|Subject:||Re: [fluid-dev] Override envelope release time?|
|Date:||Thu, 12 Apr 2018 14:11:38 +0200 (CEST)|
Hello Mark & Tom
>Is this use-case worth to keep absolute generators?
A part for experimenting i don't see any use case for "absolute" generators.
>Or is there any other way of achieving what Mark is asking for?
I'am sorry i dont see an other way.
Musically speaking if the goal is to have a very fast release to satisfy an "aggressive staccato playing", i think that
that FluidSynth should bring the minimum release to 10 ms (rather then 16 ms). A limit below 10 ms doesn't seem to give any difference
and produces pops and clics ( because the internal resolution is around 1 ms (with sample rate at 44100Hz) ).
> Message du 11/04/18 15:45
> De : "Mark Raynsford" <address@hidden>
> A : "Tom M." <address@hidden>
> Copie à : address@hidden
> Objet : Re: [fluid-dev] Override envelope release time?
> On 2018-04-11T14:50:40 +0200
> "Tom M." <address@hidden> wrote:
> > > I'd like to set the R time to 0 globally so that notes immediately cut
> > > of when their respective note-off event is received.
> > The SF2 spec requires a minimum release time of around 16 ms. That's how "immediate" you can get.
> Is that "if you specify 0 in a soundfont, fluidsynth will translate
> that to a 16ms release" or is that "fluidsynth is not capable of
> producing less than a 16ms release"? I mean, I'm happy to violate the
> spec for my own purposes. :)
> I'm basically trying to get a very aggressive staccato sound, and
> that's sort of difficult to do if instruments insist on tailing off in
> a soft (and arguably realistic) way.
> I can probably tolerate a 16ms release in any case.
> > > Is it possible?
> > Well... yes it should be, but in fact it is not.
> > What you're requesting is a job for fluid_synth_set_gen2() and something like this should do it:
> > foreach (c : midichannel)
> > fluid_synth_set_gen2(synth, c, GEN_VOLENVRELEASE, 0, TRUE, TRUE);
> > This sets the generator responsible for volume release to an absolute value of 0 for each midi channel. However: This feature never worked as it's implementation is incomplete. Therefore it was concluded to remove it upon next major version bump: https://lists.nongnu.org/archive/html/fluid-dev/2017-12/msg00004.html
> > @Marcus + JJC: Is this use-case worth to keep absolute generators? Or is there any other way of achieving what Mark is asking for?
> Mark Raynsford | http://www.io7m.com
> fluid-dev mailing list
> [ OpenPGP digital signature (0.2 Ko) ]
|[Prev in Thread]||Current Thread||[Next in Thread]|