ViewVC logotype

Contents of /linuxsampler/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 2377 - (show annotations) (download)
Thu Oct 4 18:16:26 2012 UTC (7 years, 9 months ago) by schoenebeck
File size: 50923 byte(s)
* Various "const" and "restrict" optimizations.

1 Version CVS HEAD (?)
3 * packaging changes:
4 - fixed building with newer MinGW-w64
5 - Mac OS X: support the new dir for Core Audio SDK
6 - Mac OS X: fixed building outside source directory
7 - made sure all source files for hostplugins are included when
8 doing "make dist"
9 - removed empty directories left from the cvs to svn migration
10 - fixed compilation with gcc 4.6.1
11 - another "make dist" fix, for the LV2 plugin
12 - made --enable-pthread-testcancel default on Mac OS X
13 - minor Makefile fix for building lscpparser when build and source
14 directories are separate
15 - Mac OS X: made it possible to specify plugin installation dir to
16 configure
17 - Mac OS X: Makefile fix for the install-strip target
18 - fixed compilation with gcc 4.7
19 - fixed configure script error with old autoconf versions
20 - lsatomic.h: use gcc provided atomic functions if building with
21 gcc 4.7 and C++11
22 - modernized configure script
24 * general changes:
25 - Refactoring: moved the independent code from
26 the Gigasampler format engine to base classes
27 - command line option '--profile' is currently disabled, since the
28 respective profiling code is currently broken
29 - Introduced support for optional environment variable
30 "LINUXSAMPLER_PLUGIN_DIR" which allows to override the directory
31 where the sampler shall look for instrument editor plugins
32 (patch by Luis Garrido, slightly modified).
33 - implemented Roland GS NRPN 1ArrH which allows to set volume per note
34 - implemented Roland GS NRPN 1CrrH which allows to set pan per note
35 - implemented Roland GS NRPN 1DrrH which allows to set reverb send per
36 note (in this implementation of the sampler its simply hard routed to
37 the 1st effect send of the sampler channel, no matter what the actual
38 effect type is)
39 - implemented Roland GS NRPN 1ErrH which allows to set chorus send per
40 note (in this implementation of the sampler its simply hard routed to
41 the 2nd effect send of the sampler channel, no matter what the actual
42 effect type is)
43 - implemented support for internal LADSPA effects
44 - raised limit of program change queue from 100 to 512
45 (as suggested by Alex Stone)
46 - implemented sine LFO, pulse LFO and saw LFO
47 - experimental support for per voice equalization
48 - added command line option --exec-after-init
49 - Introduced new C++ API method:
50 EngineChannel::InstrumentFileName(int index)
51 allowing to retrieve the whole list of files used for the loaded
52 instrument on an engine channel (a.k.a. part). Some GigaStudio
53 instruments for example are splitted over several files like
54 "Foo.gig", "Foo.gx01", "Foo.gx02", ...
55 - Added new C++ API method Sampler::GetGlobalMaxVoices().
56 - Added new C++ API method Sampler::GetGlobalMaxStreams().
57 - Added new C++ API method Sampler::SetGlobalMaxVoices().
58 - Added new C++ API method Sampler::SetGlobalMaxStreams().
59 - Various "const" and "restrict" optimizations.
61 * Gigasampler format engine:
62 - implemented the "round robin keyboard" dimension
63 - fixed round robin and random dimensions for cases when number of
64 dimension zones is not a power of two
65 - made round robin use a counter for each region instead of each
66 key
68 * SFZ format engine:
69 - Initial implementation (not usable yet)
70 - added support for v2 multiple stage envelope generators
71 - added a fine-tuned v1 envelope generator instead of using the
72 one from the gig engine
73 - fixed memory leak and memory handling errors
74 - added support for trigger=first, trigger=legato and sw_previous
75 - allow non-numerical key values ("C#4" for example)
76 - "key" opcode now sets pitch_keycenter too
77 - fixed error when unloading instrument with same sample used by
78 multiple regions
79 - added some opcode aliases, like loopmode for loop_mode, to be
80 more compatible
81 - added support for trigger=release and rt_decay
82 - added support for off_mode=normal
83 - added support for random, seq_position, seq_length and volume
84 - added v1 LFO opcodes to parser (no support in engine yet)
85 - added support for amp_veltrack and amp_velcurve_N
86 - fine-tuned the default velocity curve
87 - added support for transpose
88 - fixed crash when using sw_down/up
89 - improved logic for sw_lokey/hikey/up/down/last
90 - added more v1 aliases to parser
91 - reduced memory usage for sfz data
92 - RT-safeness: avoid malloc in audio thread
93 - fixed a bug that could cause voice stealing to fail
94 - optimized sample lookup
95 - improved support for exclusive groups (group, off_by and
96 off_mode)
97 - added support for controller triggered regions (on_locc/on_hicc)
98 - added support for loop_mode=one_shot
99 - fixed playback of 16 bit wav files on big endian CPUs
100 - added support for Ogg Vorbis sample files
101 - use loop markers from sample file if loop_start and loop_end are
102 not set in sfz file
103 - implemeted filters. Filter types: lowpass, bandpass, bandreject
104 and highpass. 1, 2, 4 and 6 pole filters. Opcodes: fil_type,
105 cutoff, resonance, fil_veltrack, fil_keytrack, fil_keycenter,
106 cutoff_cc, cutoff_chanaft.
107 - bugfix: zero ampeg_sustain didn't work
108 - added support for velocity effect on amplifier envelope time
109 (ampeg_vel2attack, ampeg_vel2decay, ampeg_vel2sustain and
110 ampeg_vel2release)
111 - added support for EG hold (ampeg_hold)
112 - added support for sample offset (offset)
113 - use the newly introduced signal units model
114 - implemented opcodes lfoN_delay,
115 lfoN_freq, lfoN_pan, lfoN_cutoff, lfoN_resonance
116 - implemented opcodes ampeg_delay, ampeg_vel2delay
117 pitcheg_delay, pitcheg_start, pitcheg_attack, pitcheg_hold,
118 pitcheg_decay, pitcheg_sustain, pitcheg_release, pitcheg_vel2delay,
119 pitcheg_vel2attack, pitcheg_vel2hold, pitcheg_vel2decay,
120 pitcheg_vel2sustain, pitcheg_vel2release, pitcheg_depth
121 - implemented opcodes pitchlfo_delay, pitchlfo_freq,
122 pitchlfo_depth, fillfo_delay, fillfo_freq, fillfo_depth,
123 amplfo_delay, amplfo_freq, amplfo_depth
124 - implemented opcodes fileg_delay, fileg_start, fileg_attack,
125 fileg_hold, fileg_decay, fileg_sustain, fileg_release, fileg_vel2delay,
126 fileg_vel2attack, fileg_vel2hold, fileg_vel2decay, fileg_vel2sustain,
127 fileg_vel2release
128 - implemented opcode lfoN_wave
129 - implemented opcode pitchlfo_depthccN
130 - implemented opcodes lfoN_phase, lfoN_phase_onccX,
131 lfoN_pitch, lfoN_pitch_onccX
132 - implemented opcodes pitchlfo_fade,
133 fillfo_fade, amplfo_fade, lfoN_fade, lfoN_fade_onccX
134 - implemented opcodes fillfo_freqccN,
135 pitchlfo_freqccN, amplfo_freqccN, lfoN_freq_onccX
136 - implemented opcodes ampeg_delayccN, ampeg_startccN,
137 ampeg_attackccN, ampeg_holdccN, ampeg_decayccN, ampeg_sustainccN,
138 ampeg_releaseccN, egN_timeX_onccY, egN_levelX_onccY
139 - lfoN_* and egN_* opcodes defined in group sections
140 are now taken into account
141 - implemented curves
142 - implemented opcodes volume_onccN, volume_curveccN
143 - implemented opcode volume_smoothccN
144 - implemented opcodes fillfo_depthccN, amplfo_depthccN,
145 lfoN_volume, lfoN_volume_onccX, lfoN_volume_smoothccX,
146 lfoN_freq_smoothccX, lfoN_pitch_smoothccX, lfoN_pan_onccX,
147 lfoN_pan_smoothccX, lfoN_cutoff_onccX, lfoN_cutoff_smoothccX,
148 lfoN_resonance_onccX, lfoN_resonance_smoothccX, lfoN_delay_onccX
149 - implemented opcode end
150 - implemented opcodes egN_amplitude_onccX,
151 egN_volume, egN_volume_onccX, egN_cutoff, egN_cutoff_onccX,
152 egN_pitch, egN_pitch_onccX, egN_resonance, egN_resonance_onccX
153 - implemented opcodes xfin_lokey, xfin_hikey,
154 xfout_lokey, xfout_hikey, xf_keycurve, xfin_lovel, xfin_hivel,
155 xfout_lovel, xfout_hivel, xf_velcurve, xfin_loccN, xfin_hiccN,
156 xfout_loccN, xfout_hiccN, xf_cccurve
157 - implemented opcodes pan_onccN, pan_smoothccN,
158 pan_curveccN, egN_pan, egN_pan_curve, egN_pan_onccX,
159 egN_pan_curveccX, pitch_veltrack
160 - when failed to parse a sfz file
161 print the line number on which the error occurs
162 - use common pool of CC objects to minimize RAM usage
163 - implemented opcodes amplfo_delay_onccN,
164 amplfo_fade_onccN, fillfo_delay_onccN, fillfo_fade_onccN,
165 pitchlfo_delay_onccN, pitchlfo_fade_onccN
166 - implemented opcodes fileg_delay_onccN,
167 fileg_start_onccN, fileg_attack_onccN, fileg_hold_onccN,
168 fileg_decay_onccN, fileg_sustain_onccN, fileg_release_onccN,
169 fileg_depth_onccN, pitcheg_delay_onccN, pitcheg_start_onccN,
170 pitcheg_attack_onccN, pitcheg_hold_onccN, pitcheg_decay_onccN,
171 pitcheg_sustain_onccN, pitcheg_release_onccN, pitcheg_depth_onccN
172 - implemented automatic aliasing *ccN <-> *_onccN
173 - *lfo_freqccN wasn't working when the respective *lfo_freq
174 was not set or was set to zero
175 - lfoN_freq_onccX wasn't working when lfoN_freq
176 was not set or was set to zero
177 - implemented opcodes resonance_onccN, resonance_smoothccN,
178 resonance_curveccN, cutoff_smoothccN, cutoff_curveccN
179 - implemented opcodes fillfo_depthchanaft,
180 fillfo_freqchanaft, amplfo_depthchanaft, amplfo_freqchanaft,
181 pitchlfo_depthchanaft, pitchlfo_freqchanaft
182 - implemented opcodes pitch_onccN,
183 pitch_curveccN, pitch_smoothccN, pitch_stepccN
184 - implemented opcodes volume_stepccN, pan_stepccN,
185 cutoff_stepccN, resonance_stepccN, lfoN_freq_stepccX,
186 lfoN_volume_stepccX, lfoN_pitch_stepccX, lfoN_pan_stepccX,
187 lfoN_cutoff_stepccX, lfoN_resonance_stepccX
188 - implemented opcodes eq1_freq, eq2_freq, eq3_freq,
189 eq1_freqccN, eq2_freqccN, eq3_freqccN, eq1_bw, eq2_bw, eq3_bw,
190 eq1_bwccN, eq2_bwccN, eq3_bwccN, eq1_gain, eq2_gain, eq3_gain,
191 eq1_gainccN, eq2_gainccN, eq3_gainccN
192 - implemented opcodes delay, delay_onccN, delay_random,
193 delay_samples, delay_samples_onccN
194 - implemented opcodes egN_eq1gain, egN_eq2gain, egN_eq3gain,
195 egN_eq1gain_onccX, egN_eq2gain_onccX, egN_eq3gain_onccX, egN_eq1freq,
196 egN_eq2freq, egN_eq3freq, egN_eq1freq_onccX, egN_eq2freq_onccX,
197 egN_eq3freq_onccX, egN_eq1bw, egN_eq2bw, egN_eq3bw, egN_eq1bw_onccX,
198 egN_eq2bw_onccX, egN_eq3bw_onccX, lfoN_eq1gain, lfoN_eq2gain,
199 lfoN_eq3gain, lfoN_eq1gain_onccX, lfoN_eq2gain_onccX, lfoN_eq3gain_onccX,
200 lfoN_eq1gain_smoothccX, lfoN_eq2gain_smoothccX, lfoN_eq3gain_smoothccX,
201 lfoN_eq1gain_stepccX, lfoN_eq2gain_stepccX, lfoN_eq3gain_stepccX,
202 lfoN_eq1freq, lfoN_eq2freq, lfoN_eq3freq, lfoN_eq1freq_onccX,
203 lfoN_eq2freq_onccX, lfoN_eq3freq_onccX, lfoN_eq1freq_smoothccX,
204 lfoN_eq2freq_smoothccX, lfoN_eq3freq_smoothccX, lfoN_eq1freq_stepccX,
205 lfoN_eq2freq_stepccX, lfoN_eq3freq_stepccX, lfoN_eq1bw, lfoN_eq2bw,
206 lfoN_eq3bw, lfoN_eq1bw_onccX, lfoN_eq2bw_onccX, lfoN_eq3bw_onccX,
207 lfoN_eq1bw_smoothccX, lfoN_eq2bw_smoothccX, lfoN_eq3bw_smoothccX,
208 lfoN_eq1bw_stepccX, lfoN_eq2bw_stepccX, lfoN_eq3bw_stepccX
209 - implemented opcodes eq1_vel2freq, eq2_vel2freq,
210 eq3_vel2freq, eq1_vel2gain, eq2_vel2gain, eq3_vel2gain
211 - sfz parser: allow double spaces in sample filenames
212 - sfz parser: allow absolute paths for sample filenames
213 - use linear decay and release for filter and pitch EG
214 - bugfix: only the first amp_veltrack definition in a file was
215 used
216 - bugfix: looping was disabled if loop_start was set to 0
217 - allow regions with end=-1 to turn off other regions using the
218 group and off_by opcodes (#168)
219 - made end=0 play the whole sample
220 - fixed support for lochan and hichan opcodes (#155)
221 - fixed crash when using lochan/hichan opcodes (#187)
223 * SoundFont format engine:
224 - Initial implementation (not usable yet)
225 - RT-safeness: avoid malloc in audio thread
226 - fixed a bug that could cause voice stealing to fail
227 - fine-tuned amplitude EG (by switching from gig to sfz EG)
228 - initial implementation of Vibrato LFO and Modulation LFO
229 - initial implementation of cutoff filter
230 - use linear decay and release for filter and pitch EG
232 * Gigasampler format engine:
233 - bugfix: pitch LFO controller "internal+aftertouch" was broken
234 - bugfix: filter keyboard tracking was broken
235 - filter performance fix (an unnecessary copy was made of the
236 filter parameters in each sub fragment)
238 * Host plugins (VST, AU, LV2, DSSI):
239 - AU bugfix: failed to destroy its audio/MIDI devices
240 - Listen to all interfaces on Mac OS X (INADDR_ANY)
241 - VST bugfix: If the host called resume() before and after
242 changing sample rate or block size, the number of channels was
243 incorrectly set to two. This caused silence in Cubase 5.
244 - save engine type (gig, sfz or sf2) in plugin state
245 - VST: when opening Fantasia, look for both 32 and 64 bit Java on
246 64 bit Windows
247 - AU: changed number of output channels from one stereo to 16
248 stereo
249 - VST: made it possible to build the VST plugin for Mac
250 - AU: link AU plugin dynamically if --disable-shared isn't
251 specified
252 - LV2 "state" extension support (patch by David Robillard)
253 - VST bugfix: instrument loading hang and crashed the host when
254 the plugin was loaded a second time (#174)
255 - plugin bugfix: instrument loading hang when the plugin was
256 loaded a second time (this time it's for Linux and Mac, previous
257 similar fix was for Windows)
258 - thread safety fixes for the instrument loading thread
259 - LV2: use the new lv2 package if present
260 - VST: try to open Fantasia automatically on Linux and Mac too (on
261 Linux, the Fantasia jar should be placed in <prefix>/share/java)
262 - VST: fixed crashes on Linux Ardour and EnergyXT
263 - DSSI bugfix: it wasn't possible to change engine type. The MIDI
264 port and audio channel routing for DSSI plugins are now visible.
265 - LV2: use urid and atom extensions instead of deprecated uri-map
266 and event
267 - LV2: lv2 package 1.0 is now required to build the LV2 plugin
268 - LV2: changed number of output channels to 16 stereo, just like
269 the VST and AU plugins
270 - LV2: fixed save/restore of SFZ state (patch by David Robillard)
271 - LV2: made LV2 plugin buildable on Windows and Mac
272 - VST: implemented retrieval and switching of programs using the
273 sampler's internal MIDI instrument mapping system
275 * MIDI driver:
276 - ALSA MIDI driver supports now "NAME" device parameter, for overriding
277 the ALSA sequencer client name
278 - removed limit of maximum amount of MIDI ports per MIDI device, since
279 there is no reason for this limit
280 - MME: fixed memory handling bug found with cppcheck
281 - MME: removed compiler warning
282 - CoreMIDI: implemented driver specific port parameter "CORE_MIDI_BINDINGS",
283 which allows to retrieve the list of CoreMIDI clients / ports and to connect
284 to them a la JACK, via the usual sampler APIs
285 - CoreMIDI: added driver specific port parameter "AUTO_BIND", if enabled
286 the driver will automatically connect to other CoreMIDI clients' ports
287 (e.g. external MIDI devices being attached to the Mac)
289 * audio driver:
290 - ASIO driver fixes for newer gcc versions (fix from PortAudio)
292 * LSCP server:
293 - added support for sending MIDI CC messages via LSCP command
294 "SEND CHANNEL MIDI_DATA CC <sampler-chan> <ctrl> <val>"
295 - added LSCP command "GET AVAILABLE_EFFECTS"
296 - added LSCP command "LIST AVAILABLE_EFFECTS"
297 - added LSCP command "GET EFFECT INFO <effect-index>"
298 - added LSCP command "CREATE EFFECT_INSTANCE <effect-index>"
299 - added LSCP command
300 "CREATE EFFECT_INSTANCE <effect-system> <module> <effect-name>"
301 - added LSCP command "DESTROY EFFECT_INSTANCE <effect-instance>"
302 - added LSCP command "GET EFFECT_INSTANCES"
303 - added LSCP command "LIST EFFECT_INSTANCES"
304 - added LSCP command "GET EFFECT_INSTANCE INFO <effect-instance>"
305 - added LSCP command
306 "GET EFFECT_INSTANCE_INPUT_CONTROL INFO <effect-instance> <input-control>"
308 <effect-instance> <input-control> <value>"
309 - added LSCP command "GET SEND_EFFECT_CHAINS <audio-device>"
310 - added LSCP command "LIST SEND_EFFECT_CHAINS <audio-device>"
311 - added LSCP command "ADD SEND_EFFECT_CHAIN <audio-device>"
312 - added LSCP command
313 "REMOVE SEND_EFFECT_CHAIN <audio-device> <effect-chain>"
314 - added LSCP command
315 "GET SEND_EFFECT_CHAIN INFO <audio-device> <effect-chain>"
316 - added LSCP command "APPEND SEND_EFFECT_CHAIN EFFECT <audio-device>
317 <effect-chain> <effect-instance>"
318 - added LSCP command "INSERT SEND_EFFECT_CHAIN EFFECT <audio-device>
319 <effect-chain> <effect-chain-pos> <effect-instance>"
320 - added LSCP command "REMOVE SEND_EFFECT_CHAIN EFFECT <audio-device>
321 <effect-chain> <chain-pos>"
