Schism Tracker development

No.1723/1: Fullscreen Freezes the Program

George — Sat, 18 May 2013 16:54:13 GMT

Mac OSX Bug Report

BUG: When I tried to put the program into full screen, the window went away and stopped responding, if you force stop the program and restart it, then it will try to go into fullscreen and then it will not have any windows displayed and it will be frozen. If you go to the window bar the rainbow pinwheel beachball, whatever you want to call it, comes up and you can't click on any of the drop-down menus. Deleting and reinstalling the program doesn't help.

Please help, email me at gmorr3@sbcglobal.net

No.1722/1: Fullscreen Freezes the Program

George — Sat, 18 May 2013 16:54:07 GMT

Mac OSX Bug Report

Please help, email me at gmorr3@sbcglobal.net

No.1709/2

George — Sat, 18 May 2013 16:51:46 GMT

Mac OSX Bug Report

BUG: When it tried to put the program into full screen, the window went away and stopped responding, if you force stop the program and restart it, then it will try to go into fullscreen and then it will not have any windows displayed and it will be frozen. If you go to the window bar the rainbow pinwheel beachball, whatever you want to call it, comes up and you can't click on any of the drop-down menus. Deleting and reinstalling the program doesn't help.

Please help, email me at gmorr3@sbcglobal.net

No.1707/4

George — Sat, 18 May 2013 16:51:30 GMT

Mac OSX Bug Report

Please help, email me at gmorr3@sbcglobal.net

No.1711/3

George — Sat, 18 May 2013 16:51:14 GMT

Mac OSX Bug Report

Please help, email me at gmorr3@sbcglobal.net

No.1700/5

George — Sat, 18 May 2013 16:51:03 GMT

Mac OSX Bug Report

Please help, email me at gmorr3@sbcglobal.net

No.1715/3

George — Sat, 18 May 2013 16:50:10 GMT

Mac OSX Bug Report

Please help, email me at gmorr3@sbcglobal.net

No.1715/2

nbensa — Tue, 23 Apr 2013 11:50:19 GMT

Sorry. Ignore 1715. This IS bug 1551. I finaly were able to apply the patch.

No.1715/1: -ref 2871 breaks jeff93.it (Drifting Onwards)

nbensa — Tue, 23 Apr 2013 05:55:06 GMT

Hello,

Some channels (percussion) are not played.

At first, I thought this is bug #1551
http://schismtracker.org/scdev/res/1551.html

But I cannot apply the patch posted and so I cannot confirm this is the same bug.

I did hg bisect:

$ hg bisect -g
The first bad revision is:
changeset: 2871:cf8f327d96d2
user: Storlek

Comment too long; full version here.

No.1700/4

Saga — Thu, 11 Apr 2013 13:17:32 GMT

With linear slides, Impulse Tracker continues pitch slides into infinity (well, probably until it wraps around or maybe there is an insanely high limit), but with Amiga slides it should indeed stop.

No.1700/3

JosepMa/JAZ — Wed, 10 Apr 2013 18:49:46 GMT

I recently got into this in Psycle and after looking into the problem, I found that it is documented:

This is from Impulse Tracker doc:

      Fxx   Pitch Slide up
      FFx   Fine Pitch slide up
      FEx   Extra fine pitch slide up

              Operates in the same manner as the Exx commands, but
              slides the pitch up. If the pitch gets 'too high', then
              the channel is turned off.

"Too high" means that the (amiga) period goes below 1. (Since a period 0 has no meaning and periods are integer values in these trackers). I am unsure what the effect is with linear slides, but the same logic might apply.

No.1711/2

Saga — Wed, 20 Mar 2013 10:47:01 GMT

In fact, this is a MOD file, renamed to S3M. SchismTracker seems to use a wrong period to frequency conversion table. I don't remember the gory details, but it has something to do with PAL vs NTSC timing.

No.1711/1: S3M file instrument played in staccato-like fashion

jengelh — Tue, 19 Mar 2013 15:37:56 GMT

In http://inai.de/scdev/ , you will find an S3M file, and how schismtracker-20120105 and the XMP player render it to WAV (encoded to ogg for small downloads).

Playback the S3M with schismtracker and set channel 1 to solo mode. In order 2+ (pattern 4), you will notice that schismtracker does sort of "glottal stops" when the instrument is invoked anew. XMP does not seem to do this, but which implementation is right?

