Hi Spikey,

I fully agree that the description of the panpot implementation in the LA board is confusing enough. The thing is, we can configure values in range 0..14 via a SysEx when writing to the patch temp memory. However, the corresponding port of LA32 has only 3 bits, which permits 8 distinct values. I'd therefore say that we have 15 available panpot values but these will lead to only 8 distinct pan levels. The situation is the same regardless of the timbre structure: those "stereo" structures also only provide for 8 distinct pan positions for each partial in the pair.

Note, this quirk also means that there is no central pan point - we cannot create a partial that would sound in the perfect middle. Of course, many of us felt unhappy about this, and thus we tried to improve the situation a bit in mt32emu with the "nice panning" option, that extends the effective panpot range to all 15 values (see the code).

Hope that helps!

Interesting, thanks for the explanation. So in reality there's only 8 positions, and they differ between the MT-32 and CM-32L. That simplifies things nicely.