Please solve: Warning: divide by 0

[elanhickler]

d:\_programming\rs-met\libraries\juce\modules\rosic\dynamics\rosic_softkneecompressor.cpp(88): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\math\rosic_complex.cpp(92): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\math\rosic_complex.cpp(92): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\analysis\rosic_waveformdisplaybuffer.cpp(169): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\math\rosic_specialfunctionsreal.cpp(91): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\math\rosic_specialfunctionsreal.cpp(75): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\datastructures\rosic_string.cpp(235): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\datastructures\rosic_string.cpp(233): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\dynamics\rosic_softkneeexpander.h(112): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\math\rosic_complex.cpp(14): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\filters\rosic_prototypedesigner.cpp(844): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\math\rosic_complexfunctions.cpp(146): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\math\rosic_complex.cpp(92): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\filters\rosic_polezeromapper.cpp(501): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\math\rosic_complex.cpp(92): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\filters\rosic_polezeromapper.cpp(476): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\math\rosic_complex.cpp(92): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\filters\rosic_polezeromapper.cpp(447): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\math\rosic_complex.cpp(92): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\filters\rosic_polezeromapper.cpp(346): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\math\rosic_complex.cpp(92): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\filters\rosic_polezeromapper.cpp(234): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\math\rosic_complex.cpp(92): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\math\rosic_polynomialalgorithms.cpp(9): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\math\rosic_complex.cpp(92): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\math\rosic_complexfunctions.cpp(204): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\math\rosic_complex.cpp(92): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\math\rosic_complex.cpp(92): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\math\rosic_complex.cpp(92): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\math\rosic_complex.cpp(92): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\math\rosic_complexfunctions.cpp(203): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\filters\rosic_prototypedesigner.cpp(1009): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\math\rosic_complex.cpp(14): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\filters\rosic_prototypedesigner.cpp(920): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\filters\rosic_prototypedesigner.cpp(864): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\math\rosic_complex.cpp(92): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\filters\rosic_polezeromapper.cpp(501): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\analysis\rosic_waveformdisplaybuffer.cpp(21): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\filters\rosic_prototypedesigner.cpp(944): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\filters\rosic_prototypedesigner.cpp(1033): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\datastructures\rosic_string.cpp(315): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\filters\rosic_prototypedesigner.cpp(1100): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\filters\rosic_filteranalyzer.cpp(184): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\dynamics\rosic_softkneeexpander.h(112): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\filters\rosic_prototypedesigner.cpp(1120): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\dynamics\rosic_softkneeexpander.h(112): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\jura_processors\effects\jura_funcshaper.cpp(175): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\filters\rosic_prototypedesigner.cpp(384): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\filters\rosic_prototypedesigner.cpp(501): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\filters\rosic_prototypedesigner.cpp(501): warning C4723: potential divide by 0
d:\_programming\rs-met\libraries\juce\modules\rosic\filters\rosic_infiniteimpulseresponsedesigner.cpp(233): warning C4723: potential divide by 0

[RobinSchmidt]

hmm - i'm not getting these warnings. this is warning level 4, right? in what project do they occur?

[elanhickler]

spiral generator. couldn't you just not divide by zero? not good practice i think.

[elanhickler]

#define INF (1.0/dummyFunction(0.0))

whyyyyyyy

Edit: do a fresh build (delete build files) to get the warnings.

[RobinSchmidt]

not sure anymore why i did it this way. this is code from about 15 years ago. can probably be replaced with a better expression now, like std::numeric_limits::infinity() - which was not available before c++11 (or was it? hmm...maybe...not sure). ...however, i think, i don't really divide by 0 but only "potentially". it typically occurs, when i set the denominator inside some branches of if-statements and the compiler cannot know, that actual code always takes some branch where the value is set to some nonzero value (the compiler must sometimes assume, that no branch is taken, when in practice, you would always use one of them - or something) i have to look into it - but i'm still not getting these warnings. are you using my latest commit of the library?

[RobinSchmidt]

...and could you use your modified envelope-generator class from your folder, so i/we don't always have to replace the file in my library?

[elanhickler]

...and could you use your modified envelope-generator class from your folder, so i/we don't always have to replace the file in my library?

Don't know how to do that without compile errors. Was planning on eventually wrapping the envelope generator class.

Yes I'm using your latest commit.

[elanhickler]

Just got your LATEST latest commit, I think I am not getting the warnings anymore. Will keep an eye on it.

[RobinSchmidt]

good. i have now set the warning level for my module to 3 (corresponding to "medium" in projucer) in the code. level 4 (high) complains a lot about things that i actually consider non-issues

[RobinSchmidt]

Don't know how to do that without compile errors. Was planning on eventually wrapping the envelope generator class.

hmm...should i try to do that for you? ...for free, of course - i think, it should be a small thing

edit: i mean not the wrapping but using a different file

[RobinSchmidt]

i think, it will just be a matter of renaming the class (maybe use a prefix "se" or "eh" or whatever you like) and adding it to your synth-framework. ...and then using the renamed class instead of the original one in your project

[elanhickler]

yeah you could try it for me. I tried it and got stuck.

[RobinSchmidt]

so, it is the file rosic_EnvelopeGenerator_elanEdit.h/cpp in Shared/EalnSynthLib, right? i think, there are at least two options:

(1) just manually add the files to each jucer project where they are needed (as you have done with ElanSynthLib in SpiralGenerator)

(2) wrap the code into a juce library module - i specifically created the se_framework library module for that purpose. it currently contains only the licensing code, but the idea is that you could put customized (shared) dsp-code, gui-code, utility-code, etc. there, too. i would propose to make a folder dsp in se_framework and put the envelope generator code together with ElanSynthLib. then you don't have to bother with manually adding the code files to your jucer projects. you just add the se_framework module to the module list and drag in all code at once

[elanhickler]

I already tried (1), was hoping you'd give it a shot. Adding to the jucer project is trivial, that's not the issue.

(2) No idea how modules work, I can learn about that another time.

[RobinSchmidt]

done.

[RobinSchmidt]

...oh...just fixed a little bug

[RobinSchmidt]

haha! it's fun to play with the parameters of spiralgenerator. i'm getting funny sounds out of it!

[RobinSchmidt]

...but i think, the funniness of the sounds is mostly due to the pitch detoriating in weird ways when tweaking parameters

[elanhickler]

Pitch deterioration is good for fx. Also you should be using spiral generator with a 2d scope, everything makes a lot more sense. But I plan on addressing pitch deterioration. There are interesting timbres that are desirable but unfortunately tied to the pitch deterioration.

[RobinSchmidt]

yeah - using a 2D scope for sound analysis and design is probably something i should do much more anyway. i like it when synthesized sounds are intrinsically stereo

[elanhickler]

ok thanks for sorting out the envelope issue.

My code is a bit of a mess, I'll be cleaning it up for the next version.

I'd like to release a paid version of Spiral Generator with anything to anything modulation and other features, but I am waiting for you to give me GUI features. Also it needs polyphony.