No.1707/3

Bojan Landekic — Sun, 17 Mar 2013 03:22:48 GMT

Hey GreaseMonkey, where did you get the stats on IT 2.12's volume ramping and anticlick? I tried Googling it but couldn't find nothing.. :-/

No.1709/1: schism is not very good at detecting free instruments

Yolanda Mercer — Tue, 12 Mar 2013 03:27:23 GMT

File: 0:00

When loading a file, schism seems to think that blank instrument slots before the highest numbered instrument aren't really blank. Any time you load a sample in instrument mode and select "Create new instrument", and the corresponding instrument slot isn't the highest numbered instrument, the instrument will get created after the highest numbered one.

I don't feel like I'm being very clear, so as an example here's a file with samples and instruments in slots 01 and 20. Loading a sample in slot 02-19 will create an instrument in slot 21. This will hold true even if you delete the instrument in slot 20 and save/reload the file (well actually it goes in slot 20 since that one got deleted properly).

No.1707/2

Saga — Sat, 09 Mar 2013 14:25:02 GMT

Getting rid of these clicks in Schism Tracker can actually be achieved by much easier means than implementing a completely new anti-click mechanism.
Instead of setting the channel length to 0, you should set the fadeout volume to 0 and enable the CHN_FADEOUT flag. It has the same effect (i.e. volume commands can't "pick up" a note after it has been cut), but it avoids the clicks.

No.1707/1: Anticlick would be a good idea

GreaseMonkey — Fri, 08 Mar 2013 22:38:10 GMT

File: 4:17, chinariffic

I remember what was apparently 2 years ago, rendering my piece for 8bitcollective's "chinese whimpers" (chinese whispers vol 2), and noticing that the bass clicks quite horribly.

IT 2.12's diskwriter introduced a simple anticlick. My implementation of it is as follows:

Each audio channel ("virtual" as opposed to "pattern" channel) has a set of "anticlick" values, one per output channel, which are initialised to 0. There is also a set of global "anticlick" values (one per output), which are initialised to 0.
On a note cut NNA (explicit cut, SCx, note that follows another note where NNA is set to note cut), add the audio channel's anticlick values to the global anticlick values
At mix time, set all audio channel anticlick values to 0.
For each audio channel that's actually being played, set the anticlick values (1 per output channel) to the last value that comes out of it.
Add the ramping down of the global anticlick values linearly interpolated down to 0 at 1/400 of a second (the IT 2.12 volume ramping rate - IT 2.11 uses 1/500) to the output stream.
Set the global anticlick values to 0.

Comment too long; full version here.

No.1705/2

Saga — Mon, 04 Mar 2013 19:46:50 GMT

Compressed stereo samples. Yay.

No.1705/1: Stereo loading bug (another one)

Saga — Wed, 20 Feb 2013 01:02:30 GMT

File: 0:07

You are probably very well aware of this, but since there are actually applications out there that can write IT-compressed stereo samples now (munch.py and OpenMPT), it should be pointed out that Schism Tracker only loads the left channel of such samples.

No.1703/2

Evan C. Robinson — Sun, 03 Feb 2013 07:39:45 GMT

Confirmed. Strange!

Category: User Input
Complexity: Trivial
Priority: Normal
Screen: Instrument List, Sample List
Status: Accepted

No.1703/1: multichannel playback and naming things

Tobias Clements — Tue, 29 Jan 2013 14:15:44 GMT

Here's a minor bug: if multichannel playback is enabled (F3/F4, Alt-N), entering an instrument or sample name will increase the playback channel whenever a key that would play a note is pressed, even though no note actually plays.

No.1699/2

Ivette Woods — Wed, 23 Jan 2013 02:34:00 GMT

Page up/down keys work perfectly with all combinations of control/alt/shift on my computer, and I'm running the same version. Only difference is I have a US keyboard.

I swear I've heard this same bug before somewhere, but I can't for my life remember how it was resolved.

No.1700/2

68k — Tue, 22 Jan 2013 11:23:30 GMT

Confirming that the s3m sounds quite different in ST3 or IT from Schism.

In IT the (Amiga) F80 slides cause the note frequency to quickly go past the point where IT just cuts the note off instead. Schism doesn't seem to have a point like this, so you get some crazy aliased noise instead.

No.1700/1: Bug in pattern #6 (Order: 5/53) at channel 7 in song "Panic by Future Crew".

milo64 — Sat, 19 Jan 2013 05:40:39 GMT

File: 3:55

Finally, this bug shall be fixed.

On Schism Tracker 20120105 built Jan 19 2013 11:14:14.
Compiled using the following commands:
./configure
make
make check
make install

In song "Panic by Future Crew". Downloaded from scene.org.
In order 5/53 pattern #6 at channel 7.
A very high frequency beep occurs.