322 - added LSCP command "SET FX_SEND EFFECT <sampler_channel>
323 <fx_send_id> <effect_chain> <chain_pos>"
324 - added LSCP command "REMOVE FX_SEND EFFECT <sampler_channel> <fx_send_id>"
329 * Bug fixes:
330 - Fixed crash which may occur when MIDI key + transpose is out of range
331 - minor valgrind fixes
332 - fixed crash which occurred when changing an already deployed sampler
333 channel to a different engine type
334 - fixed crash when deleting a sampler channel or changing engine
335 type while an instrument load was in progress
336 - bugfix: playing a note while changing the instrument could cause
337 a crash, or give "Handing back unknown region" error messages
338 - bugfix: calling SET VOICES (which Fantasia does on start and
339 refresh) could cause instruments to be unloaded
340 - Mac OS X: fixed name collision of enum in EffectControl and
341 wrong name of destructor in AudioOutputDeviceCoreAudio.cpp
342 - Mac OS X: fixed hanging threads
343 - Windows: disabled the previous thread fix on non-Mac systems, as
344 it caused hanging threads on Windows
345 - Fixed possible crashes due to corrupted MIDI/audio device list
346 after MIDI/audio device creation failure
347 - When creating MIDI instrument map entries with "PERSISTENT" type, the
348 instruments were uselessly precached with zero samples, however it still
349 took the full preloading time and on 1st program change the respective
350 instrument was completely reloaded again.
351 - fixed handling of rapid bank select and program change messages
352 sent to the same sampler channel (patch from the Open Octave
353 project, slightly adjusted)
354 - fixed crash when trying to create an effect instance with controls
355 which min and/or max values depend on the sample rate
356 - fixed bug #162
357 - bugfix: LADSPA_PATH was not evaluated correctly when containing
358 multiple paths (#165)
359 - thread safety fixes for the instrument loading thread
360 - bugfix: instrument loading crashed for sfz and sf2 in Ardour
361 (#176)
362 - more thread safety fixes for the instrument loading thread
363 - sfz/sf2 engine: fixed crash when using small audio fragment size
364 - Mac OS X: fixed crash when unloading plugin on 10.7 and later
365 - Mac OS X: fixed process hang when unloading 32-bit plugin (bug
366 introduced in previous fix)
367 - fixed crash when a channel received a program change while
368 playing a note in a key group
369 - fixed erroneous error message piping in VoiceBase.h
370 (labelled "Disk stream not available in time")
372 Version 1.0.0 (31 July 2009)
374 * packaging changes:
375 - autoconf bugfix: the PKG_CONFIG variable wasn't initialized properly,
376 causing e.g. the libgig test to fail when
377 "./configure --disable-jack-driver" was used
378 (patch by Alexis Ballier)
379 - fixed compilation with gcc 4.3
380 - fixes for building on OS X (thanks to Ebrahim Mayat for testing)
381 - fixed configure so it detects x86_64 (#107)
382 - fixes for building with newer MinGW versions
383 - fix for building with bison 2.4 (#111)
384 - fixed building with libgig installed in a non-standard directory
385 - minor fix in configure for mmsystem.h detection on MinGW
386 - Windows: look for editor plugins and Fantasia using base
387 directory of liblinuxsampler dll (look in the same directory and one
388 directory above)
389 - configure script fix: removed unconditional use of SSE
390 - fixed building with sqlite installed in a non-standard directory
391 - when cross-compiling, don't try to create instruments.db
392 - fix for new mingw-w64 version, which has usleep
394 * general changes:
395 - bugfix: on some POSIX systems instrument editor plugins refused to
396 load as we used a non-portable Linux specific struct field
397 (fixes bug #70, patch by Ronald Baljeu)
398 - fixed endless loop which occured when loading an instrument editor
399 plugin DLL on Windows systems failed
400 - fixed memory leaks that occurred when liblinuxsampler was unloaded
401 - optimized the SynchronizedConfig class so it doesn't wait
402 unnecessarily long after an update
403 - added support for notifying instrument editors on note-on / note-off
404 events (e.g. to highlight the pressed keys on the virtual keyboard
405 of gigedit)
406 - added support for triggering notes by instrument editors (see above)
407 - be verbose on DLL load errors (on Linux)
408 - fixed di-harmonic triangle LFO implementation (this LFO implementation
409 is only used by very few systems, most pick the int math
410 implementation, which is usually faster)
411 - fixes for audio drivers with varying buffer sizes
412 - experimental support for running LinuxSampler as a DSSI, LV2 and
413 VST plugin
414 - notification events for stream/voice count statistics are now sent
415 only when there are actual changes
416 - added memory ordering constraints to improve stability on
417 multi-core and multi-cpu systems
418 - maximum voices and disk streams can now be altered at runtime
419 - fixed CPU feature detection on x86_64 (maybe fixes #108)
420 - automatic stacktrace mechanism is now turned off by default and can
421 be switched on by command line option "--stacktrace" (the automatic
422 stacktrace mechanism seems to be broken on most systems at the moment)
423 - C++ API method InstrumentManager::LaunchInstrumentEditor() now returns
424 a pointer to the launched InstrumentEditor object
425 - added optional 3rd party user data parameter for following
426 liblinuxsampler C++ API methods: InstrumentEditor::Main(),
427 InstrumentEditor::Launch(),
428 InstrumentManager::LaunchInstrumentEditor()
429 - theoretical fix: made SynchronizedConfig follow C++0x memory
430 model more strictly
431 - fixes for using large audio device buffers
