Granulate Live Input with Memory Writer and SampleCloud

Question

Hi,

I want to granulate the input mic in real time in the SampleCloud.

I'm using a Memory Writer, then use the record in the SampleCloud, sample field parameter. I checked the FromMemoryWriter Box. 

Everthing works but i have like some digital cyclic artefact in the output and don't find the reason,

wrong parameter in Memory Writer ?

Here is the simple sound .

If someone has already done this or have an idea why there are these artefact.

Thanks

Answer 1

Hi Samy,

Have you tried the example called

direct + Granular Reverb with Frequency (Rate) control on grain playback

You can search for it in the Sound Library using the Sound Browser search for Sound name. It's in

Kyma/Kyma Sound Library/Effects & Processing

This example is designed to allow for rate changes in the granular playback. Please let us know if this works better for what you are doing. Thanks!

Comments

Thanks SSC, i will try this sound to see if it works.

Answer 2

Hey Samy

I just took a look at your Sound and your memory writer is only set to capture 200 ms. As your microphone will be recording into the same buffer of only 200 ms you will most likely get a lot of waveform discontinuity at the recording head. Your default GrainDuration is set to 500 ms when you play the Sound, so just within one grain you play the buffer twice and will most likely get 2 discontinuities (artefacts).

My suggestion would be to increase the capture duration of your memory writer to say 10 seconds and make the TimeIndex and TimeIndexJitter parameters "hot" so you can control them live. See if that helps

Answer 3

Thanks SSC and Andreas,

all works fine, the capture duration was too small.

And the sound direct + Granular Reverb with Frequency (Rate) control on grain playback is great !!!