But when I play this song on a mod decoder (some kind of lib), the beep sound is gone.

Comment too long; full version here.

No.1699/1: CTRL-PgUp in Pattern Editor has no effect

Topy44 — Fri, 18 Jan 2013 16:25:44 GMT

CTRL-PgUp (Move to top of pattern) has no effect. PgUp (on its own) and CTRL-PgDn work as expected.

Version 20120105 running on Windows 7
German keyboard layout (tried with US int., no change)

I can be reached at #schism on FreeNode for further tests/information.

No.1196/4

Leanne Constance Vinson — Sun, 30 Dec 2012 21:41:07 GMT

>>1697
That's an artifact of modplug's tendency to start all new notes on a separate channel with a fast fadeout of the old note, in lieu of a real anticlick mechanism.

No.1196/3

68k — Sun, 30 Dec 2012 12:57:01 GMT

File: 0:08

For posterity, I'm going to add on to this to say that the channels-in-use counter is not correct. If a sample is playing in one channel and it is cut off by another, even in sample mode, Schism briefly registers two channels as being active. This results in the maximum channels in use counter being inflated.

Example file uses 1 channel max, then 4 channels. Counter registers 2 then 8.

No.1551/8

68k — Sun, 30 Dec 2012 12:47:32 GMT

Since someone else is talking about this error lately, I'll confirm that the patch from >>1567 does fix the issue.

No.1678/6

Eva Combs — Tue, 25 Dec 2012 08:08:18 GMT

>>1691
Volume envelopes in instruments are still broken in hg-build, which renders it mostly unusable.

No.1686/6

delt !DEimOs08uM — Sun, 23 Dec 2012 16:50:14 GMT

oh, your memory must be better than mine about this stuff... i (probably) stand corrected :D

No.1686/5

Dejuan S. Murray — Sun, 23 Dec 2012 15:32:28 GMT

>>1692
Yes there is - if you don't change it back, then every note ever will apply the new NNA settings! It needs to copy the instrument setting to the channel's copy. IIRC, Modplug originally didn't make a copy at all, which was... terrible.

No.1686/4

delt !DEimOs08uM — Sun, 23 Dec 2012 00:05:57 GMT

...iirc in Impulse Tracker there's a per-note NNA variable as part of the note data structure, or something like that.

This variable gets.. like you said, "copied" from the instrument data each time a note is played - except if you use a S7x command, which supersedes the instrument's NNA value. (there is no need to "revert" anything)

No.1678/5

Sophia Dianna Blevins — Sat, 22 Dec 2012 13:20:40 GMT

>>1684
yes. It's not hard, just a few lines. Someone can do this ;)

>>thread
Those dialog mouse problems were resolved a long time ago. If you build from hg it'll work. Unfortunately I don't think all the compiled versions available for download are up to date with that fix.

No.1686/3

Dejuan S. Murray — Sat, 22 Dec 2012 13:18:42 GMT

I'm not sure it is the same. Note that you have no instrument numbers in that column after the first note. Add them and it seems to play correctly.

I THINK what's going on here is that, when you write an instrument number, it makes a copy of the instrument and attaches it to the channel. S7x modifies that copy directly, and once you change something, it's never "reverted" like it should be.

Really the copy should probably be happening when a note OR instrument is triggered. I think. It's been a very long while.

Category: Playback
Complexity: Moderate
Priority: Normal
Status: Accepted

No.1626/4

Tierra Shields — Sat, 22 Dec 2012 13:11:46 GMT

I don't think the 4/8 keys bother searching back for a number. They should, yes.

Category: User Input
Complexity: Moderate
Priority: Normal
Screen: Pattern Editor
Status: Accepted

No.1626/3

Peter Kasting — Sat, 22 Dec 2012 05:45:02 GMT

Note: This still happens with the 20120105 build.

No.1686/2

Peter Kasting — Sat, 22 Dec 2012 05:40:37 GMT

Note: I suspect this has the same root cause as http://schismtracker.org/scdev/res/1578.html .

No.1686/1: S7x should set NNA per-note, not per-channel

Peter Kasting — Sat, 22 Dec 2012 05:38:21 GMT

File: 2:33, Break It Down

In Impulse Tracker, if you use S7x to override the NNA for a particular note, it sets that note's NNA to the desired value and has no other effect. In Schism Tracker, the new NNA seems to be applied to all subsequent notes on the channel.

A test file is attached. Compare the sound in pattern 3 in Schism Tracker (where there's a great deal of smearing as a whole bunch of notes get S76 [== note continue] and all play at once) versus Impulse Tracker.

