[Top][All Lists]

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

Re: [fluid-dev] Fluidsynth instrument attenuation

From: Edenyard
Subject: Re: [fluid-dev] Fluidsynth instrument attenuation
Date: Wed, 21 Jul 2010 16:32:37 +0100
User-agent: Mozilla/5.0 (X11; U; Linux i686; en-GB; rv: Gecko/20100713 Thunderbird/3.1.1

Thanks for clarifying this, Chris. Now I know for sure what the correspondence is, I can safely re-tweak my ranks for the intended balance.


On 21/07/10 15:30, S. Christian Collins wrote:
Gerald, the 1:0.4 correspondence is the "correct" behavior for
compatibility with the thousands of SoundFonts that have been created
expecting this behavior. This is what I proposed, so as of FluidSynth
1.0.9, 1 dB of attenuation specified at the instrument/preset level will
only give you 0.4 dB attenuation. I believe this was the case with 1.0.8
and earlier as well, but I had to be specific about this unique case in
my proposal since FluidSynth previously calculated all dB scales at 0.4
as well (for modulators, etc.), when only the instrument/preset
attenuation should be calculated at this scale.

If FluidSynth has changed this behavior since, I don't know, since I
haven't really looked closely at it in a while.


On 07/21/2010 03:19 AM, Edenyard wrote:
Many thanks for your response, Chris. The PDF was very interesting,
even if I don't fully understand some of the programming type content
of it.

Do you know whether your proposed modifications were actually
implemented in FS? If so, can one now get a 1:1 correspondence between
dBs of attenuation set versus dBs of attenuation actually obtained?

Finally (bearing in mind this attenuation matter and the proposed use
with Jorgan, etc.), would you have any recommendation for which
version of FS I ought to be using to update my system? There are mixed
messages floating around about the success (or otherwise) of the
various versions/updates of FS and I'm a bit confused over which
version would be most reliable. I'm still on version 1.0.6 which has
worked for my system with Jorgan for 5 years, albeit with various
'artefacts' in evidence!

Many thanks for your help!


On 21/07/10 01:11, S. Christian Collins wrote:
Preset/Instrument level attenuation attenuates 0.4 dB for every 1 dB
that you specify (according to the values reported by Vienna Soundfont
Studio 2.xx). The reason for this behavior is compatibility... this is
how Creative/E-MU designed their synth engines, and the thousands of
SoundFonts out there expect this behavior. You can read more details
about this in this PDF file
<http://www.schristiancollins.com/temp/20080817_changes.pdf>, which was
my proposed list of changes for FluidSynth 1.0.9.


On 07/19/2010 11:28 AM, Edenyard wrote:
I'd be grateful if someone here could kindly unravel a bit of a
mystery in Fluid Synth for me regarding instrument attenuation.

I've been making a new organ soundfont using samples from the
excellent new Silver Octopus romantic set and I'm now tryingto get the
right balance between the various ranks by adjusting the instrument
attenuation for each instrument. I've found quite a bit of conflicting
information in different places about the relationship between what
the attenuation setting is set at and what actual attenuation Fluid
Synth (and others) actually apply to the instrument.

Also, it seems as though the relationship has changed sometime over
the last few releases. I've tried googling and get presented with
quite a lot of programming type stuff that unfortunately leaves me
rather in the dark as a non-programmer.

So - in a nutshell, is 10dB of attenuation as set going to give me
10dB of actual attenuation for my instrument, or is it something else?
Also, which version of FS should I be using to get the right answers?

In case it matters, I'm using Jorgan in conjunction with FS 1.0.6 on a
Linux Slackware system running a 2.4 kernel - although this may get
updated as part of the mods I'm doing.

Sorry if this question is really trivial, but it's stumping me at
present! Thanks in anticipation for any help.


fluid-dev mailing list

reply via email to

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