I was working on a module yesterday, and decided to book it and work on it later. When I loaded the module earlier today, it kept crashing after it got to a specific part in the song.

A friend and I have determined that it is the pitch sequence on instrument 03 (the snare drum) -- It would crash the tracker if played by itself only, and would not crash if the pitch sequence was changed.

Attached below is the module with the bug (in a .zip file, because Github doesn't accept .BTM files), I'm not sure if this is an adequate bug report or if this has been reported before - if it has, my apologies.

TEST.zip

More unused j0CC ideas:

• Early Note effect ^xx similar to 0Gxx but reversed
• Define regions where row density is increased *= 2 (for 1/8 beat) or 3 (triplets) or 3/2 (triplets but possibly harder to implement and keep aligned when the region ends)
• Multiple rows at the same time, but affected independently by 0Gxx or ^xx.
  • unsure if useful in bambootracker. definitely would be useful for manipulating nes dpcm delta-counter.

gui/mainwindow.cpp: In member function ‘void MainWindow::on_instrumentListWidget_customContextMenuRequested(const QPoint&)’: gui/mainwindow.cpp:944:8: error: ‘class QAction’ has no member named ‘setShortcutVisibleInContextMenu’; did you mean ‘setShortcutContext’? edit->setShortcutVisibleInContextMenu(true); ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ setShortcutContext [...] Makefile:1639: recipe for target 'mainwindow.o' failed make: *** [mainwindow.o] Error 1

make produces lots of warnings related to mainwindow.cpp and one error above. OS is Ubuntu 18.04 LTS, all dependecies installed, I think.

Looking at a discussion fa213ae there is a need for that.
Compiling from source on Windows is a chore, compared to Linux distros, since:
-there is no centralized package manager for distributing development libraries (Qt?)
-MinGW/MSYS can be a hassle

CI (continuous integration) services like AppVeyor can be used to test/build source code on Windows VM and then deploy the fresh devbuilds in binary form without a need to wait for a full release.
https://www.appveyor.com/docs/

Hi,

Was working on something for a couple hours. Not sure what happened but all of a sudden I am no longer able to reopen my .BTM file as it appears with an error:

I have attached the relevant .BTM file in the following zip.
djtBMX_BTM_bug.zip

Not sure what could suddenly cause this but I suspect it might have something to do with deleting and adding instruments?

Thanks,
-MaliceX

Ubuntu 18.04 x64, master commit 1d5cf1b + 1 unrelated instrument editor change.

sd3winter-dev.btm.zip

• Hold Shift+Up and allow the cursor to wrap to the bottom of the song. Selection does not appear.
• Hold Shift+Right and selection appears.
• Ctrl+X to initiate program crash.
• Debugging in Qt Creator is impossible unless you remove try-catch from main().

int BambooTracker::getStepNoteNumber(int songNum, int trackNum, int orderNum, int stepNum) const, stepNum = 32 and my song has 32 rows/pattern.

I'd like to compile this on my own, to see how well compiling on Linux goes.

Could we possibly get a list of required libraries/programs and compile instructions?

I opened BambooTracker and still getting this message.
" The procedure entry point "CancelIoEx" could not located in the dynamic link library "Kernel32.dll" "

I should not work on Older machines. :(

Sorry for my old pcc And I didn't get new one

I don't know how it happened or is it a real bug.
I made a little test - just 2 patterns, one was empty, two instruments (one was FM, second was an unused SSG instrument),50 Hz tick freq, few notes there and there and I saved result. Now when I try to open it I only see this nonsense:
array::at:__n (which is 255) >= _Nm (which is 128)

If it's any help, there is a module in question - kona.btm.zip

This tracker is surely promising, but i've noticed some issues:

1. After adding my first instrument and messing over with the parameters, i created a new one and somehow it overrides with the other instrument, it seems that the operators apply to every instrument you create at the moment.
2. The lack of the ADPCM channel, i'm sure this isn't really a problem but just letting myself know about that.
3. The volume column seems to scale volume in a reversed order. 00 being the loudest and 7F being the quietest.
4. Are you going to have support for loading Deflemask (.dmp) patch files?

When exporting a file, Bamboo auto-fills the "System" box with PC-9801. This isn't necessary accurate.

When FM/SSG levels are matched, it's nearly exactly the same as an OPNA with no resistor on the SSG output and no amplification. On initial startup, Bamboo's 0/0 mixing is similar (if not the same as) a PC-88 with a Sound Board 2. With the SSG set to -3dB relative to FM, the output is again, very similar to (if not the same as) a PC-9801-86 board and one of my own Makoto boards for MSX. Makoto uses Yamaha's suggestion of a 470 ohm pulldown on the SSG output, and it's likely that the -86 does as well, making this the "proper" PC-98 mixing level.

Adding Mix Select boxes in the module properties menus for Sound Board 2 and 9801-86/Makoto guarantees proper sound when playing .vgm or .S98 on hardware with these platforms without a user having to enter the menu and reset it themselves.

This does not mean the options for arbitrary mixing should be removed, only to add selections for proper hardware levels.

Hello

It's great to have licenses/list.md

however qt is not in there as sources, afaik, so it's just linking against it.

do you have parts of source codes of mame, vgmplay, silk icons? if so, where in the directory tree?

i need to specify them in debian/copyright, example:

https://sources.debian.org/src/protracker/2.3d.r147-1/debian/copyright/

Yet another compilation problem - this time something is very wrong with SCCI, as it produces a lenghthy list of errors in scci.h:
https://gist.github.com/papiezak/5df54cea1d5d66ff480b641b2e875bd0

Ubuntu 18.10, all dependencies installed. There were no build problems before cf99e95

This is a bit hard to explain, but to put it shortly, some envelopes values are set in the wrong order when a new instrument is loaded as an .ins file, which often results on really bizarre sounds.

This screenshot shows the envelopes of a bass drum instrument. One window shows the incorrect settings, and the other showing how it should be. To "fix" this problem, I had to load the mentioned instrument through DefleMask, save it as a .dmp file, and then load said file to BambooTracker in order to get the correct settings.

I'm not sure if this problem is also present when other instrument file types are loaded, but yeah.

Here's another attachment, it has a .btm file to show the sound of both the incorrect and correct bass drums. It also has the bass drum instrument, as both .ins and .dmp files:
BambooTracker FMEnvelopeBug.zip

• Move cursor to right, not down, after entering effect type (possibly effect param as well)
• Shortcut to "jump up/down to next highlighted beat"?
• Editable step? 0 or greater.

When editing sequences in the instruments, dragging the scrollbar works properly, but the arrow buttons reverse up and down.

Applies to FM: TL, AR, DR, SR and applies to SSG: Tone/Noise, Envelope, Arpeggio, and Pitch.

When editing either SSG or FM, using the instrument in a channel of a different type (for example, typing a note in SSG with an FM instrument selected by accident) crashes the tracker immediately.

Let's compare Deflemask's YM2612 and Bamboo Tracker outputs:
Bamboo: https://instaud.io/359e (F#3, effect 02xx)
Deflemask: https://instaud.io/359f (the same)

As you can hear, OPN2 core in Deflemask goes much lower than Bamboo's OPNA, while they should sound pretty much the same. It seems that in Bamboo Ym2608 is one octave higher than it should be (C-0 is C-1, C-1 is C-2 and so on). Therefore there is no lowest octave. Some drums such as kicks or snares from PC-98 games are impossible to replicate here.

Steps to replicate:

1. Open a module :http://cherry.dhcp.phys.ethz.ch/majfant2quieter.btm
2. Play it
3. Edit the first instrument by changing FB value to 0
4. Hit play button once again.
5. With any luck, the song will become much louder, almost like every instrument was playing at max volume

I don't know if this is a bug or just things not being implemented yet, but

only main board / works. Number row * (Shift+8) does not, and neither does numpad / or numpad *.

Setting the Buffer length to 1ms followed by confirming and reopening the configuration dialogue will display TextLabel as the currently selected Buffer length. Doesn't seem to have any effect on playback though.

Are there any sample songs available? I've only found:

http://cherry.dhcp.phys.ethz.ch/majfant2quieter.btm

SSG 2 and 3 apparently end up looping the current pattern on editor when exporting to VGM and WAV
Not sure if this happens in all modules but it doesn't seem like there's anything I've done that caused it.

The track is supposed to sound like this (recorded through an external program)
https://dl.dropbox.com/s/qquzuh0ur9khfke/bamboo%20tracker%20recording.mp3
But when exporting to wave while the first pattern is selected, the WAV/VGM output sounds like this
https://dl.dropbox.com/s/zqao8zp1eatozrh/bamboo%20tracker%20wav%20export%201.mp3
When the second pattern is selected, it sounds like this
https://dl.dropbox.com/s/nx6jyoy7rivplw9/bamboo%20tracker%20wav%20export%202.mp3

Here's the module file
begin bossa.zip

Let's compare for example hi-hat instrument from EVE: burst error. In Hoot player it sound like this:
https://instaud.io/2ZYr
Yet in Bamboo Tracler, the same instrument sounds completely different:
https://instaud.io/2ZYs

What is the reason of this difference? There are several percussive instrument that sound slightly off, at least compared to Hoot.

Instrument in question: Closed Hi-hat.zip

Currently bambootracker, like ft, seems to use per-channel pattern indexing. If I use the same pattern in multiple order-rows, I frequently end up changing music I wrote previously (almost universally above my current position) and did not intend to change.

I also sometimes make the same mistake when reusing instrument sequences.

BambooTracker

• Pressing Insert multiple times in order editor results in pattern numbers "0,6,5,4,3,2,1". Maybe Insert should move to next order-row.
• All instruments default to sequence 0. (At least I appreciate the "Users: ..." field)
• Pressing Esc doesn't close instrument editor (bug or intentional?)

Inheritance

(Basically these are ideas I wanted to implement into j0cc-ft, but never got around to doing the refactoring. Hopefully this codebase is newer and easier to work with. I can help write the code if you want.)

You can substitute order/instrument and pattern/sequence:

My ideas is that each order orders[i0] "owns" an array of patterns/sequences. To reuse one order's pattern, the other order orders[i1] "inherits" the pattern orders[i0][chan].

My concept for inheritance/references is that orders[i1][chan] is a Union[pattern, another order] (tagged union, sum type, std::variant, etc).

• Simple approach: Set orders[i1][chan] = Ref(i0) for some optional Ref wrapper class.
• To allow for reordering orders[], maybe orders[i0] == uuid0 and refers to patterns[uuid0][chan] for all chan.
  • patterns is a hashmap[uuid] -> hashmap/list[chan] -> pattern. IDK if you use enums or strings/etc. as channel numbers.
  • If (for specific chan) pattern [i1, chan] inherits from [i0, chan], orders[i1] == uuid1 and patterns[uuid1][chan] == uuid0 and points to orders[uuid0][chan]

My experiences with FT

This solves the issue where I create several instruments which all share sequence 0, then I enable/edit sequence 0 multiple instruments. (In 0CC, I can workaround by pressing "select next empty" whenever enabling a checkbox.)

This also solves the issue where adding an expansion chip defaults to all order[row][chan] = 0. (Which could be solved easily by setting order[row][chan] = row.)

0CC has "populate unique patterns" and "merge duplicate patterns". Often I reuse pattern 0 (with volume/effects) in multiple places, populate unique, then only change volume/effects in 1 place but not all copies. (I just did this a few days ago).

Avoid automatically modifying parent when editing child

When writing patterns, I often edit a lower pattern and accidentally mangle a higher one (earlier in time, composed earlier). (I don't know if I often edit an above pattern and accidentally mangle a lower one).

When copying instruments, I used to shallow-clone an instrument and edit the copy, mangling the parent. (Now I usually deep-clone, but sometimes I still make this mistake and open the file in another 0cc window to recover my unmodified instrument. I wish instrument undo existed.)

• Solution: Clearly mark inherited patterns using a different background color (both in order table and pattern editor?).
  • Idea: Should parent patterns be marked as well? I don't know if accidentally editing the parent and mangling a child happens often, or if deliberately editing all similar tracks is more likely.

Editing children should not change the parent. Choices:

• Copy on write? Any change to child pattern creates a distinct pattern.
• Extensible children? (more applicable to patterns than instrument seqs)
  • Append extra effects into existing rows. Effects only exist in child, and add to (do not overwrite) existing effects.
  • Erase specific cells (effects).
  • Alter notes (I don't like this idea).

When changing mixing settings, they're not saved properly. SSG always resets to 0dB when reopening the tracker. FM resets itself to +9 for values above +9, +6 for values between +3 and +6, and 0dB for values below +3dB.

I'm fairly new to FM, just tried using bambootracker, now I'm trying to switch algorithms. Currently alg4 has op2 as one output, but many other algorithms only support op4 as output. Is it useful for people other than me to be able to switch operator settings around?

Solutions:

• Ability to copy all FM envelopes as text: global settings on first line, 4 ops on subsequent lines
  • open text editor and move lines to move operators around
• copy/paste buttons for individual ops?

I'm experiencing a noticable stuttering of the Pattern Editor when playing a song in Follow Mode with any Buffer length >3ms.

Here is a .WEBM showcasing this problem.

(This is on Linux, NixOS 18.09, running the freshly compiled release v0.1.4)

Display order numbers and the order count on the status bar in hexadecimal.

Unsetting the "Show row number in hex" option only affects the currently selected order in the order list.

It doesn't affect

• the status bar display
• the numbering of all orders in the order list that aren't the currently selected order

A screenshot (working on German i18n, but the bug's replicable without it)

System:

• Ubuntu 18.04.2 LTS x86_64
• Commit 4be0b10 (behind the Appveyor commit, but I doubt it had any effect either)

The keybinds for adding notes, as the README mentions, are:

There are two rows of a piano keyboard:

Current octave+1
 2 3  5 6 7  9
Q W ER T Y UI O

Current octave
 S D  G H J  L
Z X CV B N M, .

This works fine on a QWERTY keyboard.

Using a non-QWERTY however, like the German QWERTZ, French AZERTY or, well, the Russian JCUKEN using Cyrillic script can either take some getting used to or might just not work at all. (I can only speak for myself, using a QWERTZ keyboard.)

I propose that, at least, these keys should be rebindable. They can stay the default, I highly doubt that the majority of users will be using a non-QWERTY layout, but the user should be able to change the mapping and save it how they see fit. The groupbox for changing the keys to change the current octave etc. could be moved into the same settings window tab as well.

Perhaps offering some presets for different layouts as well would be the icing on the cake.

Regarding the following file: https://ocv.me/tmp/1.btm

When opening this file, BambooTracker crashes. Here is the output:

Qt has caught an exception thrown from an event handler. Throwing
exceptions from an event handler is not supported in Qt.
You must not let any exception whatsoever propagate through Qt code.
If that is not possible, in Qt 5 you must at least reimplement
QCoreApplication::notify() and catch all exceptions there.

Segmentation fault

The error is somewhere in FileIO::loadInstrumentPropertySectionInModule, inside case 0x00: // FM envelope when i == 4.

Please do not try to rescue data from this file. I am only uploading this file in case you need sample data to help reproduce this crash :-)

I was playing around with all of the operator parameters, and may have created an instrument which is nonsensical/illegal. One of the things I recall doing was to create an LFO/Operator sequence for instrument 00. Parameters: ☐ LFO, Operator: AL, ☑ Sequence, and the sequence itself was [2 4 6] (Loop). When creating a blank song with this configuration, a similar crash is reproduced.