No.1681/3

esaruoho — Fri, 07 Dec 2012 12:27:32 GMT

Thanks, was just hoping that, as if by magic, these two had been done:
http://schismtracker.org/scdev/res/1678.html#i1684
and the ctrl-o (f2 feature, really neat) addition to F3..

No.1678/4

esaruoho — Fri, 07 Dec 2012 12:20:16 GMT

Is there any chance of making this saveable in Schism configuration, so that it defaults to Wav, if the user selects it to be Wav=default? Any chance whatsoever?

No.1681/2

Lane Marsh — Fri, 30 Nov 2012 00:14:44 GMT

Take a look at the hg repository and you'll see that nothing much as been done with the code since January.

No.1681/1: newer build than january 2012 for OSX?

esaruoho — Wed, 21 Nov 2012 17:47:26 GMT

hi, is there any chance of a newer build than january2012 for OS X? :)

No.1678/3

failotron — Thu, 08 Nov 2012 08:19:01 GMT

it works! thanks so much!

No.1678/2

Antoinette Silva — Wed, 07 Nov 2012 20:13:21 GMT

I was able to select WAV with keyboard:

[Tab][Down][Down][Down][Tab]

No.1678/1: Sample export in WAV - issue

failotron — Wed, 07 Nov 2012 14:41:17 GMT

File: 386x195, Screen Shot 2012-11-07 at 2.26.0(...).png

hi there

i'm fairly new to Schism Tracker, so please excuse if i'm asking the stupid obvious - even though i didn't really find a thread strictly relating to this issue.

i'm trying to export some samples from an .it file in order to use it in other programs, but when it comes to the export window (alt-t), i can't click on the WAV, AIFF, AU or RAW function. the only format i can use for exporting is the default ITS, which i can't open or use for anything external.
http://i.imgur.com/MLnzr.png

i'm running a Mac OS X 10.7.4 Lion

any help is more than welcome, thank you in advance!

áron

No.1550/7

Melisa Carey — Tue, 30 Oct 2012 05:00:30 GMT

Also, in case it matters, when loading the left channel, the right channel leaks in a bit in the first half. I'm pretty sure that at one point, it was the left channel leaked into the right.

No.1550/6

Ronda K. Hutchinson — Sat, 27 Oct 2012 16:50:03 GMT

I don't think it ever worked correctly.

No.1550/5

Casey Gus Brady — Fri, 26 Oct 2012 22:54:14 GMT

I remember mrsbrisby saying (at least once) that this bug was fixed. I do not remember a time when this bug was fixed.

No.1670/5

Saga — Sun, 14 Oct 2012 11:38:20 GMT

Hello vibrato, "nice" to meet you.

No.1670/4

GreaseMonkey — Fri, 12 Oct 2012 21:23:55 GMT

Now that I've had time to think of it, here's some code that does it right:

v = v*vib_depth;
if(old_effects) v = ~(v<<1);
bool go_down = (v<0);
if(go_down) v = ~v;
v = (v+32)>>6.

if(v >= 16)
{
if(go_down) slide_down(v>>2);
else slide_up(v>>2);
} else if(v >= 1) {
if(go_down) slide_down_fine(v);
else slide_up_fine(v);

Comment too long; full version here.

No.1670/3

GreaseMonkey — Fri, 12 Oct 2012 10:10:41 GMT

ARGH CRAP sorry for the triple post. (-2vvib_depth+16)>>5 for Old Effects.

No.1670/2

GreaseMonkey — Fri, 12 Oct 2012 10:08:50 GMT

Whoops, my bad. You do (v*vib_depth+32)>>6.

I haven't fully sorted out Old Effects. (v*vib_depth+16)>>5 comes extremely close, even being sample-accurate-give-or-take-a-one-digit-number on the first bit in b-joit.it.

No.1670/1: Hello everyone, I'm Vibrato. Let me ruin your life.

GreaseMonkey — Fri, 12 Oct 2012 05:36:26 GMT

File: 0:02, Hxx test lin/new no-rt

Good news everyone, I've found out how vibrato works in not-old-effects linear-slides mode!

Bad news, I don't know if anything does it right at all (aside from the player I'm working on, but even then there's a chance I may be horribly wrong).

Get the value from the sine table.
If it is negative, we INVERT THE BITS (DON'T v = -v, do v = ~v) and use the slide down tables, otherwise we use the slide up tables.
v = (v*vib_depth+8)>>4;
If v <= 15, do a fine slide by v, otherwise do a normal slide by v>>2.

