fluid-dev
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[fluid-dev] Re : Re: Re : no effects applied using multiple stereo chann


From: CERESA Jean-Jacques ENAC/ENAC
Subject: [fluid-dev] Re : Re: Re : no effects applied using multiple stereo channels
Date: Thu, 09 Apr 2015 18:45:21 +0200

 Le 09/04/15, "Tom M." <address@hidden> a √©crit :
fluidsynth .......-a jack -o synth.audio-channels=3.......
...and applied my reverb settings. Since my soundcard doesn't have 3 outputs..
The  internal FluidSynth structure is relative adaptive to using an audio card with multiple audio ouput.
However, the relationship settings to get this result need explanations:
1.1) The choice of the audio driver XXX (-a XXX). This driver must able to handle the intended audio card.
with multiple audio outputs. This means that this driver need to be instructed (with the help of  specifics "settings").
Unfortunaly in the case of Jack audio driver i am ignorant about Jack capabilities and settings !. I hope
someone can help you about those. The audio path  (audio only) is the following:
 
--->Voices synthesis and effects -->Mixer with N buffers---> Audio driver XXX for "multiple output" --->Audio card with "multiples output"
 
1.2) The internal fluidsynth mixer pass audio to the audio driver input via multiples buffers . The number N
of  buffers is set by -o  synth.audio-channels = N. This is a parameter for the internal fluidSynth mixer .
There is no doubts that this number N need be coherent with the specifics settings discuted in (1).
Particulary if N=3 and the oudio card have only 1 output, it is normal that some audio voices produced
by the mixer will be never heared !
Apart from that, why is there actually a need for mapping all the
reverb to output 0? Wouldn't it be more useful to map them to their
corresponding output? Meaning, according to you example...
2.1)The fundamental reason is that effect unit (reverb) consumes a lot more processing time than the synthesis process
by itself. With effect output routed to only one buffers, we need to execute the effect only 1 time.
With N audio output and effect output routed to to their corresponding audio buffers, we need to execute the effect N
time. As a general rule, in performance situation, it is more appropriate to do software musical synthesis process and
effects process on distinct hardware machine.
However, the way FluidSynth works with effect unit (reverb & chorus) is compliant with Soundfont 2 specifications.
 
jjc
 
 

reply via email to

[Prev in Thread] Current Thread [Next in Thread]