Hey, good job on the first release! I'm surprised by how much work did you put into the SSG portion, but in my opinion you missed (or maybe didn't implement yet, who am I to know) two thhings that would make using the SSG that much easier:

• Envelope autotuning (AutoEnv)
  If you set the SSG instrument to use an envelope (saw/triangle), you can control its pitch in the pattern window. However, if you choose a tone+envelope instrument (Tri w/Saw w), only the tone pitch gets modified while the envelope pitch stays static. It would make things easier if you could add a flag that sets the envelope pitch along with the tone pitch (so when you play F-2, both envelope and tone pitches are F-2). Moreover, if that gets implemented, an ability to change the pitch ratios (for example ratio 1:2 would play the tone at F-1 while the envelope is played at F-2,like in Vortex Tracker 2) would be highly appreciated.
• Noise+tone/envelope combos
  One thing that has been omitted was the combination of Noise with Tone or Envelope (or both). It makes for really interesting sounds. It actually is there, I just couldn't manage to find it. (If anyone is still interested, set the Waveform to Tri or Saw, and set the Tone/Noise to T+N)

I haven't digged into the FM portion quite yet (because i'm a total n00b when it comes to FM), but so far i like where it's going!

Is it possible to output/export to MML? It would make it a lot easier to play on a real OPNA

Would it be possible to include this file as BambooTracker.png in the icon directory?

I've created a FreeBSD port that is currently fetching the file separately to use with the .desktop file.

I have created an icon for BambooTracker that I would like to propose.
It's drawn in a pixel-art style, using only colours from the 12-bit master palette of the PC-98.
The 16x16 version is unique, while all other sizes are integer upscales of the 32x32 version.

I have attached a preview image and the .ico file in a ZIP archive.

BambooTracker icon.zip

Hi, I would like to provide a French translation for the GUI, but it seems that there isn't a way of doing so yet. Could you please add it?

A good feature to add would be support for using MIDI keyboards to enter notes.

This segment of code:

https://github.com/rerrahkr/BambooTracker/blob/b315fb027f3c8c3fb5f3c609f8f35c070b70858a/BambooTracker/opna_controller.cpp#L1338-L1340

contains a logical error according to Valgrind, which it found when playing a module:

