/[svn]/linuxsampler/trunk/ChangeLog
ViewVC logotype

Contents of /linuxsampler/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 2974 - (show annotations) (download)
Fri Jul 22 15:51:40 2016 UTC (7 years, 8 months ago) by schoenebeck
File size: 68094 byte(s)
* ScriptVM: Implemented automatic suspension of RT safety
  threatening scripts.
* Bumped version (2.0.0.svn25).

1 Version SVN trunk (?)
2
3 * SFZ format engine:
4 - added support for <global>, <master> and #define (patch by Alby M)
5
6 * general changes:
7 - fixed printf type errors (mostly in debug messages)
8 - use unique_ptr instead of auto_ptr when building with C++11
9 - Added RTAVLTree class which is a real-time safe ordered multi-map, thus
10 allowing to sort data efficiently in real-time safe manner.
11 - RTList class: added methods for moving/inserting elements to arbitrary
12 position within a list.
13 - RTList class: added method fromPtr() for creating an Iterator object from
14 a raw element pointer.
15 - Attempt to partly fix resetting engine channels vs. resetting engine, an
16 overall cleanup of the Reset*(), ConnectAudioDevice(),
17 DisconnectAudioDevice() API methods would still be desirable though,
18 because the current situation is still inconsistent and error prone.
19 - Active voices are now internally grouped to "Note" objects, instead of
20 being directly assigned to a keyboard key. This allows more fine graded
21 processing of voices, which is i.e. required for certain instrument
22 script features.
23 - Fix: Release trigger voices were not spawned on sustain pedal up
24 (CC #64) events.
25 - Fix: Release trigger voices can now also distinguish correctly between
26 note-off and sustain pedal up events.
27 - All engines: Increased ramp speed of volume smoother and pan smoother
28 (while slow rate of crossfade smoother is preserved) to allow quick
29 volume and pan changes by instrument scripts for instance.
30
31 * packaging changes:
32 - removed unnecessary dependency to libuuid
33 (originated by libgig's usage of it)
34 - Automake: set environment variable GCC_COLORS=auto to allow GCC to
35 auto detect whether it (sh/c)ould output its messages in color.
36
37 * Real-time instrument scripts:
38 - