Cyclic modulation of amplitude in the reverb sound

[jimhen3ry]

FluidSynth version

2,2,8

Describe the bug

There is a cyclic modulation of amplitude at a rate of about once each second in the reverb sound. This is deliberate to to cancel unwanted "ringing" (resonant frequency).

Expected behavior

Add reverberation without introducing an unwanted amplitude modulation.

###Related discussion
#1139

Steps to reproduce

Add reverb to the output.

Additional context

This may be an inherent characteristic of the current reverb engine that cannot removed. The solution may be to allow the use of alternate reverb engines as proposed in #1157.

[spessasus]

I think I found a good reverb engine that could replace the current FDN reverb for now to fix this issue: The Dattorro Reverb!

It already has implementations which I believe could be adapted for fluidsynth:

https://github.com/khoin/DattorroReverbNode - JS, public domain (!)

https://github.com/el-visio/dattorro-verb/tree/main - C, MIT License.

Having tested the first one, I can confirm that it is not affected by this issue nor #1496. It seems quite performant too.

The only problem as far as I can tell are different parameters (e.g. no roomSize). They would have to be adapted to work with the fluidsynth API.

@derselbst what do you think?

[derselbst]

I'm still convinced that #1157 is the way to go. I'd need to try Dattorro reverb myself and then see if / how it fits into fluidsynth.

[spessasus]

I didn't say that it wasn't the way to go. I believe that it indeed is the ultimate solution, though given the amount of work it requires, I don't see it getting implemented soon. Dattorro is my suggestion for a temporary solution as a "good enough" implementation that doesn't have the two flaws. It makes fluidsynth's reverb usable while we wait for the proper overhaul of the reverb system.

[derselbst]

Dattorro is my suggestion for a temporary solution

Nothing lasts longer than a temporary. I'm not willing to do that, sry.

[derselbst]

FYI: #1760