Waveform FREE > Micro Drum Sampler: What is "Voice Stealing"

[nedreud]

Be gentle with me! My first post and I've only been using Waveform FREE for one day.

I've watched some YoutTube tutorials about the full Drum Sampler in Pro that has choking groups, which I assume you can use to stop playback of the Open Hi-Hat when a Closed Hi-Hat is triggered. This option doesn't appear to be available in the simpler Micro Drum Sampler, but I did spot this "Voice Stealing" option in the 3-bar menu:

MicroDrumSamplerVoiceStealing.png

So, questions:

1. What is "Voice Stealing"?
2. Does this option do anything in the Micro Drum Sampler?
3. Can this somehow be used to mute Open Hi-Hats when a Closed Hi-Hat is triggered, aka, choking?

[FigBug]

Micro drum samples can only play some number of sounds at once. (I think it's 32 or 64, I can't remember).

So, what happens when you play 33rd note and the previous 32 haven't finished. With voice stealing on, the oldest sound gets killed, and a new one starts. With voice stealing off, nothing happens. New note doesn't play.

This feature can't be used to do choke groups.

[nedreud]

Micro drum samples can only play some number of sounds at once. (I think it's 32 or 64, I can't remember).

So, what happens when you play 33rd note and the previous 32 haven't finished. With voice stealing on, the oldest sound gets killed, and a new one starts. With voice stealing off, nothing happens. New note doesn't play.

This feature can't be used to do choke groups.

Brilliant! Thank you very much for the explanation. Makes perfect sense. I searched the internet but couldn't find anything anywhere about "Voice Stealing" in relation to Waveform

[Peter Widdicombe]

In the old days, many synths were limited to 6 or 8 voices, so that was necessary.

[fde101]

The Kurzweil keyboards with their VAST engine have an interesting approach to this. Individual sounds in VAST are built in layers. Each layer consumes one voice of polyphony. If you have 24 voices (as with the original K2000) and are using a sound built from 5 voices, with four notes already playing you are consuming 20 of the 24 voices.

Playing a 5th note requires 5 voices with only 4 available. To get the 5th, VAST looks at the layers that are consumed and uses an algorithm to try to determine which layer is currently contributing the least to the sound - for example if one has an envelope that expired, or its volume is extremely low, etc., then it is more likely to be stolen than one that is playing loudly and is still early in its envelope.

This is a much smarter approach than the usual one of just stealing the "oldest" one currently playing, though maybe a bit overboard for something like a drum sampler.