Firmware V2.0C

[XORadmin]

Hi Thomas,

A question on pattern row and NRPNs. I'm just trying to understand what I'm seeing, because I think it would help me to see into how mappings work.

I have pattern with a BRK 000 on row 03.
In mappings I have TRCK TRK1 PTRW > MCON D-16 N--0

The values I'm getting look like:
[0, 0, 0, 0]
[0, 0, 0, 40]
[0, 0, 0, 84]
[0, 0, 0, 124]
[0, 0, 1, 40]
[0, 0, 1, 84]
[0, 0, 1, 124]
[0, 0, 2, 40]
[0, 0, 2, 84]
[0, 0, 2, 124]
[0, 0, 3, 40]
[0, 0, 3, 84]
[0, 0, 3, 124]
[0, 0, 4, 40]
[0, 0, 4, 84]
[0, 0, 4, 124]
[0, 0, 5, 40]
[0, 0, 5, 84]
[0, 0, 5, 124]
[0, 0, 6, 40]
[0, 0, 6, 84]
[0, 0, 6, 124]
[0, 0, 7, 40]
[0, 0, 7, 84]

In every case PTRW seems to yield about double the number of the actual pattern row in the MSB, and some ticks in the LSB. Row 00 shows four initial 0s, every normal row shows up three times, and whatever is the last row shows twice. So if BRK is on row 07, before reset the last numbers are:
[0, 0, 14, 40]
[0, 0, 14, 84]
[0, 0, 14, 124]
[0, 0, 15, 40]
[0, 0, 15, 84]

This all seems to be BPM independent.  If you have a moment, can you give a hint as to what's happening here?

To start with, check out what values the command is generating. In that case the current row is scaled to the full range. Row 0-63 equals to the values 0 - FFF. But the rows are generated by the ticks which means that you would get also in between values instead of only 64 different ones. (you can scale that down of course if you prefer that)

Next is the output and what you see in your midi data is the NRPN Data which is 7 bit encoded per byte. It needs some bit shifting to decode whats going out there. In fact what it generates is the following: Row 0-63 equals the NRPN Data of 0 - 16336 (it's actually normally 16383 at max but the source precision is lower than that and this is the scaled value next to it). [0,0,126,36]