432 - Windows: add the installation directory to the DLL search path
433 when loading an editor plugin (solves problems with VST and
434 gigedit on systems with other GTK versions installed)
436 * audio driver:
437 - removed the nonsense audio channel constraint (which was hard coded to
438 max. 100 audio channels) for most audio drivers
439 - JACK audio driver did not offer a device parameter "SAMPLERATE" as
440 opposed to the LSCP specs
441 - bugfix: the SAMPLERATE parameter of some drivers (e.g. JACK)
442 reflected the wrong value
443 - fixed a memory management bug in ASIO driver
444 - Makefile fix: JACK_CFLAGS wasn't used
445 - JACK: use jack_client_open instead of the deprecated
446 jack_client_new
447 - added (experimental) CoreAudio driver
448 - applied old fixes to the ASIO driver that were included in the
449 previous binary release but accidentally never committed to CVS
450 (fixes #117)
451 - fixes for ASIO on mingw-w64 (iasio wrapper is not needed on
452 win64)
453 - VST: added support for sample rate and buffer size changes
454 - VST: close editor (Fantasia) when the VST is removed
455 - VST: avoid opening Fantasia more than once for each VST instance
456 - VST: export main function as "main" on Linux too (fix for
457 energyXT)
458 - VST: changed number of output channels from one stereo to 16
459 stereo
460 - added channel routing, fxsends and midi maps to the settings
461 stored in the plugin state
462 - performance optimization of AudioChannel::MixTo() and
463 AudioChannel::CopyTo() methods using GCC vector exensions
464 (if available)
465 - ASIO fixes: avoid initializing the device twice, avoid throwing
466 exception when getting parameters from a disconnected device
468 * MIDI driver:
469 - added JACK MIDI driver
470 - dispatch bank select as ordinary CC as well, the user might seriously
471 want to (mis)use it for some purpose ("fixed" in all current MIDI
472 input drivers: ALSA, CoreMIDI, JACK, MidiShare, MME)
473 - bugfix: pitch bend wasn't working with JackMidi, VST, LV2, MME,
474 CoreMidi or AU
475 - fixed mingw-w64 compilation error in MME driver
476 - made program change handling in MIDI thread real-time safe by
477 moving the logic to a non-RT thread
478 - fixed minor memory leak in ALSA driver
480 * instruments database:
481 - avoid time consuming samples scanning when adding instruments
482 to the instruments database
483 - added support for handling lost files in the instruments database
484 - Implemented option for adding instruments in separate directories
485 in the instruments database
486 (patch by Chris Cherrett & Andrew Williams, a bit adjusted)
487 - work-around for missing fnmatch function on Windows to make
488 instrument database compilable
489 - added instrument database support on Windows
490 works with both standalone mode and VST plugin.
491 the instruments DB file is located in
492 %USERPROFILE%\.linuxsampler\instruments.db which allows different
493 databases for each windows user
494 if no DB is present it automatically creates the .linuxsampler subdir
495 and then creates an empty DB
496 - fixed recursive import, which was broken on Windows
498 * Gigasampler format engine:
499 - fixed a memory leak that could happen when a channel was deleted
500 while notes were playing
501 - made it possible to load an instrument even if the audio thread
502 isn't running
503 - added partial support for the "Controller Triggered" MIDI rule,
504 enough for piano gigs with pedal noise samples
505 - minor fix: only mark FX sends as being modified if really the
506 respective FX send MIDI controller was used
507 - added support for GM global device volume SysEx message
508 (can be selected at compile time to either apply globally to the
509 whole sampler [default] or only to the sampler channels that are
510 connected to the respective MIDI input port the SysEx message arrived
511 on)
512 - bugfix: notes triggered at position 0 in the audio buffer were
513 sometimes wrongly killed in the same buffer, causing no sound to
514 be played
515 - added support for chromatic / drumkit mode Roland GS Sysex message
516 which is usally used to switch a part between chromatic sounds and
517 drumkit sounds (as opposed to their common default setting of MIDI
518 part 10 being a drumkit part and all other ones chromatic parts), in
519 LS however one can switch between the first 16 MIDI instrument maps
520 defined for the sampler
521 - made it possible to create multiple sample channels even if the
522 audio thread isn't running
523 - fixed crash when removing channel with active voices (#116)
524 - bugfix: on sample reference changes (instrument editor), only
525 un-cache the respective sample if it's really not used by any
526 sampler engine anymore
527 - re-cache samples in case they were changed by an instrument editor,
528 e.g. when a sample was added while playing (#82)
529 - fixed hanging notes which occured when note-off event had the exact
530 same time stamp as the voice's note-on event and occured both in the
531 same audio fragment cycle (fixes bug #112)
532 - added support for the "fine tune" and "pitch bend range"
533 instrument-level gig parameters
534 - fixed minor artifacts in pitch bend handling
535 - added support for GS Reset SysEx message
536 - allow gig files to use unlimited downward pitch shifting
537 - added a limit check for upward pitch shifting
538 - bugfix: sometimes, when playing a note twice fast, the second
539 note was silent
540 - fixed crash happening when a pitch bend event arrived at the
541 same time a new instrument was loading
543 * LSCP server:
544 - added new LSCP event "CHANNEL_MIDI" which can be used by frontends to
