This is very cool. Thanks Pete, SSC, and all others involved.

On the latency introduced in the long version, is the latency based on the ResponseDuration, processor availability, or a fixed value?

It's a fixed value that depends on the sample rate. -- CarlaScaletti - 07 Jul 2005

The Start/End point controls are a very nice addition and adds a level of playability that is missing in similar implementations. Not sure if possible but it would be great to able to set the end point prior to the start point and have the response scan backwards.

The easiest way to do this would be to use a Sample as the Response and to reverse the direction and Start/End pointers on the Sample (while continuously capturing the Response using an expression of the form: s tick -- CarlaScaletti - 07 Jul 2005

Alternatively you could do it manualy, as long as you use the same trigger for the sample player (playing the responce) and the capture field in the cross filter. The new responce will replace the old, while the sound is still playing. In fact the old input signal will continue to finnish off being prossesed by the old stored responce while the new incoming input signal is being affected by the new incoming responce. Make sure that the responce sample player has it's decay time set to something like 1 day other wise you would have to hold down the trigger for the duration of the capture time. Makeing the Cross filter read the responce backwards (or even at a different rate) after the capture had already happened would be far too prossesor hungry, but you can reload the responce at different rates (and directions) while you're playing (without a brake in the sound) just by making the responce sample player do the changes and then triggering it in.

-- PeteJohnston - 08 Jul 2005

Yeah, this is pretty much what I ended up doing. There is some very interesting and cool effects to be had with these sounds. I'll post some examples here shortly. I am running into some processor assignment issues, but I have to take that up with SSC and/or refine my understanding and ablity to force processor assignments. I generally have the majority of issues when creating processing feedback loops.

-- BenPhenix - 08 Jul 2005

the "c-word" hasn't been used anywhere i can find here.. (convolution!)... how close is this process to convolution?

The concept and usage is similar but there are some important differences. Basically it's an FIR filter implementation that employs some clever tricks invented by Pete. -- CarlaScaletti - 07 Jul 2005

well, an amazing new palette of sound possibilities.. thank you pete, and SSC!!

-- TaylorDeupree - 06 Jul 2005

I'll jump on the "me too" wagon - this is exceptional! Thanks to all involved.

-- BillMeadows - 09 Jul 2005

I have been enjoying this new range of sonic journeys a lot over the last few days - spent 3 days in the studio, and used the new CrossFilter fundtions with Turkish Oud and some manual and automated percussion (used Capy to control motors to play the percussion using Chaos scripts and MIDI controllers)

You can hear some of the outcomes here http://www.activatedspace.com/sounds/AcousticElectronic.html

Thanks to everyone involved, I have been wanting exactly this for some times - woww, CrossFilter and Wacom - one happy bunny :-)

-- GarthPaine - 10 Jul 2005

Could it be that good sound of this algo (it does sound really excellent) is probably due to the FIR impliamentation, rather than using fft's to the same effect?

-- AmbroseField - 11 Jul 2005