.50 Instrument panning not saved correctly (reopened)

Saga Musix · March 08, 2008, 19:55:20

Full Version:
OpenMPT v1.17.02.50

Has the bug occured in previous versions? If yes, please specify version(s): (This option is optional, and doesn't need research. But in case you know...)
not tested

Description of the bug:
Instrument panning is not saved correctly. If you choose 256 as a paning value, it will be reset to 0. So the maximum value is 252, because panning is saved in steps of 4 units.

How often does it happen?:
tested in IT format.

How to reproduce (step by step description):
try 255 and 256 ans an instrument panning position.

Winter · August 25, 2008, 13:53:08

Hello there,

this problem is annoying me! I always get this problem! Everytime I save the panning for each channel and reload the IT-file again, all! pannings are reset!
The bug appears only in large IT-files or detailed and complex projects.
If you open a new file, set a panning, save, reload, then everything is ok!

Sometimes it's a game of luck. I always set the panning befor saving to do not abandon hope, because sometimes it happens that the panning is saved correctly. No idea why or why not!

This is really a strange bug! And you need good anit-bug spray!

my MP Version: 1.17.02.48
Channels: 17!
Patterns: 105 different Patterns

Oh! Now I see that the pannings of channel 15 and 16 are safed correctly, but not the channels 5-8 (pans: 32,64,192 and 224)

psishock · August 25, 2008, 14:04:35

temp solution, to keep you from further head-aches on saving/loading:
set panning positions in pattern data (either with Pxx or Xxx commands), not in "instruments tab" or "general tab".
Those codes will not dissapear for sure.

bvanoudtshoorn · August 25, 2008, 14:58:24

Winter, please also upgrade to the latest version of OpenMPT - you might find that the bug is fixed (or at least behaves differently

) on a newer version.

You should monitor this thread for info about new versions. As I write this, OpenMPT is up to 0.52. You can download it here - download "mptrack.exe" to wherever you'd like.

Saga Musix · August 25, 2008, 19:53:16

oh right, seems like it has been fixed and i haven't noticed it.

bug closed.

Relabsoluness · August 25, 2008, 20:15:21

Quote from: "Jojo"oh right, seems like it has been fixed and i haven't noticed it. bug closed.

Fixed? I haven't noticed either

Winter's description of 'really strange bug' doesn't sound promising, though. If someone has a file with which the bug can be reproduced, it could be valuable for debugging purposes.

Winter · August 25, 2008, 21:27:10

bvanoudtshoorn, I was wondering why you all are talking about .50 and now .52, because I could not find it anywhere in the download-section.
Thanks for the link!

But I am sorry! This bug isn't fixed.
I made the test with the new exe. I panned, saved, loaded, wrong panning!

I can not give you my .IT-file, because the track will be released on a netlabel. You know... exclusive rights and so on.

LPChip · August 25, 2008, 22:07:36

Quote from: "Winter"bvanoudtshoorn, I was wondering why you all are talking about .50 and now .52, because I could not find it anywhere in the download-section.
Thanks for the link!

But I am sorry! This bug isn't fixed.
I made the test with the new exe. I panned, saved, loaded, wrong panning!

I can not give you my .IT-file, because the track will be released on a netlabel. You know... exclusive rights and so on.

But I'm sure you can make a new track, setup a test case that has a similar thing (just fill many patterns with garbage if you have to) and send that one?

Saga Musix · August 26, 2008, 11:15:12

Quote from: "Relabsoluness"
Quote from: "Jojo"oh right, seems like it has been fixed and i haven't noticed it. bug closed.
Fixed? I haven't noticed either Winter's description of 'really strange bug' doesn't sound promising, though. If someone has a file with which the bug can be reproduced, it could be valuable for debugging purposes.

i don't recall the complete situation, but i i just tried to save a module with those instrument pannings and it worked...

Winter · August 26, 2008, 18:52:45

Okay here is a contaminated .IT-file! :twisted:

Good luck!

Click here!

Relabsoluness · August 26, 2008, 20:05:26

Based on the testcase and other examinations, it seems that the reason is in the 'surround'-setting. If channels surround setting is enabled, pan value for that channel will not be saved to file - it just saves the information that the channel surround is enabled. And when loaded, pan is set to 128. This behavior has been there since mpt 1.16.

2Winter
Thanks for testcase. Can you confirm that surround-setting is the cause of the problem?

Saga Musix · August 26, 2008, 20:31:42

this behaviour seems to be logical because surround disables stereo position, but this is not a reason not to save the stereo position imo.

Winter · August 26, 2008, 22:05:52

Ok, let's see...

Indeeed!

Thanks, Relabsoluness!
Now that's the trigger!

Although it's a pity, because I want to use stereo-effects and surround!
Even after a restart! ;-)

Saga Musix · August 27, 2008, 11:57:24

winter: it is just not possible because of the way surround works. trackers use "fake surround" by reversing one of the stereo channels. this does not work correctly with panned samples.

älskling · August 27, 2008, 14:35:49

Quote from: "Jojo"winter: it is just not possible because of the way surround works. trackers use "fake surround" by reversing one of the stereo channels. this does not work correctly with panned samples.

It should work with 4 channel output instead of prologic.