545 react on MIDI data arriving on certain sampler channels (so far only
546 Note-On and Note-Off events are sent via this LSCP event)
547 - added new LSCP event "DEVICE_MIDI" which can be used by frontends to
548 react on MIDI data arriving on certain MIDI input devices (so far only
549 Note-On and Note-Off events are sent via this LSCP event)
550 - added new LSCP commands: FIND LOST DB_INSTRUMENT_FILES and
552 - added new LSCP command: SEND CHANNEL MIDI_DATA which can be used by
553 frontends to send MIDI messages to specific sampler channel
554 - added two additional fields to GET FILE INSTRUMENT INFO command -
556 - bugfix: the bank number provided by MIDI_INSTRUMENT_INFO notifications
557 was incorrect
558 - Optimized the retrieval of the MIDI instrument mappings
559 - added new LSCP command "SET VOICES" to globally alter the maximum
560 amount of voices, added new LSCP event "GLOBAL_INFO:VOICES" which will
561 be triggered respectively
562 - added new LSCP command "SET STREAMS" to globally alter the maximum
563 amount of disk streams, added new LSCP event "GLOBAL_INFO:STREAMS"
564 which will be triggered respectively
565 - bugfix: retry if "select" returns EINTR (this fixes a crash when
566 a gigedit file dialog is opened)
567 - close all connections when LSCPServer is deleted
568 - hard close of all sockets on exit
569 - bugfix: SET CHANNEL MIDI_INPUT_TYPE didn't work with the MME
570 driver
572 * Bug fixes:
573 - fixed a crash which occurs when removing a sampler channel waiting
574 to start instrument loading after another channel
575 - fixed a crash which occurs when removing a sampler channel with
576 instrument loading in progress (bug #113)
577 - fixed termination caused by uncaught exception when adding MIDI
578 instrument with PERSISTENT load mode
579 - fixed possible iterator invalidations when resetting the sampler
580 - fixed memory leaks when issuing the following LSCP commands:
585 - fixed possible compilation error when sqlite is not present
586 - fixed orphaned pointers when setting maximum voices limit (bug #118)
587 - fixed crash when changing the audio output device of a sampler
588 channel with loaded instrument and start playing notes
589 - fixed endless loop in Engine::SuspendAll() (bug #120)
590 - fixed a low-level atomic load function that was broken on 64-bit
591 PowerPC, which probably could cause crashes on that platform
592 - fixed a memory management error which could cause a crash when a
593 plugin was unloaded
594 - bugfix: two private structs had the same name, which could cause
595 problems if the linker chose the wrong constructor
596 - fixed low-level ConditionServer usage bug that caused lockups on
597 Windows
600 Version 0.5.1 (6 December 2007)
602 * packaging changes:
603 - added autoconf checks for pthread library
604 - added autoconf check for pthread bug found on certain NPTL-enabled
605 glibc versions (see Gentoo bug report #194076)
606 - added autoconf checks for MS Windows
608 * general changes:
609 - bugfix: the thread used by an editor plugin didn't die when the
610 editor closed
611 - bugfix: Ringbuffer.h: fill_write_space_with_null() did not zero
612 out all the space. operator--() did not apply size_mask after
613 decrementing the read_ptr. DEFAULT_WRAP_ELEMENTS set to 0 in
614 order to avoid problems with the _NonVolatileReader functions.
615 - bugfix: Stream.h: added a missing divide by BytesPerSample in
616 GetWriteSpace(). Since this function is currently only used in
617 the stream's qsort() compare function, it didn't trigger any
618 bugs.
619 - Resampler.h, Synthesizer.h: cubic interpolation now works in
620 24bit mode too. Faster method to read 24bit words on little
621 endian machines (x86): replaced 3 byte reads + shifts with a 1
622 unaligned 32bit read and shift
623 - experimental support for MS Windows (MIDI input via MME, AUDIO
624 output via ASIO)
625 - made handling of SIGINT signal (Ctrl-C) a bit more robust
626 - support for monitoring the total number of active disk streams
627 (new LSCP commands: GET TOTAL_STREAM_COUNT,
630 * AUDIO driver:
631 - added Windows ASIO low latency audio driver
633 * MIDI driver:
634 - added MME Windows MIDI driver
636 * LSCP server:
637 - added support for Windows style path / filenames, however with
638 forward slash path separators instead of backslash
639 (i.e. "C:/foo/bar.gig")
640 - allow naughty liblscp to send non-string device parameters within
641 apostrophes as well
642 - added new LSCP commands: "GET FILE INSTRUMENTS <file>",
643 "LIST FILE INSTRUMENTS <file>" and
644 "GET FILE INSTRUMENT INFO <file> <index>" for retrieving informations
645 about an arbitrary instrument file on the system where the sampler is
646 running on
648 Version 0.5.0 (15 October 2007)
650 * packaging changes:
651 - config.h is not going to be installed along with liblinuxsampler's
652 API header files anymore
653 - only the API relevant header (and source) files will be exported to
654 the Doxygen API documentation (explicitly listed in Doxyfile.in)
655 - added completely new XCode project files for Mac OSX which is now
656 capable to execute our autoconf environment, thus no need anymore
657 to maintain the compile time configuration file (osx/version.h) for
658 OSX manually (patch by Toshi Nagata)
659 - fixed buggy boolean --enable-foo / --disable-foo configure script
660 parameters
661 - global.h now only covers global definitions that are needed for the
662 C++ API header files, all implementation internal global definitions
663 are now in global_private.h
664 - atomic.h is not exposed to the C++ API anymore
665 - no need to include config.h anymore for using LS's API header files
666 - fixed warnings in API doc generation
668 * general changes:
