Organic puresine preset produces no sound by default

[BHSPitMonkey]

The Organic puresine preset (data/presets/Organic/puresine.xpf) appears to produce no sound out-of-the-box. It seems that with the Organic plugin, the first oscillator's volume setting has a dominating effect on the overall instrument volume. In puresine.xpf, the volume for the first oscillator is set to zero (vol0="0"), so nothing is heard without changing this oscillator's volume first.

(Anecdote: If I turn the first oscillator's volume all the way up, it sounds exactly like I remember this preset sounding a few years ago.)

My take is this: either

• the preset is correct and there is a bug in the Organic plugin, or
• the plugin is correct and this particular value was not originally meant to be zero.

[diizy]

On 04/10/2014 04:38 AM, Stephen Eisenhauer wrote:

The Organic puresine preset (|data/presets/Organic/puresine.xpf|)
appears to produce no sound out-of-the-box. It seems that with the
Organic plugin, the first oscillator's volume setting has a dominating
effect on the overall instrument volume. In puresine.xpf, the volume
for the first oscillator is set to zero (|vol0="0"|), so nothing is
heard without changing this oscillator's volume first.

(Anecdote: If I turn the first oscillator's volume all the way up, it
sounds exactly like I remember this preset sounding a few years ago.)

My take is this: either

• the preset is correct and there is a bug in the Organic plugin
• the plugin is correct and this particular value was not originally
  meant to be zero

Probably it's #1. I don't think most of the presets have been touched in
a long time...

[softrabbit]

IIRC those standard, shared oscillators can have a "slave" oscillator attached, and when mixing they mix the signal from that and then apply volume, and that's how Organic works.

So instead of v[0]*wave[0]+...+v[n]*wave[n] you get v[0]*(wave[0]+v[1]*(wave[1]+...v[n]*wave[n]))))))))

Fixing this in the oscillator class(es), if it's even possible, will most likely change the sound of at least TripleOsc, too.

[diizy]

On 04/10/2014 10:15 AM, Raine M. Ekman wrote:

IIRC those standard, shared oscillators can have a "slave" oscillator
attached, and when mixing they mix the signal from that and then apply
volume, and that's how Organic works.

So instead of |v[0]wave[0]+...+v[n]wave[n]| you get
|v[0](wave[0]+v[1](wave[1]+...v[n]*wave[n]))))))))|

Fixing this in the oscillator class(es), if it's even possible, will
most likely change the sound of at least TripleOsc, too.

—

I don't think it's a problem with the oscillator class, because even in
triple osc, you can have osc 2 audible while osc 1 is silent.

Mixing is a simple addition of each oscillator, and their volumes should
be independent of each other.

My guess is that Organic is simply using the osc class incorrectly...
I'll have to take a look at that later.

[softrabbit]

Interesting, looks like both Organic and Triple have their oscillators chained together in a pretty much identical way. It was probably the chaining that in the past has lead me to think this behavior was intentional and not a bug.

Oscillator::updateMix does look OK, which would fit well with mixing working fine in Triple. Which means... what? That those oscillators created in organicInstrument::playNote get created with the wrong algorithm (probably AM) after all?

[diizy]

On 04/10/2014 02:39 PM, Raine M. Ekman wrote:

Interesting, looks like both Organic and Triple have their oscillators
chained together in a pretty much identical way. It was probably the
chaining that in the past has lead me to think this behavior was
intentional and not a bug.

|Oscillator::updateMix| does look OK, which would fit well with mixing
working fine in Triple. Which means... what? That those oscillators
created in |organicInstrument::playNote| get created with the wrong
algorithm (probably AM) after all?

Nope, that doesn't seem to be the problem either:

m_modulationAlgo( Oscillator::SignalMix ),

is how the modulation is initialized.

There's some kind of other screwy thing going on...

[diizy]

Fixed by #600