Page 1 of 1

Feature Request for pipes that are 'slow to speak'

Posted: Tue Mar 17, 2009 5:46 pm
by naptastic
Sampling pipe organs is a Hard Problem that is not yet solved, and I have a novel approach I want to try, but AFAICT there is not sampler software that supports what I want to do, and I'm probably not a competent enough coder to write my own.

What I want to do is to be able to select which sample gets played based on how long a note is. Now, I realize that this probably seems impossible given the real-time nature of LS. However, given the nature of organ pipes, I think it will be possible to massage the samples into a format where this would work. The sampler would need to start by playing one file (the "longest note" sample) and be prepared to switch to another file (a "bail out" sample) which would be chosen based on how long the note had already been playing. Processing samples so they would actually work with this sampler format would certainly require some special software with lots of math, and I'm confident I can write that software with the math...

What do y'all think? Am I crazy? Have I articulated my idea sufficiently well? Does LS already do this?

Re: Feature Request for pipes that are 'slow to speak'

Posted: Wed Mar 18, 2009 8:43 am
by ggoodesa
Hi naptastic,
Do you mean having different release portions for whether a note has only been played for 1/4 sec, 1/2 sec, 1 sec, 5 sec, 10 sec, etc. I think that Hauptwerk by Milan Digital has this capability (multiple release samples based on duration of a note). I have not played a Hauptwerk sample set with this capability, but would love LS to have something similar as I am using LS for virtual Theater Organs as well as Classical Organs.

Re: Feature Request for pipes that are 'slow to speak'

Posted: Sun May 10, 2009 6:11 pm
by toplayer2
Graham is correct regarding Hauptwerk. It supports both multiple loops and multiple releases. Assuming you are really talking about altering the speech characteristics, you would need to edit the actual samples. Frankly though, there should be no reason to do this if the goal is to replicate the natural speech of pipes.

To create sample sets with multiple loops, there are only two known ways to accomplish this: use the services of or do it yourself with Adobe Audition. Multiple releases are mainly important for wet (reverberant) samples organs. Hauptwerk does have a facility to match the amplitude of the sustain and release, which should suffice for dry sampled organs. GS has a similar capability, but AFAIK LS does not.

Joe Hardy