Test case attached. The frequency values must be EXACTLY the same as each other (ignore the sound btw). This should work even if your frequencies tend to be wrong (it basically does the same math).

Furthermore, the second pattern MUST NOT RETRIGGER. If you add H00 to the ins 02 note on the first pattern, THEN it should retrigger. (Oh yeah, I forgot to mention, IT vibrato hardly ever retriggers - not even with Old Effects or Compat Gxx enabled.)

No.1550/4

Anton — Sun, 30 Sep 2012 12:01:35 GMT

Also, here's a thread from 2005 where mrsbrisby said this problem had been fixed. Did somebody break it again?

http://schismtracker.org/sc/res/473.html

No.1550/3

Anton — Sun, 30 Sep 2012 11:55:33 GMT

I'd like to bump this. I think priority should be set at making the loaders work as they did in old IT, that is loading the left channel as one sample and loading the right channel as the next sample. That way we can make tracks with stereo samples without breaking the .IT format.

On a side note, if Schism allows loading stereo samples as one sample, shouldn't it be saving those tracks as a propietary format, as they wouldn't be backwards compatible with IT?

No.1635/22

Anton — Sun, 30 Sep 2012 11:47:32 GMT

Thanks for the Windows builds Josep!

No.1635/21

JosepMa (Psycle's developer) — Sat, 29 Sep 2012 11:04:45 GMT

I updated them with newer compiles but i couldn't edit the previous post. Since i posted the links in the "sc" board, i didn't care to post them again here.

No.1635/20

Dominick Chan — Wed, 26 Sep 2012 18:02:13 GMT

Have fun: http://psycle.free.fr/josepma/

No.1635/19

brkfstsqd — Sun, 23 Sep 2012 01:15:28 GMT

these links are down, does anyone have any mirrors?

No.1578/9

delt !DEimOs08uM — Thu, 06 Sep 2012 06:16:53 GMT

uh SORRY, i apologize - i just verified, and can also confirm that Josh's bug report above is correct. When first testing, i had completely - and stupidly - disregarded the per-sample "default panning" parameter... (F3, select sample, center of screen)

However, i personally feel this has to be a bug in Impulse Tracker (maybe also and/or from a design flaw in ST3: on/off switch + actual panning value?) Anyway, in most/normal cases, there's no way that a channel marked as S91 would inadvertedly play a sample panned to its marked "default panning" ...haven't done extensive testing with NNA's etc etc.....

... but, the final verdict seems that the above (Josh's) bug report is accurate.

(Need to do more testing.....)

No.1578/8

Saga — Mon, 03 Sep 2012 00:17:55 GMT

I can also confirm that the bug report is correct; I have created a test case in Impulse Tracker 2.14 which shows this behaviour. It has been fixed in OpenMPT some months ago when this problem was reported initially; The test case for verifying this problem can be found here: http://wiki.openmpt.org/Development:_Test_Cases/IT#SmpInsPanSurround.it

No.1656/4

delt !DEimOs08uM — Sun, 02 Sep 2012 19:48:43 GMT

Sounds great, i plan to check it out in the near future. Thanks!
(using linux mainly, will try compiling the source code)

Have you submitted these patches to Storlek?

No.1578/7

Josh — Thu, 30 Aug 2012 01:35:40 GMT

I'm using Impulse Tracker 2.14. In Impulse Tracker, a sample's default panning will override the channel's surround setting. In SchismTracker, the sample's default panning is ignored while the channel is set to surround. To clarify, I'm talking about the Default Pan setting on the Sample List (F3).

No.1656/3

JosepMa (Psycle's developer) — Tue, 28 Aug 2012 21:38:25 GMT

delt: if you're using windows, try the compile i made which is posted here: http://schismtracker.org/scdev/res/1635.html#i1639

If you're using linux, you might want to try downloading the sources from that link and compiling it. A bug was added in the envelope handling when optimizing it, which causes some instruments to stop immediately. My sources have it fixed.

No.1578/6

delt !DEimOs08uM — Tue, 28 Aug 2012 18:10:17 GMT

Josh: i'm sorry, but your original bug report is wrong. Which tracker are you using to compare this behaviour?

This is exactly how panning in Impulse Tracker works like - "surround" (S91) is applied to the channel, not to individual notes. The channel stays in "surround" mode until a panning is set explicitly, either with S8x or Xxx commands, or a panning control in the volume col.

A few years ago, i had suggested to implement "S90" to restore the channel's previous panning (as it was before the S91) but this suggestion was rejected.

