This commit allows the user to drag'n'drop DLL files onto VeSTige
instrument. Furthermore VeSTige now registers itself as handler for
loading DLL files.
Try to load all shared libraries before resolving any symbols. Fixes
problems with plugins that are linked against other shared libraries
located in LMMS' plugin directory.
Do not add download ID returned by QHttp::get() to array and remove it
in the finishDownload() slot rather than adding the ID to the array as
soon as the download has finished.
With newer versions of wineg++ (1.1) errorneously adds the suffix ".exe"
to the target binary. This causes "make install" to fail. Therefore
rename remote_vst_plugin.exe if it exists.
Removed obsolete warning about setting realtime. It's neither important
nor will calling printf()/fprintf() in a RemotePlugin print anything in
LMMS' console output.
Some VST plugins save their state as chunks rather than parameters.
According code to save/restore those chunks has been taken from FST 1.9.
However restore currently is disabled as it mostly fails for whatever
reason.
Global master pitch wasn't taken into account correctly when calculating
master key of a given key. Made MIDI based instruments behave in the
opposite direction regarding pitch when changing global master pitch.
Thanks to Skiessi for pointing out this issue.
For some reason, building remote_vst_plugin.cpp with recent
wineg++/gcc-multilib on 64 bit platforms causes undefined symbols:
remote_vst_plugin-JzpHs6.o: In function `std::list<VstMidiEvent, std::allocator<VstMidiEvent> >::_M_insert(std::_List_iterator<VstMidiEvent>, VstMidiEvent const&)':
remote_vst_plugin.cpp:(.text._ZNSt4listI12VstMidiEventSaIS0_EE9_M_insertESt14_List_iteratorIS0_ERKS0_[std::list<VstMidiEvent, std::allocator<VstMidiEvent> >::_M_insert(std::_List_iterator<VstMidiEvent>, VstMidiEvent const&)]+0x35): undefined reference to `std::_List_node_base::hook(std::_List_node_base*)'
Replacing std::list with std::vector for queueing MIDI events seems to
fix this issue. Consider this as a workaround rather than constructive fix.
This commit fixes lockup of the whole audio engine as soon as LMMS'
JACK backend got zombified due to longer delays (e.g. when audio engine
is blocked while loading a project). Restart JACK backend in such cases
and show an according message to the user.
Furthermore this commit includes various cleanups and improvements.
There's been a missing evaluation of _os.depth in the 2nd ternary
operator which caused depth of filetype always be 24 bit even if 16 bit
was selected.
Due to optimizations in the graphics subsystem of Qt 4.5
songEditor::paintEvent() didn't get called regularly anymore and thus
the range of the horizontal scrollbar didn't get updated at all.
This commit makes the GUI track all changes related to the length of
the song.
The slot LocalResourcesProvider::reloadDirectory() contained various bugs
so it did not record changes in filesystem properly. This also fixes lots
of QFileSystemWatcher messages in console window when saving a new project
inside LMMS.
Follow recent changes in ZynAddSubFX CVS repository - ChangeLog says:
29 Mar 2009 (Mark McCurry)
- Started to use Doxygen within the Effects
- Started to use const within Effects
- Changing tabs->four spaces in hopes of generating a bit more
consitancy
- Began to use Initialization Lists
- Almost all changes contained in Effects until further
discussion on the style, so consistancy can be reached
Updated 3rdparty libsamplerate to version 0.1.7 and bumped requirement
to system's libsamplerate to version 0.1.7.
Besides minor stuff, version 0.1.7 has one important fix:
* src/src_sinc.c
Fix a segfault which occurs when memcpy is passed a bad length parameter.
This bug has zero security implications beyond the ability to cause a
program hitting this bug to exit immediately with a segfault.
Up to now changes of samplerate or tempo only have been tracked by the
VeSTige plugin while e.g. the VstEffect didn't forward such changes to the
RemotePlugin. Moving according code from VeSTige to VstPlugin base class
fixes this issue.
Most of the MDA VST plugins require the SplittedThreading model in order
to run properly. All plugins whose unique ID starts with "mda" will be
driven with the SplittedThreading model.
When doing FM/PM, the value of sub-oscillator is globally/locally added to
phase. However when doing this, a ratio between fixed default samplerate and
actual samplerate needs to be taken into account in the phase calculation.
This commit fixes missounding render output when rendering projects with
FM/PM inside at higher samplerates. (Thanks to Skiessi for his ladspa2.mmpz
which clearly brought up this bug)
This commit fixes non-usable instrument bank browser in ZynAddSubFX,
especially in win32 version. Adding the ZynAddSubFX preset directory (inside
LMMS' preset directory) to config.cfg.bankRootDirList allows easy instrument
browsing using ZynAddSubFX' instrument bank browser (closes#2627950).
Calling alignedMemClear() when getNextBuffer() returns 0 is superfluous and
causes crashes in some situations. Simply get out of the outer loop as soon
as quit==true.
When global samplerate changes, settings of a LADSPA effect are stored in
a temporary object, the effect is reloaded and the settings are restored.
However, controller connections weren't restored properly as the connections
were not finalized. This commit fixes the issue.
When importing MIDI files notes were always mistransposed by 9 semitones.
This commit fixes correction of notes' keys so imported MIDI files sound
just like they should.
In Qt3 we had to take care of drawing background properly if there were
semitransparent child widgets. This code had been ported to the Qt4 version
but is obsolete and causes a crash when for example dragging a Vibed preset
onto an existing instrument track. This commit removes this code.
(closes#2688270)
If SF2_PANNING_SUPPORT is set, overlapping notes of the same key played on
different MIDI channel caused some notes to hang. This commit fixes the issue.
If SF2_PANNING_SUPPORT is set, changed patches were applied to current
channel after calling fluid_synth_noteon(...) - some notes still were played
with the old patch until m_channel made one round trip.
This commit ensures that all defined data types in include/lmms_basics.h are
suffxed with "_t" to keep consistency as well as to fix issues on systems
with older STL/libstdc++.
check for nodeName being "connection" - do not try to load an instrument
called "connection" if one of the basic parameters of the InstrumentTrack
is automated
