Re: [fluid-dev] FluidSynth 1.0.6 "Music to my ears"

[Josh Green]

On Mon, 2005-06-13 at 17:44 +0100, Gerald Pye wrote:
> Josh Green wrote:
> > I'm proud to announce that FluidSynth 1.0.6 "Music to my ears" is now
> > available for download.  Of particular note is that many synthesis bugs
> > have been fixed....
> 
>       Many thanks, Josh, for that!
> 
>       I was hoping that the new version would solve another problem that I'm 
> having, 
> but it hasn't - and I hasten to add that I don't believe that the problem I'm 
> having is attributable to Fluid in any case. But I'd appreciate any ideas as 
> to 
> how to solve it:
> 
>       I've assembled myself a soundfont (using Swami 0.9.2) which is made up 
> from 
> .WAV sound samples from two different sources. Unfortunately, it seems that 
> the 
> samples from one source were recorded at about 10% of the level of those from 
> the other source and I'm having great difficulty in working out how to 
> balance 
> them - they ought to sound roughly the same level in the finished soundfont, 
> but 
> obviously they don't!
> 
>       Using Swami, I've tried adjusting the attenuation ('Vol Envelope' tab) 
> of the 
> presets that are too quiet but setting the attenuation to negative levels 
> (i.e., 
> increasing the gain) doesn't seem to have any effect. Selecting positive 
> values 
> of attenuation makes the preset quieter, as expected.
> 
>       Since the number of quiet presets is much smaller than the number of 
> loud ones, 
> it makes sense (or so I thought!) to bring the quiet ones up to the level of 
> the 
> loud ones but I don't seem to be able to achieve that. I also don't want 
> everything to end up down in the noise and quantising distortion by running 
> at 
> too low a level.
> 
>       Does anyone have any useful pointers as to how I can fix this problem? 
> I 
> appreciate that this isn't necessarily anything to do with Fluid as such, but 
> Fluid is the one thing that's making my project at all possible so I hope I'm 
> not offending anyone by posting this query here!
> 
>       Cheers,
> 
>               Gerald.
> 

Hello, Gerald.

Not a problem at all posting that question here, seems like an
appropriate place.  The answer is that SoundFont files don't have
negative attenuation (you may see something like this at the Preset
level, but when it actually gets applied at the instrument layer, only
positive attenuation is allowed).  With DLS files this is not the case,
but thats another story.  So you either need to attenuate all other
sounds to the level of the quietest one, or better would be to normalize
all sounds with an external audio editor.  Currently Swami 0.9.x is
lacking in some features that would make this process much easier.  In
fact I'm not sure if sample exporting even works with Swami 0.9.2 (its
essentially not being developed anymore, since I've been working on the
new version).  I probably should release another version of 0.9.x with
some of these features in there, since it is the version that is being
used.  Hope that helps.
        Josh Green

signature.asc
Description: This is a digitally signed message part