Some Bugs NKI, SF2

[eve93100]

Recently, while converting some Kontakt instruments, the program threw an error. I remember that another time I reported this, you mentioned that you still didn't fully understand how this part of the Kontakt format worked. This is the error:

ConvertWithMoss 16.1.1
Detecting and converting multi-samples from Kontakt NKI/NKM to DecentSampler...

Analyzing: C:\Users\Erick\Documents\Producer\Instruments_Conv_in

Analyzing: C:\Users\Erick\Documents\Producer\Instruments_Conv_in\HEADHUNTER-X5D.nki
Detected Kontakt Container format (created with Kontakt 8.1.0.0 - monolith) - Little-Endian.
Unknown internal modulation header.

Finished.

Regarding SF2, in some cases when you convert an instrument to SF2, the "Tune" parameter is doubled with respect to the original value; that is, if the value is "7", when converted the value is "14", and if the value is "-20", when converted the heat obtained is "-40". This does not occur in all converted instruments.

I've attached examples.

HEADHUNTER-X5D.zip

Accord Clari.zip

The file "HEADHUNTER-X5D" is the one with the header problem.

The file "Accord Clari" is the SF2 example; it's in NKI format. You have to convert it to SF2 to reproduce the problem. If you can't reproduce it on your machine, please let me know so I can send you examples converted on my machine.

Thank you for your attention!

[git-moss]

Thanks for the detailed and helpful report.
Issue 1: Same problem with different exception. I will simply ignore all of them for the time being (will be in the next update.
Issue 2: I could "only" spot a rounding error which occurs for some negative cent values (which were off by 1). Where do you see this doubled values? Could it be a bug in the application you use?