669 - replaced the old, confusing MIDI program change mechanism by a
670 flexible MIDI instrument mapper which allows to map arbitrary
671 (MIDI bank, MIDI program) pairs with arbitrary
672 (engine type, instrument file, file index) triplets which will be
673 loaded on the respective channel when such MIDI program change
674 messages arrive, beside that, each entry allows to define a life-time
675 strategy for the instrument, so the user can define whether the
676 instrument should i.e. be permanently loaded (to be able to switch
677 quickly among sounds) or i.e. loaded just on demand when the
678 respective program change arrives (to safe RAM space), as well as a
679 global volume factor for each entry, so the user can adjust the volume
680 dynamic of his mapped instrument collection without having to modify
681 the instrument files, also one can manage arbitrary amount of such
682 MIDI instrument maps and assign each sampler channel individually
683 a certain map, so that i.e. separation of normal instruments and
684 drumkits is possible
685 - new notification events for tracking changes to audio/MIDI devices,
686 MIDI instrument maps, MIDI instruments, FX sends, global volume.
687 - sampler was limited to load max. 200 instruments in the background
688 due to a constant size RingBuffer FIFO which is now replaced by a
689 dynamic (unlimited) size std::list FIFO
690 - added FX sends, these allow to route audio signals to arbitrary audio
691 output channels for being processed by external effect processors
692 (i.e. jack-rack), the send levels are controllable via arbitrary MIDI
693 controllers
694 - global (sampler wide) volume can now be controlled at runtime
695 - Implemented new, improved notification system
696 - fixed compilation errors regarding OSX
697 (patch by Toshi Nagata)
698 - implemented instruments database
699 - added support for escape sequences to the instruments database
700 - added highly experimental support for on-the-fly instrument editing
701 within the sampler's process (by using instrument editor plugins),
702 you'll notice the new "Registered instrument editors:" message on
703 startup, added a new LSCP command:
704 "EDIT CHANNEL INSTRUMENT <sampler-channel>"
705 to spawn a matching instrument editor for the instrument on the
706 given sampler channel, the plugin path can be overridden at compile
707 time with ./configure --enable-plugin-dir=/some/dir
708 - added experimental code for synchronizing instrument editors hosted
709 in the sampler's process to safely edit instruments while playing
710 without a crash (hopefully) by either suspending single regions
711 wherever possible or if unavoidable whole engine(s)
712 - fixed several issues in fundamental "Thread" class: set scheduling
713 policy and priority on thread level, set a minimum stack size for
714 thread (TODO: a reasonable value yet to be tested), bugfix: non-RT
715 threads simply inherited properties of starting thread instead of
716 setting their own policy and priority
717 - minor fix in our automatic stack trace mechanism on crashes, the
718 main process did not wait for the stack trace process to finish
719 its output
720 - fixed some minor memory leaks
721 - reenabled assembly features support, at the moment only for
722 enabling a fast denormal FPU mode (x86 platforms supporting SSE2)
723 - minor assembly fix in x86 features detection (don't use the PIC
724 register, to avoid relocations in the text segment at runtime)
725 - POSIX callback functions of Thread.h are hidden
727 * MIDI driver:
728 - dispatch bank select (MSB and LSB) messages
730 * audio driver:
731 - the ALSA audio output driver parameters now reflect the correct
732 parameter value ranges for the respective selected sound card
733 (patch by Till Wimmer, a bit fixed and extended)
735 * Gigasampler format engine:
736 - if a filter is used and EG2 finishes before EG1, let the voice
737 die when EG2 has finished (this fixes a problem with clicks and
738 voice starvation for some gigs)
739 - playback is no longer disabled during instrument loading
740 - all notes playing on a channel that changes its instrument keep
741 playing with the old instrument until they get a note off
742 command
743 - EG fix: a release value of zero could cause noises or crash
744 - handle MIDI coarse tuning messages (MIDI RPN #0 MSB #2 LSB)
745 - EG fine tuning: when attack is zero the EG starts at a level
746 above max sustain level, which means that there is a decay phase
747 even if sustain is 100%
748 - more EG fixes: the level could sometimes go below zero and cause
749 noises or crashes
750 - minor fix of EGDecay (patch by Toshi Nagata)
751 - fixed compiler error when --enable-override-filter-type was
752 supplied to the configure script (fixes #46)
753 - disk thread: queue sizes are now proportional to CONFIG_MAX_STREAMS
754 instead of a fix value
755 - behavior fix: on MIDI CC# 65 (portamento on / off), 126 (mono mode),
756 127 (solo mode) only kill voices if the respective mode really
757 changed
759 * LSCP server:
760 - fixed compile time error for old Bison versions
761 (i.e. v1.28 found on MacOS 10.4, patch by Toshi Nagata)
762 - parser now supports extended ASCII character set
763 (up to ASCII code 255, i.e. includes now umlauts and accents)
764 - filename arguments in LSCP commands now allow to use escape
765 sequences, that is directly literal as one of: \', \", \\, \n, \r,
766 \f, \t, \v, or as octal ASCII code value like \132, or as hex ASCII
767 code value like \xf2) (fixes bug #24)
768 - the following LSCP commands now also support escape sequences for at
769 least one of their text-based parameters (i.e. name, description):
776 - returns verbose syntax errors (line and column where syntax error
777 occured, the unexpected character and the actually expected, possible
778 character(s), the latter only if less than 5 possibilities)
779 - made sure that LSCP syntax is not affected by gigedit locale
780 settings
781 - bugfix regarding strings parameter lists: all comma separated lists
782 of strings were treated as being one string containing commas
783 (fixes #57)
785 * Bug fixes:
786 - fixed crash occurring on certain LSCP scripts (Bug 39)
787 - another thread safety fix for lscp "load engine" and "set
788 channel audio output device"
789 - fixed a crash which occurs when reassigning the same engine
790 on a sampler channel with connected MIDI device
791 - fixed a crash which occurs when changing the number of ports of a MIDI
792 device connected to a sampler channel to number less then or equal
793 to the index of the port to which the sampler channel is connected.
794 - The previous bindings were not been disconnected when altering
795 the ALSA_SEQ_BINDINGS parameter. Introduced a NONE keyword for
796 unsubscribing from all bindings (e.g. ALSA_SEQ_BINDINGS=NONE).
797 - The active stream/voice count statistic was incorrect.
798 - notification events were not been sent for some sampler
799 channel changes
800 - added default min and max values to restrict the number of allowed
801 audio output channels and MIDI input ports
802 - the connection to the PCM interface is now closed when destroying
803 an audio output device
804 - files with slash in their path or filename could not be loaded
806 * test cases:
807 - updated and fixed (haven't been touched in a while)
809 Version 0.4.0 (24 November 2006)
811 * packaging changes:
812 - changed deprecated copyright attribute to license; added ldconfig
813 to post-(un)install steps; added devel package for liblinuxsampler;
814 to linuxsampler.spec (RPM)
815 - install necessary development header files for allowing 3rd party
816 applications to link against liblinuxsampler
817 - liblinuxsampler's API documentation can be generated with 'make docs'
818 (Doxygen required)
819 - added benchmark to automatically detect the best triangle LFO
820 implementation (currently either an integer math solution or a
821 di-harmonic approximation), automatic detection can be overriden
822 with --enable-signed-triang-algo=x and --enable-unsigned-triang-algo=x
823 configure script argument though (mandatory for cross-compilation)
824 - do not automatically pick optimized gcc flags if the user already
825 provided some on his own (as CXXFLAGS)
826 - added compile time option to disable processing of All-Notes-Off MIDI
827 messages
828 - added compile time options to allow disabling the various audio and
829 MIDI drivers
830 - fixed automatic GCC CPU switch detection on PPC
831 (patch by Ebrahim Mayat)
833 * Gigasampler format engine:
834 - extensive synthesis optimization
835 (reimplementation of EGs and LFO(s), removed synthesis parameter
836 prerendering and the synthesis parameter matrix in general, splitting
837 each audio fragment into subfragments now where each subfragment uses
838 constant pitch and filter coefficients. The volume coefficient is
839 linearly interpolated inside a subfragment, unless
840 --disable-interpolate-volume is set.)
