Offline Editor For Project Files (*.NRD)

[eliclement]

Updated. Sorry for the trouble.

No problem. Thanks for the fast hotfix release :-)

For the offline editor, I finished the modular track patterns and now I'm working through the other track patterns...

1. How is the 2 byte drummatrix structured?
   
2. How do I export/import the bulk pattern files in version 1.27(A)?
   

Cheers!

[XORadmin]

Updated. Sorry for the trouble.

No problem. Thanks for the fast hotfix release :-)

For the offline editor, I finished the modular track patterns and now I'm working through the other track patterns...

1. How is the 2 byte drummatrix structured?
   
2. How do I export/import the bulk pattern files in version 1.27(A)?
   

Cheers!

2 bytes = 16 bits = 16 triggers.

Bulk export: shift+ok after entering the filename.
Import automatically.

[proxemics]

@XORadmin FYI the download link seems to be broken... "no file".

Top two links (sequencer) broken
bottom two (cv16) working

[XORadmin]

@XORadmin FYI the download link seems to be broken... "no file".

Top two links (sequencer) broken
bottom two (cv16) working

Fixed. It took 3 days until someone reported it :-) Thanks.

[proxemics]

Yes, but one of the days was xmas.

[mgd]

Each tracktype (Modular/CV16/Trigger16/Audio) got a different layout. The drummatrix itself is hidden in 2 bytes ;-)

But lets show the most basic ones and I will deliver the other ones later (especially CV16 where I have to pack the data in 12 bit words).

[stuff snipped]

Midi
[1] Note 1
[1] Velocity 1
[1] Note 2
[1] Velocity 2
[1] Note 3
[1] Velocity 3
[1] Note 4
[1] Velocity 4
[1] Program Change
[3] FX1 (1 byte fx command, 2 bytes payload)  
[3] FX2 (1 byte fx command, 2 bytes payload)  
[3] FX3 (1 byte fx command, 2 bytes payload)  
[1] Patch
[1] Table
[1] Groove

Here's my guess at CV16 - please correct:

CV16
[1] CV1 bits 0-7
[1] CV1 bits 8-11 and CV2 bits 0-3
[1] CV2 bits 4-11
[1] CV3 bits 0-7
[1] CV3 bits 8-11 and CV4 bits 0-3
[1] CV4 bits 4-11
[1] CV5 bits 0-7
[1] CV5 bits 8-11 and CV6 bits 0-3
[1] CV6 bits 4-11
[3] FX1 (1 byte fx command, 2 bytes payload)  
[3] FX2 (1 byte fx command, 2 bytes payload)  
[3] FX3 (1 byte fx command, 2 bytes payload)  
[1] Patch
[1] Table
[1] Groove

So basically a variant of the Midi layout.

Please correct, especially w/r to the bit numbering

Kind regards,
Michael

[XORadmin]

Here's my guess at CV16 - please correct:

CV16
[1] CV1 bits 0-7
[1] CV1 bits 8-11 and CV2 bits 0-3
[1] CV2 bits 4-11
[1] CV3 bits 0-7
[1] CV3 bits 8-11 and CV4 bits 0-3
[1] CV4 bits 4-11
[1] CV5 bits 0-7
[1] CV5 bits 8-11 and CV6 bits 0-3
[1] CV6 bits 4-11
[3] FX1 (1 byte fx command, 2 bytes payload)  
[3] FX2 (1 byte fx command, 2 bytes payload)  
[3] FX3 (1 byte fx command, 2 bytes payload)  
[1] Patch
[1] Table
[1] Groove

So basically a variant of the Midi layout.

Please correct, especially w/r to the bit numbering

Kind regards,
Michael

You are a smart person :-)

Lets add also the audio part and then we are there (for now):
[1] Note 1
[1] Sample/wave 1
[1] Volume1
[3] FX1.1 (1 byte fx command, 2 bytes payload) 
[3] FX1.2 (1 byte fx command, 2 bytes payload) 
[1] Note 2
[1] Sample/wave 2
[1] Volume 2
[3] FX2.1 (1 byte fx command, 2 bytes payload) 
[3] FX2.2 (1 byte fx command, 2 bytes payload) 
[1] Groove

Some do realize that there is 1 byte left here :-)  What to do with that?

[mgd]

Hi Nerds!

[updated once more ]

Find attached my again updated Perl script to transform an exported [bulk] pattern file into a text version and also back again. The sequencer parts are now appended side by side in one block at the end of the file. I intentionally did not try to come up with XML or JSON, because
- for one it would be trivial to amend the script accordingly if someone wants to do it
- my main goal is and was to edit projects in editors like emacs or notepad++
- at this point I want to be sure I got everything right

Things missing:
- Video patterns - the format is not yet described and I don't have a video NSA to test it, however each row is dumped as 21 hex bytes
- raw "voltage" value in CV16 lanes are only an approximation and sometimes off by 1 (smaller and/or larger) than the original values. I do linearly interpolate the values and my formula obviously differs ever so slightly from the one Thomas is using
- very little to no plausibility checks for parameters

Except for bug fixes and of course format changes and additions (e.g. Video) I expect this to be done for the time being.

A very brief usage section is at the top of the Perl Script.
Feel free to ask questions.

Kind regards,
Michael

PS: I had to append a ".txt" to the name of the script to be able to upload.
PPS: A few counts in the file format description are off by 1, namely the number of Patterns and Sequence rows. The description gives the last valid ID which is 1 smaller than the count since the IDs are 0-based. Details regarding this and when/where "endian-ness" matters are commented in the Perl script.

[proxemics]

Wow, Michael.... impressive quick coding!

I'll try to try it out. Do you have a timeline for the text->nerd side?

Thanks for this!

[mgd]

Wow, Michael.... impressive quick coding!

I'll try to try it out. Do you have a timeline for the text->nerd side?

Thanks for this!

Conversion from text to NerdSEQ pattern is implemented in the current version. My test bulk export was converted to text and then back to NerdSEQ. The final file was byte wise identical to the original. While this does not rule out that I missed something, I'm confident that most things should work just fine.

I'm happy to fix any bugs that I'm notified about.

Hope this is useful for others as well - I will definitely use it