Print

[crescent]

Hi,

It is nice to have Gravis-Patches support! 

It's very very hidden however, and different to Impulsetracker you can't choose which sample of an instrument you would like to use.

What I like is that you can load a pat as an instrument!

Here are the bugs:

- Not possible to use it as a sample
- Usually loaded at wrong pitch
- Multisample patches are loaded totally mistuned to each other. Just try to use one of those instruments!

All in all a great "hidden feature", which needs some work still!

[raina]

Hidden? Don't think so, .pat importing is mentioned in the documentation. The pitch issues are known but what do you mean by not being able to use .pats as samples? Just edit the instrument key mapping, copy samples around or do whatever you like. Like if you want to use sample 3 from a patch, use the Copy instrument/sample dialog (instrument editor -> Copy) to copy that sample into sample slot 0 of another instrument. Is this what you were after? Another way would be to load it as a raw sample and just crop out the instrument headers and unwanted samples.

[crescent]

Thanks! That clarifies a lot!

Hidden: My mistake! In older versions of Milky it was necessary to type in *.pat by hand.
I oversaw the improvement by trying to load them (the Impulsetracker way) as samples-
Actually there are only the pitch issues left for the "known bugs list".

[raina]

pailes said something before about not bothering with perfecting the old format importing as it isn't possible to support it 100% anyway (.pat can hold more than 16 samples, .xi cannot) but I wouldn't mind if the pitch issues were sorted either. Actually Milky is trying to detect the correct relative note for samples when they are loaded but the different sampling rates and pitches used in the default GUS bank for example are complicating things.

Updating the bug list for completeness..

[pailes]

I was playing around with various properties stored in the PAT file to properly detect the relative note and the finetune of the given sample in the patch, but it didn't work out. If someone has knowledge about it I'd like to hear about it.

[crescent]

Surprisingly there is very few about the pat-format on the net.
It took me nearly an hour to find these links:

A patchguide:
- http://mstation.org/patguide.html

Some info of someone who had similar problems (but finally succeeded):
- http://alsa.opensrc.org/index.php/WildMidi

Something about the format:
- http://www.softhelp.ru/fileformat/patfmt/Pat.htm

Another description of the file format:
- http://www.onicos.com/staff/iz/formats/guspat.html

The last document says it: There is no fine tune- 
Could it be the "Root frequency" you are looking for?


Good luck!   

[KungFuFurby]

This is quite a jaw dropper... I never knew it was supported in MilkyTracker! I'll go ahead and let the people at my IT tracker clone, Schism Tracker, know about this staggering find.

[pailes]

I gave it another try today and it seems that I have succeeded in properly retrieving the relative note number out of the root frequency of the sample. Anyone any GUS patches to try?

[raina]

Hooray! Here's the Ultrasound bank: and here it's not anymore.

[crescent]

You host them on your own site? Nice!