How to extend realistic sound textures with MGranularMB

Official support for: meldaproduction.com
Post Reply New Topic
RELATED
PRODUCTS

Post

I'm looking for a method to extend a sound texture, that is to produce multiple random loops or slices of a texture-like audio, recombined in a random order with continuous little variations, in order to extend its length at will (like producing three minutes of a water stream sound from an eight-seconds sample of it, without audible repetitions; instead of a water stream, it could be an applause, a bubbling water sound, a birdsongs sound landscape, a voice muttering, a typewriter sound… sound textures of various kinds). I want this to be done in a realistic way, not simply by looping (it would cause audible repetitions), not by unrealistic stretching or slowing down, without altering the transients and without transposing the pitch. Would it be possible to achieve this with MGranularMB, which I have?

Post

Interesting idea. Sorry for not really answering your question, but my first thought regarding this would be: with MGranularMB (which is a really great plugin, btw!) you have to keep in mind this situation: let's say you have 8 seconds of audio you are somehow "looping" in a clever way with the plugin to have a longer, still realistic sounding result. In a DAW you always had to play from inside the 8 seconds loop to also hear the result later in the arrangement of the DAW. Or you could / should render / freeze it, though. Just some quick thought, hehe.

I would rather cut and blend such a short recording myself and try to expand it to be longer. But I am looking forward to what kind of ideas other people might come up with!
System: Win 10 64 bit / i9 9900K (8x 3.6 GHz) / 16 GB DDR4-3200 RAM / 1TB M.2 SSD + 2x 500 GB SSD / RME Babyface / Reaper

Tagirijus.de

Post

You could just loop the original sample to transform it with MGranular (hopefully) and record the output with MRecorder.

Post

Ah this would be your plan, I see. I first thought that your idea is to let MGranularMB basically "resynthesize" the missing audio material. But your idea would be to constantly messing with the incoming audio signal and chopping it. Not sure, how authentic the new material could be with MGranularMB in such a case. Hm.
System: Win 10 64 bit / i9 9900K (8x 3.6 GHz) / 16 GB DDR4-3200 RAM / 1TB M.2 SSD + 2x 500 GB SSD / RME Babyface / Reaper

Tagirijus.de

Post

Tagirijus wrote: Tue Feb 09, 2021 9:00 am I first thought that your idea is to let MGranularMB basically "resynthesize" the missing audio material.
Yes, that would be a great solution, even better. It would probably require to use a feedback feature (and some sort of delay, probably). I do not know MGranular enough to understand if this is possible.

I know that there are at least two other softwares that allow for doing what I want, but I would like to understand if it is possible with a Melda product, so MGranular came to my mind.

Post

Speaking of feedback in combination with MGranular I would think of trying it in MXXX. There you can set up feedbacks etc.
System: Win 10 64 bit / i9 9900K (8x 3.6 GHz) / 16 GB DDR4-3200 RAM / 1TB M.2 SSD + 2x 500 GB SSD / RME Babyface / Reaper

Tagirijus.de

Post

A granular sampler would be better for the job, so that it plays random positions, then it should sound relativ clean.
do you have msoundfactory? Then you can try this module.:
viewtopic.php?f=138&t=541705
long hold time, and only random position.

Post

No, I don't own MSoundFactory (yet), but trying to use MGranularMB as a sort of granular sampler playing at random positions was exactly my idea, and the other two softwares I mentioned do exactly that.

Post

The problem is, that MGranularMB first cuts the audio into grains and then rearrange in time. So you can create kind of random, but with overlaps and gates or more density. It cannot retrieve the audio in different places. You can of course use all the voices to create a thick cloud from the grains, but doesn't sound very nice without pitch changes in my experience.

Post

Faiky wrote: Tue Feb 09, 2021 1:39 pm The problem is, that MGranularMB first cuts the audio into grains and then rearrange in time. So you can create kind of random, but with overlaps and gates or more density. It cannot retrieve the audio in different places. You can of course use all the voices to create a thick cloud from the grains, but doesn't sound very nice without pitch changes in my experience.
Thank you, very good answer. This was more or less exactly what I was hoping to know. Maybe it is not the right tool, in my case (fortunately, I have another option to experiment with). Thank you very much!

Post Reply

Return to “MeldaProduction”