No.1656/2

delt !DEimOs08uM — Tue, 28 Aug 2012 18:00:19 GMT

Oh, forgot to mention... i remember in the old dos/it.exe days, i ran into a similar issue, in one of my old .it modules, an instrument played very weakly and muffled – turns out the filter values were all the way down. This isn't the case here, there seems to be nothing wrong with the module itself.

No.1656/1: Bug with instruments, possibly filter/pitch env?

delt !DEimOs08uM — Tue, 28 Aug 2012 17:57:12 GMT

Hello, it's been several months, sorry about that....

There's a bug that as far as i can tell, has been introduced a few months ago – seemingly between the end of 2011 and around may of 2012. Listen to my demo song "secretly wishing" - the instrument #9 (jump to order 5, last two channels) does not play at all. Nothing's been changed in the module, and it plays fine both in Impulse Tracker and pre-2012/05 versions of schism.

In recent versions of scismtracker, the sample plays fine, but the instrument acts like it's not mapped to it.

No.1635/18

esaruoho — Sat, 11 Aug 2012 06:11:38 GMT

>>1654

woah.. ok, i have no way to know how to do that so huh :D

No.1635/17

Mohamed Stephenson — Sun, 05 Aug 2012 09:26:26 GMT

The big problem with OS X is that there's no M4 file installed by default with SDL. You've got to rip it from, say, the Linux version and I think patch it up a bit. I don't think it comes with a sdl-config either, but I don't quite remember.

No.1635/16

JosepMa (Psycle's developer) — Thu, 02 Aug 2012 14:02:29 GMT

On the main page, there's a link to build for OS X, but describes how to cross-compile it from linux.

The link to build for linux has some things that might help on building for other platforms.

Concretely, try:

$ autoreconf -i
$ ./configure
$ make

if autoreconf does not work, then, you might try what I wrote in the compile_win32.txt file (it does the same as autoreconf)

$ aclocal
$ autoconf
$ autoheader

Comment too long; full version here.

No.1635/15

esaruoho — Thu, 02 Aug 2012 11:49:12 GMT

i tried another autoconf and this time got no errors. also:

./configure; make; make install
configure: error: cannot find install-sh or install.sh in "." "./.." "./../.."
make: *** No targets specified and no makefile found. Stop.
make: Nothing to be done for `install'.

so i dunno

No.1635/14

esaruoho — Thu, 02 Aug 2012 11:42:41 GMT

ok installing of SDL.framework changed autoconf errors into:

configure.ac:61: error: possibly undefined macro: AM_PATH_SDL

  If this token and others are legitimate, please use m4_pattern_allow.
  See the Autoconf documentation.

so i guess there's some progress, no other errors

No.1635/13

esaruoho — Thu, 02 Aug 2012 11:41:46 GMT

i installed SDL.framework to /Library/Frameworks and still get the same error.

No.1635/12

esaruoho — Thu, 02 Aug 2012 11:34:06 GMT

ok i was told to run aclocal, which gave

configure.ac:61: warning: macro `AM_PATH_SDL' not found in library

No.1635/11

esaruoho — Thu, 02 Aug 2012 11:15:45 GMT

that ./configure; make; make install just shot me an error and nothin else. well, i'm way out of my league here.

No.1635/10

esaruoho — Thu, 02 Aug 2012 11:13:43 GMT

yeah i have no idea. tried autoconf and got

configure.ac:33: error: possibly undefined macro: AM_INIT_AUTOMAKE

  If this token and others are legitimate, please use m4_pattern_allow.
  See the Autoconf documentation.

configure.ac:40: error: possibly undefined macro: AM_PROG_CC_C_O
configure.ac:49: error: possibly undefined macro: AM_CONDITIONAL
configure.ac:52: error: possibly undefined macro: AM_PATH_PYTHON
configure.ac:61: error: possibly undefined macro: AM_PATH_SDL

No.1635/9

esaruoho — Thu, 02 Aug 2012 11:05:07 GMT

>>1644

how do i build for OS X? i have 10.6.8 so that should be good enough, right? i don't have xcode tho. what should i do?

No.1635/8

JosepMa (Psycle's developer) — Wed, 01 Aug 2012 15:06:39 GMT

What? Why do you need me to run a source repository? The whole source with the modifications are in the third link (first two are binaries and third one is source).

(I only have experience with CVS and SVN, not with mercurial. I installed tortoiseHg which i've used to get the sources of schism).

If you want to contact me directly, you can use hotmail/messenger: jazet_e at hotmail.com (better email me first, i don't add people that i don't know).

No.1635/7

Marie Stokes — Wed, 01 Aug 2012 00:44:26 GMT

>>1643
I have no way at all to build for OS X at the moment. Heck, can't even build for the OS I'm running.

