MC-101

Some experiments about sysex on my Roland MC 101 firmware 1.82 (reference manual with all updates included upto 1.81, sorry large pdf)

Reply from identify message [ 7E 7F 06 01] is [7E 7F 06 02 41 5E 03 00 00 00 01 00 00]

Full scan of address map with RQ1 [41 10 00 00 00 5E 11 xx yy zz 00 00 00 00 7F] for xx,yy,zz from 00 to 7F dumped in this FILE.

The project loaded in the MC-101 was this ONE.

Some analysis with the help of the Fantom midi implementation gives this RESULT.

Be careful, when replying, the MC-101 change the address given in the RQ1 message, the mapping from the requested address to the reply address seems quite strange ...

I.E. : asking 7F bytes at 53 22 56 00 gives the following reply : 

F0 41 10 00 00 00 5E 12 0A 68 2B 00 00 00 00 00 00 00 00 00 00 7F 40 7F 7F 00 01 40 40 08 00 08 00 08 00 08 00 08 00 7D F7

the reply address is seen as 0A 68 2B 00 .... quite strange ? but I thinks this perhaps the computed address in ram of the MC ... anyway the data seems consistent with Fantom documentation.

Structure of sysex seems consistent with an MC 707 too. Structure related to tracks are 8 in size even if the MC-101 got only 4 tracks.

Map can be hinted as follow:

00 00 00 00 :1D: ? System ?

10 00 00 00 - 10 00 3A -- : ? Config/system and tracks datas ? 

20 00 00 00 - 28 3E 00 -- : ? clips datas ? base structure repeated 128+8 times (16 clips times 8 tracks) + (8 currents edited tracks ?)

30 00 00 00 - 32 0E 41 -- : [ Z-Core Tone] as Fantom repeated 128 + 8 times 

17 per track : # 0-15 : for each clip of track 1 (if tone track, and tone per clip set) + # 16 for track

Tested : Z-Core tone #17 is tone for clip 1 of track 2 for a default init project ... sending change are ok (tested with the name of the tone) 

32 10 00 00 - 35 26 45 -- : [Drum Kit] as Fantom repeated 128 + 8 times (1 per 128 clips) + 1 per tracks (8) ?

40 00 00 00 - 41 07 02 -- : ? perhaps effects (one sub struct if quite like [MFX] ? repeated 128 + 8 times as usual

50 00 00 00 - 50 7F 00 -- : 64 [Z-Core Tone] names are in substruct ... but why 64 tones here ? 

52 00 00 00 - 53 E3 45 -- : 64 [Drum Kit]  as Fantom ... names are in substruct ... but why 64 kits here ?  

some works needs to be done to asses thoses datas.

Actually only RQ1 messages where sent, DT1 messages have to wait for some data analysis before tests to avoid "bricking" device ?

Some RQ1 message for 7F bytes only replied with 7E bytes as they should have returned 7F bytes as from a 01 00 field ...

Perhaps I should have asked for 00 00 01 00 bytes in the RQ1 message ? but the full scan took a week to be done, so I will not redo it soon

Done with a python script, not too fast to avoid messages overlap as it can lock device ... (need to remove battery as the on/off switch was non operative ... certainly due to buffer overflow in the MC-101 MIDI system ...)