.41 Temporary test build bug report thread

[grantek]

Here's one I bet no one could answer.

Has anyone got MPT working under Linux, with 2 soundcards? One integrated (which is the one MPT tries to use, but I dont use it), and one PCI that I want to use, but it doesnt register.

Now THAT's a tech support question!  :lol:

If you don't use any 2-card functionality, just disable the internal one in the BIOS.

[rewbs]

anboi,

i still have weird performance issues i mentioned in another thread but have not had time to track down which version number they start at

Let me know when you do, I'm eager to figure that one out.

using reorder/remove in the channel manager makes the wrong channels be muted/unmuted according to the markings

Thanks, fix will be in next revision.

something weird with the screen redraw in some instances when changing to pattern view from something

Hmm I think I need more info to reproduce that one.

trying to play a note lower than G#-0 results in... nothing! d'oh!

For IT files, there's a hard coded upper limit for the period length (which goes up when the freq goes down). Has been there since pre-open source days.  I couldn't figure out the rational for that limit, so it will experimentally be removed in the next revision. Look out for bugs that kick in at very low notes.

bug with setting of highlighting - [...] it complains about  primary highlight needing to be higher value than secondary or something along those lines

Are you sure this is a bug? When highlighting is set to match time signatures, the rows/beat will be used for secondary highlights, and the rows/measure will be used for primary highlights. It is normal that the latter must be greater, since you should have at least one beat per measure. Does that clarify things or does it still feel wrong?

when typing in instrument allocations (which sample is mapped to which note in the instrument view) changing the pitch moves the cursor down [...] clicking miles up on note part only moves it up one, instead of jumping to where clicked, seems strange.

Agreed that this is irritating, will be changed in next revision.

also logic of order is silly. pitch - pitch - sample no. should be pitch - sample -pitch

I kind of agree, but don't think it's worth changing since many people will be used to the current layout and will yell at a swap. Certainly not worth the effort of making it an option...

can't type user wave names into imposcar cos modplug steals the keypresses.

I have added a "Pass keypresses to plug" switch available from the pluginwindow's "options" dropdown menu; will be in next revision.

sorry for not having all the necessary info in some cases but i think most are relatively clear, i'll try and help as much as possible with any specifics you might need.

Thanks.

[rewbs]

Hi VSh, thanks for the bug report.

If Midi Library was imported from some SF2 file with stereo samples and some midi file was opened after that, samples aren't recognized as stereo (only mono) and midi sounds very bad.

Hmm yeah SF2 support is a bit flaky, not sure how much time I will be able to invest in improving it. Can you point me to the specific SF2 file you're using?
If there's a quick fix I'll check it out.

[rewbs]

I'm just uploading v1.17.02.36. See beginning of thread for details.

[BooT-SectoR-ViruZ]

(+ <rewbs> volume commands without notes now send volume MIDI CC to plugs, but as far as I've seen, no plugs seem to respond to it.)

:cry: damn


wow, my name as betatester is in the product-info  

thx man

[Anonymous]

Hi VSh, thanks for the bug report.

Thank you for this good tracker.

Hmm yeah SF2 support is a bit flaky, not sure how much time I will be able to invest in improving it. Can you point me to the specific SF2 file you're using?

It is the best to try with Reality_GMGS.SF2 from sf2midi.com. Some samples sound good (as "52: Choir Aahs" or "10: Music Box"), but "0: Piano 1" sounds very bad in midi.

[rewbs]

Here you have a silly but very small example of what happens
http://www.oocities.com/rncekel/prueba1.it

Instrument 3 is really a triangle. But if you open the Edit menu->Cleanup-> Rearrange samples, instrument 2 will still point out to sample 3, that will now be empty.

Sorry rncekel, i forgot to put the fix for this in the latest build. The problem is a small typo that prevents the last intrument from being adapted to the new sample positions. Fix will be in next revision. Meanwhile you can add a dummy instrument in the last position before doing re-arrange samples, and all should be fine.

[rncekel]

I don't know what happens, I supposed I am doing something wrong, but I can't find the new version 1.17.02.36, but the old one 1.17.02.35.

[rewbs]

I don't know what happens, I supposed I am doing something wrong, but I can't find the new version 1.17.02.36, but the old one 1.17.02.35.

No you're probably looking in the right place, but Sourceforge's CVS web frontend is taking ages to update.

[BooT-SectoR-ViruZ]

still not there  :cry:

[Relabsoluness]

I don't know what happens, I supposed I am doing something wrong, but I can't find the new version 1.17.02.36, but the old one 1.17.02.35.

No you're probably looking in the right place, but Sourceforge's CVS web frontend is taking ages to update.

No new files are available even when using cvs-update - shouldn't it be updated much faster than the web frontend?

[rewbs]

I don't know what happens, I supposed I am doing something wrong, but I can't find the new version 1.17.02.36, but the old one 1.17.02.35.

No you're probably looking in the right place, but Sourceforge's CVS web frontend is taking ages to update.

No new files are available even when using cvs-update - shouldn't it be updated much faster than the web frontend?

Yes hmmm... maybe I didn't commit properly. I've recommitted.

[rewbs]

Relabsoluness, can you try CVS update again?
I think I'm in sync with the repository, but the web frontend is still not updating.

[LPChip]

Relabsoluness, can you try CVS update again?
I think I'm in sync with the repository, but the web frontend is still not updating.

On my end, the download points to rev. 35, and checkout is empty.

http://cvs.sourceforge.net/viewcvs.py/modplug/ModplugCheckout/?

[Squirrel Havoc]

I'll check out the new update with my Linux box. Just so you know, CVS is still reporting the old version