Adds a phase-aligned formant generator, a cosine wavetable, and a bell curve wavetable

[chilichesdog]

No description provided.

[tedmoore]

Can you explain why bell_bWrap is False? For synthesis it seems like it would always be True? Also, why is this a Parameter that could be set by a user? Why would a user set it to True instead of False?

[chilichesdog]

Miller Puckette's design in Pure Data only reads the second half of the bell curve table. It uses tabread4~ which does not wrap - instead it sends the first or last value in the table (depending on which direction the index value exceeds the bounds), which in the bell curve, is negligible (1e-7). In this case in MMMAudio, setting bWrap to False by default has the same effect, sending 0 if the index is out of range. I found in testing that the user can achieve different sounds by setting bWrap to true, thus reading the entire bell curve many times continuously rather than reading half of it only at the beginning and end of the sinusoidal phase.

[tedmoore]

Awesome. An example file demonstrating these options would be excellent. Also, a very brief description (basically what you've put above) can go in the mojo docstrings.

[tedmoore]

There is a lot of duplicated code in either path of this if/else. Can it be tightened up to make the struct more maintainable?

[tedmoore]

Is the buffers list allocated to have 6 elements? I don't see where that happens?

[spluta]

Thinking about this a bit, we need to make some changes to make this merge-able:

1. I don't want to change the basic waveforms, so don't add the waveforms to the OscBuffers.
2. a cos wave is just a sin wave with a 1/4 phase offset, so for the cos lookups just offset the lookup point by OscBuffersSize/4. we don't need a separate table.
3. the gaussian window is generally useful and can be added to Windows in Windows_Module.mojo but not basic oscillators
4. the sine window already exists in the Windows struct, so use that instead

for both the gaussian window and the sine window, you can get the value from the window using the win_env function, which handles the lookup for you.

[spluta]

One more thing. The cos waves should definitely use Osc instead of implementing their own lookup. To make this work, just give the Osc.next function a freq of 0 and a phase_offset value of your cosN_phase plus 0.25 (to make it a cos).

This will make this much more maintainable.