841 - fine tuning of the EG modulation parameters
842 - improved filter cutoff calculation by adding support for the
843 following gig parameters: Cutoff freq (used when no cutoff
844 controller is defined), Control invert, Minimum cutoff, Velocity
845 curve and Velocity range. The keyboard tracking now scales
846 cutoff frequency instead of resonance.
847 - added support for gig parameter Resonance.
848 - fixed bug in sysex handling (patch by Juan Linietsky)
849 - added global volume attenuation of -9 dB (0.35f) to prevent clipping
850 which can be overridden with --enable-global-attenuation
851 - EG fixes: made the length of "attack hold" stage more
852 accurate. Release stage can now start before attack stage
853 ends. Cancel release didn't work when sustain was zero. Attack
854 time now has a minimal value to prevent clicks.
855 - fixed pitch changes larger than one octave
856 - fixed EG3 (pitch envelope) synthesis which was neutral all the time
857 - implemented portamento mode and solo mode (a.k.a 'mono mode'):
858 all modes can be altered via standard GM messages, that is CC5 for
859 altering portamento time, CC65 for enabling / disabling portamento
860 mode, CC126 for enabling solo mode and CC127 for disabling solo mode
861 - fine tuning of the curves for volume (CC7), pan (CC10 and gig
862 parameter) and crossfade
863 - added support for the "attenuation controller threshold" gig
864 parameter
865 - added smoothing of volume changes caused by control change
866 messages
867 - sample loop parameters are now taken from the DimensionRegion
868 instead of the wave chunk
869 - fixed keyswitching for v3 gigs with a number of keyswitch splits
870 not equal to a power of two
871 - reimplementation of the filter algorithm. The new filter is more
872 accurate and supports all gig filter types, including bandreject
873 and lowpass turbo.
874 - real support for 24 bit samples - samples are not truncated to
875 16 bits anymore
876 - support for aftertouch (channel pressure, not polyphonic
877 aftertouch)
878 - LFO1 behaviour fixed (dampening from max volume instead of
879 amplifying from 0)
881 * LSCP server:
882 - fixed application exit on broken pipe error (fixes bug #20)
883 - fixed the notification messages delay due to lack of
884 network activity (fixes bug #26)
885 - fixed parser bug which occured on space(s) within device parameters
888 * audio driver:
889 - added aRts audio output driver (by no means RT safe)
891 * MIDI driver:
892 - fixed legacy sysex code which caused dispatching of MIDI SysEx
893 messages several times instead of once
894 - API extension for MIDI drivers which already supply exact time stamps
895 for events (i.e. for offline rendering based MIDI drivers)
896 - added checks for bad MIDI data
898 * linuxsampler application:
899 - show available sampler engine types on startup
901 * general changes:
902 - support for muting sampler channels and solo mode of the same, two new
903 LSCP commands ("SET CHANNEL MUTE" and "SET CHANNEL SOLO") and two new
904 fields ("MUTE" and "SOLO") for command "GET CHANNEL INFO" were
905 introduced for this, the behavior is the same like on a mixer console
906 (patch by Grigor Iliev, a bit adjusted). Also added configure option
907 --enable-process-muted-channels which can be used to enable the
908 processing of muted channels.
909 - support for sostenuto pedal
910 - support for monitoring the total number of active voices
913 - fixed some memory management errors
914 - fixed some concurrency problems that could lead to crashes when
915 LSCP commands were executed
916 - fixed crash when instrument loading failed previously on the same
917 sampler channel (fixes bug #36)
919 Version 0.3.3 (15 July 2005)
921 * packaging changes:
922 - fixed compilation with gcc 4.0
924 * Gigasampler format engine:
925 - LFO preprocessing was performed even though the respective LFO was
926 disabled by the instrument patch (minor efficiency fix)
927 - if period time of chosen audio device is too small (< MIN_RELEASE_TIME)
928 for volume ramp downs in the same fragment (needed for current voice
929 stealing implementation) then simply inform the user with a warning
930 about possible click sounds and reduce the volume ramp down
931 appropriately instead of cancelling the audio device connection
933 Version 0.3.2 (24 June 2005)
935 * packaging changes:
936 - updated autotools build files to compile on OS X
937 (conditional compilation of CoreMIDI and MidiShare drivers)
938 - hand-crafted assembly optimization code can be disabled with
939 './configure --disable-asm' (definitely not recommended)
940 - fixed 'make dist' rule to include all necessary files
941 - require automake (>= 1.5) for 'make -f Makefile.cvs'
942 (due to 'dist-bzip2' automake option)
944 * Gigasampler format engine:
945 - support for the gig parameters for "release velocity response" curves
946 - fine tuning of the EGADSR envelope
947 - volume of release triggered samples now depends on note-on velocity,
948 note length and gig parameter "release trigger decay" instead of
949 note-off velocity.
950 - revised voice stealing
951 (fixes crash and endless loop caused by voice stealing)
952 - don't reset scale tuning on instrument or audio output device change
953 - handle key group conflicts right at the beginning of each fragment
954 instead of when the respective voice is actually launched
955 (fixes undefined behavior if stolen voices belonged to a key group -
956 this case was followed by a "killed voice survived" error message)
957 - fixed minor issue with null/silence samples
958 (those stole voices even though they don't need a voice at all which
959 resulted in "voice stealing didn't work out" messages)
960 - don't reset volume, pan, pitch and MIDI controller values on
961 instrument or audio output device change
963 * LSCP server:
964 - fixed some crashes (patch by Grigor Iliev, fixes #19)
965 - fixed LSCP event "CHANNEL_INFO" notification
966 (e.g. did not notify on volume changes or MIDI program change events)
968 * linuxsampler application:
969 - added command line parameters --lscp-addr and --lscp-port to override
970 default IP address and port of LSCP server
972 Version 0.3.1 (24 May 2005)
974 * initial release

  ViewVC Help
Powered by ViewVC