==29767== Conditional jump or move depends on uninitialised value(s)
==29767==    at 0x1FF633: OPNAController::keyOnSSG(int, Note, int, int, bool) (opna_controller.cpp:1338)
==29767==    by 0x1807E5: BambooTracker::readSSGStep(Step&, int, bool) (bamboo_tracker.cpp:1780)
==29767==    by 0x17FC99: BambooTracker::readStep() (bamboo_tracker.cpp:1656)
==29767==    by 0x17B49D: BambooTracker::streamCountUp() (bamboo_tracker.cpp:1132)
==29767==    by 0x1427D1: MainWindow::onNewTickSignaled() (mainwindow.cpp:2088)
==29767==    by 0x162068: QtPrivate::FunctorCall<QtPrivate::IndexesList<>, QtPrivate::List<>, void, void (MainWindow::*)()>::call(void (MainWindow::*)(), MainWindow*, void**) (qobjectdefs_impl.h:136)
==29767==    by 0x1609F4: void QtPrivate::FunctionPointer<void (MainWindow::*)()>::call<QtPrivate::List<>, void>(void (MainWindow::*)(), MainWindow*, void**) (qobjectdefs_impl.h:169)
==29767==    by 0x15EA97: QtPrivate::QSlotObject<void (MainWindow::*)(), QtPrivate::List<>, void>::impl(int, QtPrivate::QSlotObjectBase*, QObject*, void**, bool*) (qobject_impl.h:120)
==29767==    by 0x63BC6CE: QMetaObject::activate(QObject*, int, int, void**) (in /usr/lib/x86_64-linux-gnu/libQt5Core.so.5.9.5)
==29767==    by 0x312984: AudioStream::streamInterrupted() (moc_audio_stream.cpp:142)
==29767==    by 0x1A365C: AudioStream::AudioStream(unsigned int, unsigned int, unsigned int, QString)::{lambda()#1}::operator()() const (audio_stream.cpp:19)
==29767==    by 0x1A4407: QtPrivate::FunctorCall<QtPrivate::IndexesList<>, QtPrivate::List<>, void, AudioStream::AudioStream(unsigned int, unsigned int, unsigned int, QString)::{lambda()#1}>::call({lambda()#1}&, void**) (qobjectdefs_impl.h:130)
==29767==  Uninitialised value was created by a heap allocation
==29767==    at 0x4C3017F: operator new(unsigned long) (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==29767==    by 0x186BEA: std::_MakeUniq<OPNAController>::__single_object std::make_unique<OPNAController, unsigned int const&, unsigned int, unsigned long>(unsigned int const&, unsigned int&&, unsigned long&&) (unique_ptr.h:825)
==29767==    by 0x174B85: BambooTracker::BambooTracker(std::weak_ptr<Configuration>) (bamboo_tracker.cpp:27)
==29767==    by 0x165041: void __gnu_cxx::new_allocator<BambooTracker>::construct<BambooTracker, std::shared_ptr<Configuration>&>(BambooTracker*, std::shared_ptr<Configuration>&) (new_allocator.h:136)
==29767==    by 0x164597: void std::allocator_traits<std::allocator<BambooTracker> >::construct<BambooTracker, std::shared_ptr<Configuration>&>(std::allocator<BambooTracker>&, BambooTracker*, std::shared_ptr<Configuration>&) (alloc_traits.h:475)
==29767==    by 0x163833: std::_Sp_counted_ptr_inplace<BambooTracker, std::allocator<BambooTracker>, (__gnu_cxx::_Lock_policy)2>::_Sp_counted_ptr_inplace<std::shared_ptr<Configuration>&>(std::allocator<BambooTracker>, std::shared_ptr<Configuration>&) (shared_ptr_base.h:526)
==29767==    by 0x161C4E: std::__shared_count<(__gnu_cxx::_Lock_policy)2>::__shared_count<BambooTracker, std::allocator<BambooTracker>, std::shared_ptr<Configuration>&>(std::_Sp_make_shared_tag, BambooTracker*, std::allocator<BambooTracker> const&, std::shared_ptr<Configuration>&) (shared_ptr_base.h:637)
==29767==    by 0x1606A3: std::__shared_ptr<BambooTracker, (__gnu_cxx::_Lock_policy)2>::__shared_ptr<std::allocator<BambooTracker>, std::shared_ptr<Configuration>&>(std::_Sp_make_shared_tag, std::allocator<BambooTracker> const&, std::shared_ptr<Configuration>&) (shared_ptr_base.h:1295)
==29767==    by 0x15E774: std::shared_ptr<BambooTracker>::shared_ptr<std::allocator<BambooTracker>, std::shared_ptr<Configuration>&>(std::_Sp_make_shared_tag, std::allocator<BambooTracker> const&, std::shared_ptr<Configuration>&) (shared_ptr.h:344)
==29767==    by 0x15D135: std::shared_ptr<BambooTracker> std::allocate_shared<BambooTracker, std::allocator<BambooTracker>, std::shared_ptr<Configuration>&>(std::allocator<BambooTracker> const&, std::shared_ptr<Configuration>&) (shared_ptr.h:691)
==29767==    by 0x15A954: std::shared_ptr<BambooTracker> std::make_shared<BambooTracker, std::shared_ptr<Configuration>&>(std::shared_ptr<Configuration>&) (shared_ptr.h:707)
==29767==    by 0x12FCD3: MainWindow::MainWindow(QString, QWidget*) (mainwindow.cpp:55)

It doesn't appear to cause any harm, but it should be investigated nonetheless. If this error is irreplicable on your end, I can attach 2 modules that successfully triggered this error for me.

(OS: Ubuntu 18.04.2 LTS x86_64)

You've probably thought about some of them but just to reinforce

1 - A way to remap key commands
I use a laptop without a numeric pad, so I don't have access to the default keys to change octaves, also for me it's not intuitive to have - as a key-off

2 - Volume commands
While I know that the TL values on FM makes lower numbers louder, most FM editors I know account for that and make reverse this order. I think your tracker would really benefit from it, mainly considering the SSG and RHY have louder values on higher numbers as well.

