a bit clumsy, Hence the need for a buffer where I can have random access to any part of a sound instantly.

[tebjan]

ok, that would probably be something like a FileBuffer node or so? to load an audio file into ram and access it with a play head...

[sunep]

Yes, that would be cool.

If it could address and rearrange chunks of the audio buffer, a bit like spread manipulation it would somehow give me this, especially if combined with the ability to save such a buffer.

Perhaps a set of Reader (VAudio Buffer) and Writer (VAudio Buffer) and GetSpread (VAudio Buffer) and SetSpread (VAudio Buffer) nodes that operate with time or sample as index.

If it somehow could operate in an "increment mode" way it would be even cooler. an example is that I have sorted a video according to audio level and rearranged it accordingly, currently I split the video up into a png sequence and a wav file, but the only way to export it is by using fraps to record it. it would be amazing if it was possible to set mainloop to increment mode and export in a nice way. I know vvvv is mainly real time, but I find myself using it in a non-real time manner more and more. and I would love to use audio in that way too.

[sunep]

and perhaps with the ability to overlap and crossfade between the slices. I guess that could be done by multiplying buffer together

[sunep]

Also being able to use spreads to generate a buffer would be cool, eg. if you want to make a pulse