JosepMa: Maaaaaan I'd love to work with you on this stuff. If you can put up a Mercurial repo, I can pull from that and update the official source.

No.1635/6

esaruoho — Sat, 28 Jul 2012 19:08:22 GMT

>>1642

Storlek, any chance you could use JosephMa's stuff and bake an OSX version?
thanks for your work JosephMa. yeah, I used to use zTracker, a friend altered the ctrl-shortcuts to be alt-shortcuts and even did a rudimentary MIDI loader for it, but he never finished it and promptly lost the sourcecode. zTracker was wonderful, so much freedom and so much control. so wonderful. hence why i always get kinda angry when i try to use schismtracker for midi anything since it feels like a drunk guy trying to hobble around as re: midiclock timing (pc/mac)

No.1635/5

JosepMa (Psycle's developer) — Sat, 28 Jul 2012 09:26:42 GMT

Oh, sorry, I didn't think that a separate link to whatsnew was necessary, being the other link just a 1MB .zip file. Anyway, if anyone else needs it, here it is: http://psycle.free.fr/josepma/whatsnew.txt

About your questions:

There is no support for sysex at the output. I believe there's no way to generate them from schism, so supporting it would be moot.
What I've changed in the win32 support is in how midi input understands (as in reads) sysex messages. Yet, I believe that right now, schism does not process those messages, so probably it does not matter. (Note about midi tracker: There was ztracker: http://ztracker.sourceforge.net/ which has an Impulse-tracker like interface and is made exclusively for midi. Maybe those sources could help at some point).

About IP-MIDI, I discovered that it can be disabled from the commandline using the setting –no-network . Other than that, it can be disabled by going to the midi configuration screen (Shift+F1) and set the number of IP midi ports to zero (The former does not start the thread, while the latter leaves the thread started).

IP-MIDI should not interfere with the other midi drivers. Said that, i don't know how the OSX side of the midi implementation is made, so I don't know if it could be improved.

Comment too long; full version here.

No.1635/4

esaruoho — Sat, 28 Jul 2012 05:43:49 GMT

>>1640

actually, even pasting the whatsnew here would be good, so here's my questions:

>Sysex messages might not fully work (does that matter?)

while i'm not certain schism will ever be able to do sysex "properly" (heh), i reckon it might matter to some people - but considering schism isn't a valid midi tracker (by any means!), i guess it doesn't matter if sysex too does not work.

>IP-MIDI caused the high cpu usage (one full core) in Windows (Socket was not being created, and thread became an endless loop). Same thing might happen on other platforms if setting number of IP midi ports to zero. Now I've also prevented this case too.

Comment too long; full version here.

No.1635/3

esaruoho — Sat, 28 Jul 2012 05:39:45 GMT

>>1635

two things. (well, actually three)

1)is there any chance you could make that whatsnew.txt into a clickable URL so one could actually click on it before installing it
2)and what about osx .app compile?
3)when will midi come back, and are you seriously working on making it work better?

No.1635/2

JosepMa (Psycle's developer) — Fri, 27 Jul 2012 17:44:13 GMT

I've worked a bit on the MIDI problem and localized it. It was the IP-MIDI driver the one that was causing the high cpu usage. I've also solved a problem with the win32 midi input driver.

Here are the new links:

Windows 32bit: http://psycle.free.fr/josepma/SchismTracker-hg-JM-mingw-x86-20120727.zip
Windows 64bit: http://psycle.free.fr/josepma/SchismTracker-hg-JM-mingw-x64-20120727.zip
Source code: http://psycle.free.fr/josepma/SchismTracker-hg-JM-src-20120727.zip

No.1636/3

tct1501 !TCpAlqu3Hk — Mon, 23 Jul 2012 10:55:01 GMT

http://schismtracker.org/sc/src/tct-bbc.it for the correct file. Maybe something in the mod data can be compared. I do not know.

No.1636/2

tct1501 !TCpAlqu3Hk — Mon, 23 Jul 2012 10:50:03 GMT

File: 7:10, Compassion of Christbot

Erroneous mod. The RH major can be reset, but the minor...

No.1636/1: Highlight bug

tct1501 !TCpAlqu3Hk — Mon, 23 Jul 2012 10:47:51 GMT

File: 475x205, rowhls.jpg

I stumbled upon some strange row highlight bug. As far as I can tell, I didnt do any intricate stuff when this happened.

This also happened to me with the module I made for HC01. In fact, this 'bugged' mod is in the pack as well.