3 - Change the order of input for columns that require 2 digits
Considering the SSG and most effect commands use just "00-0F", I think altering the last digit with the first input would make it faster to use, so one just needs to input 1 digit to achieve the effect. Deflemask handles this well.

4 - Instrument editor
I really like how it looks and how it works, but it's hard to input precise numbers on the values that have more of them, like DR/SR and TL. A way to either slow down the cursor (like pressing CTRL on VOPM) or just letting me input directly the numbers would fix it.

As a side note: please make the tracker able to load .OPM and .TFI files, those are the most common FM sound formats, OPM is text-based so it should be the easiest to inplement.

5 - Envelope Reset
The OPNA doesn't reset the envelopes by default before key-off so I think this is a nice feature, but turning the option off will just make every note legato, rather than not resetting the envelope. I think legato should be a different option from envelope reset.

I really liked your tracker and I'm surprised on how much it has for a first release, I hope you'll continue to improve even further! (I also hope you plan on making it output a playable format for the PC98, since there's no VGM players for it yet....)

Considering there's a lot of plain text instrument formats, like OPM, FMDrive and MML formats, I think it would be nice to have a way to support them and I think pasting from clipboard would be the easiest one. So I was thinking of a thing like Right Click -> Paste From... -> (List of formats)

OPM and MML formats such as PMD, FMP and MUCOM88 would probably be the easiest to implement since the FM instrument settings are in a similar order despite there being differences between them.

So here are some examples of the same instrument in different formats:

BambooTracker clipboard instrument definition (spacing added for definition):

            FB ALG
FM_ENVELOPE: 0,  5,
AR DR SR RR SL TL KS ML DT SSGEG
31,22,12, 6, 6, 0, 1, 0, 1,-1,
31,17,12, 9, 6, 4, 1, 6, 5,-1,
31, 8, 2, 9, 5, 4, 1, 0, 7,-1,
31,31, 5,12, 0, 0, 1, 0, 3,-1,

VOPM

@:0 CH1
//  LFRQ AMD PMD WF NFRQ
LFO: 199  64  64  2    0
// PAN FL CON AMS PMS SLOT NE
CH: 64  0   5   0   0  127  0
//  AR D1R D2R RR D1L  TL KS MUL DT1 DT2 AMS-EN
M1: 31  22  12  6   6   0  1   0   1   0      1
C1: 31  17  12  9   6   4  1   6   5   0      1
M2: 31   8   2  9   5   4  1   0   7   0      1
C2: 31  31   5 12   0   0  1   0   3   0      1

PMD

;   ALG FB
@ 0   5  0
; AR DR SR RR SL  TL KS ML DT AMON
  31 22 12  6  6   0  1  0  1    1
  31 17 12  9  6   4  1  6  5    1
  31  8  2  9  5   4  1  0  7    1
  31 31  5 12  0   0  1  0  3    1

FMP

'@ 0 A
   AR DR SR RR SL  TL KS ML DT AMON
'@ 31,22,12, 6, 6,  0, 1, 0, 1, 1
'@ 31,17,12, 9, 6,  4, 1, 6, 5, 1
'@ 31, 8, 2, 9, 5,  4, 1, 0, 7, 1
'@ 31,31, 5,12, 0,  0, 1, 0, 3, 1
  AL FB
'@ 5, 0

MUCOM88

  @ 0:{
; FB  AL
   0,  5
;  AR  DR  SR  RR SL TL KS ML DT
  31, 22, 12,  6, 6, 0, 1, 0, 1
  31, 17, 12,  9, 6, 4, 1, 6, 5
  31,  8,  2,  9, 5, 4, 1, 0, 7
  31, 31,  5, 12, 0, 0, 1, 0, 3,"??????"}

The common thing is that only BambooTracker supports SSGEG on the instrument definition, but it doesn't support AMON/AMS-EN, in OPM/PMD/FMP it should put a "-1" instead of the AMON definition. Since MUCOM88 doesn't use this as well, it should just add a -1 after the end of each line.

