category.zd
Offline
Toronto, Canada

Hello,

I am working on a little DIWhy project to build a controller for GenMDM similar to what Look Mum No Computer did here with his Megadrive Synth. For testing and prototyping I have successfully managed to get MIDI signals to the genMDM with the Max 4 Live device provided by Little-scale, and some custom mappings on my Launchkey37. And yes I played notes and got sound on all 10 channels. Didn't try CH3 special mode yet...

Anyway...

The SSG-EG parameters for genMDM seems to have some inconsistent documentation around it. In the MIDI mapping doc, the data range is 16. In the M4L device, the values range from 0-31. From 3 different software-based YM2612 tools I've played with, there are only 8 SSG patterns, and OFF. Deflemask did their own thing and added an Enable tick box with a 0-7 data slider.

It's going to be very tricky to manually figure out when each SSG pattern kicks in, so I'm wondering if anyone's managed to figure out the data values? Worst case I can... try.

Cheers!

MazHoot
Offline
France Eure Et Loire

I found that the old documentation (for previous version) is more understandable. I'm not on the computer now but I can send you. The last time I wanted it it was not available on internet and I asked catskull electronics to send me it.

I think there are also several bugs like : I don't know why you have to activate cc9 for enable fm synthesis on other channels than 1. If I don't there is nothing on default on it.

I can't respond to your question but for now the best controller I used for genMDM was lgpt on raspberry pi with midi. But your project looks promising keep going!!

category.zd
Offline
Toronto, Canada

Hey MazHoot thanks for the response. I managed to find the v101 firmware doc through little-scale's blog but unfortunately it wasn't much help. So I just decided to experiment on my own rig.

I assigned a knob on my MIDI controller to one of the SSG-EG controllers (CC 90 for OP1) and found some interesting results in what the data values actually do with the SSG-EG programming:

  • 0-31: effectively "off"

  • 32-35: pattern 1

  • 36-39: pattern 2

  • 40-43: pattern 3

  • 44-47: pattern 4

  • 48-51: pattern 5

  • 52-55: pattern 6

  • 56-59: pattern 7

  • 60-63: pattern 8

...now add 64 to all of these (yes, ALL) and the whole thing repeats.

Odd, but I'm guessing it has something to do with how the SSG-EG register has an enable bit in addition to values representing the patterns. No matter this was a useful exercise in figuring out what data values to send on that controller.

I've also encountered some odd bugs but it seems they don't always manifest. My main gripe right now is that writing to RAM presets does not retain. I'll need some kind of patch storage for my project so I don't have to rely on the Max 4 Live devices.