The only thing I can think of, is that this occurs when deleting unused samples/instruments. Makes sense, since in both cases I was going through this process before saving it to another module.

The erroneous mod with this highlight error is included in the thread.

No.1635/1: New Windows compile with additions

JosepMa (Psycle's developer) — Sun, 22 Jul 2012 17:45:27 GMT

Hello.

I've been implementing some changes into SchismTracker these past weeks and I'd like to share those changes here.

This is based on the source code from the head of the source repository, concretely, revision 2873 from two months ago.

There is an included "whatsnew.txt" file with the changes I've added, which include improvements in S3M/Adlib playback, better FFT and VU visuals and some bugfixes.

This version still has MIDI disabled. Next thing i'll try is fixing the high cpu usage when it is enabled.

Files are:
- Windows 32bit: http://psycle.free.fr/josepma/SchismTracker-hg-JM-mingw-x86-20120722.zip
- Windows 64bit: http://psycle.free.fr/josepma/SchismTracker-hg-JM-mingw-x64-20120722.zip
- Source code: http://psycle.free.fr/SchismTracker-hg-JM-src-20120722.zip

No.1634/1: More tracker compatibility test cases

Saga — Sun, 08 Jul 2012 23:05:31 GMT

I'm currently collecting all my tracker test cases that I've written over the years and posting them all on a wiki page:

http://wiki.openmpt.org/Development:_Test_Cases/IT
http://wiki.openmpt.org/Development:_Test_Cases/XM

More test cases will follow. They are mostly centered around OpenMPT and I don't test all kinds of other software like Storlek did with his test, but most of them are easy to test because you just have to compare if the left and right channel sound identical. :)

No.1578/5

JosepMa (Psycle's developer) — Sun, 08 Jul 2012 21:27:07 GMT

Link removed. There is a new link in the thread pointed by Josh

No.1578/4

Josh — Sun, 08 Jul 2012 20:23:10 GMT

There's more discussion about the Windows version crashing here: http://schismtracker.org/sc/res/5473.html

No.1578/3

JosepMa (Psycle's developer) — Sun, 08 Jul 2012 17:45:14 GMT

Hello.

I've been able to compile a windows build with mingw 3.4.2 (Ok, it's an old version, I know).

I had to do a hack for gettimeofday, but other than that, it works.
I tried to copile an x64 version with mingw 4.7.0, but that one crashes upon startup (in fact, it does mostly what the current windows build does, so if Storlek is building them with a 4.x version, we might be a bit nearer to the clue of why it crashes).

These are the compiler settings used : -Wall -Wstrict-aliasing=2 -Winit-self -O3 -DNDEBUG

It is stored here:
http://psycle.free.fr/josepma/schismtracker-20120105-win32-mingw-3.4.2.zip

No.1628/3

Xavier Brock Case — Thu, 05 Jul 2012 17:57:06 GMT

>>1629
I'm going to be pretty much gone for a while. To put it nicely and mildly, life sucks right now.

No.1628/2

Chelsea Annmarie Ayers — Sun, 01 Jul 2012 06:08:57 GMT

2 months without commits ;_;

No.1628/1: Mx and 3xx in the same XM pattern cell

Saga — Sat, 30 Jun 2012 23:30:11 GMT

So do you really want to know what happens?
3xx is ignored, and the effectiveness of Mx is doubled, i.e. "M1 310" is the same as "M2 ..."
Have fun implementing that in the XM loader. :D

No.1626/2

pkasting — Fri, 29 Jun 2012 22:25:28 GMT

Incidentally, I thought this was working at first and then stopped, so is there a setting I might have accidentally toggled that controls this? Or is my memory faulty?

No.1626/1: "8" to play row doesn't play notes w/o sample/intrument

pkasting — Fri, 29 Jun 2012 22:24:22 GMT

I normally specify an instrument number only at the top of a pattern, leaving the rest of the notes in the pattern with no value in the sample/instrument column.

Using the 20110101 Windows build, pressing "8" repeatedly to play rows only plays the one with an instrument number. Subsequent rows don't sound. This works fine in Impulse Tracker.

No.1613/9

JouluPam — Sun, 10 Jun 2012 18:56:04 GMT

Now it suddenly stopped working again.. Why, oh why?

No.1613/8

Salina Araceli Lindsay — Thu, 07 Jun 2012 05:27:25 GMT

Weird!

No.1613/7

JouluPam — Wed, 06 Jun 2012 15:56:18 GMT

Thanks everybody! Today the - key just started suddenly working again. Thanks for all the help. Hope it doesn't go away :)