Are you interested in adding a text-based export format like FT? Possibly as a native open/save file format (not just import/export) so opening a text and pressing Ctrl+S produces a text file instead of .btm?

Things I would improve from FT text export:

• 1:1 mapping between orders and visible text (achievable through #14 )
• Removing pattern numbers? Maybe makes diff/merge easier, maybe makes parsing complicated and error-prone.

Kubuntu 18.04 x64, Qt 5.9.5.

I cannot add new orders by right-clicking the black space around the orders. Instead the menu is grayed out without an indication as to what's wrong.

Instead, I have to click the actual numbers. I find this unintuitive.

Is there a keyboard shortcut to switch between order and pattern editor?

In 0CC/FT, right-clicking channel headers gives options to toggle/solo channel. (So does right-clicking the pattern itself, though the menu is different.)

In Bamboo:

• Right-clicking the pattern has the same menu as the header.
• "Mute/Solo Track" affects the channel under keyboard caret, which (unlike text editors) does not move to the mouse location. It should affect the channel under mouse cursor.
• When I "Solo Track" twice on different tracks (keyboard caret), the second time it unmutes all instead of soloing my new track.
• (bug or not?) 0CC "toggles solo" when double-clicking a channel header quite quickly (not slowly). Bamboo doesn't handle that case.

Bamboo has "mute track" (less useful than left-clicking) while 0CC has "toggle track" (identical to left-clicking). Neither tracker opens right-click menu when I press the keyboard Menu key, making the menu item useless (except 0CC but not Bamboo tells you the corresponding keyboard shortcut).

• I think Bamboo should display shortcut text for each menu action. it aids discoverability (https://news.ycombinator.com/item?id=18584381)
• Also, Bamboo should underline menu items by prepending & to letters. &Action becomes _A_ction and I can open the menu, then press A to activate that quickly from the keyboard. I feel it's faster than using a mouse. (I noticed the main menu already has underlines.)

Hey. Thanks for trying out. So first to all I found the issue

1. Open Bamboo, go instruments, right click to add or loading 2 instruments, Right click at the first instrument, edit name, when renaming instrument, go to cursor instrument 2, right click on the instrument 2, Select any options and it crash. :(

2. Click ... Or module properties, at the insert song on title write something, and press enter. And it terminated with error code.

English broken. Sorry

Hi. The program raises out_of_range from vector and crashes.
It seems in relation with SSG and the arpeggio sequence.
The problem is reproducible all the time in current BambooTracker 3e62c65.

It occurs at the end of the song file. Also attached is a Gdb trace with breakpoint on exception thrown.
Bug-MODULE.btm.gz
Bug-TRACE.txt.gz

Basically, the effect used to jump into an order doesn't seem to function properly when the module is exported to S98. It works fine when exported to VGM, however.

I'll be attaching the .btm module here, along with the .s98 and .vgm exports:
Bambootracker 0Bxx Bug Modules.zip

While the current system of storing the configuration config.json in the current directory works well enough, it would be desirable to save it to and load it from a fixed directory in a cross-platform way (meaning no hardcoded Windows path or otherwise Windows-only solution).

Using QSettings in IniFormat (or NativeFormat, which uses the Registry on Windows) seems convenient for this purpose:
http://doc.qt.io/qt-5/qsettings.html

This isn't anything super urgent, so it doesn't require to be implemented immediately, but it'd be nice to see in the future :)

I am using two separate audio outputs in Windows: one for headphones, and another for speakers. I never have both enabled at the same time, but I switch between them a lot.
BambooTracker fails to switch devices correctly.

Steps to reproduce:

• Have only one sound device enabled. Bamboo will produce sound.
• Close Bamboo
• Enable another sound output device and disable the one that was previously used.
• Open Bamboo. Now it will not produce sounds.
• To bring back the sound, open the Configuration window, go to the Sound tab and press OK.

I'm no coder, but I think this could be fixed by making Bamboo default to the, well, "default" (for Windows) sound output.

On Ubuntu 18.10 Linux I can build and start BambooTracker, when I load a .btm file I can't play it message in subject appears.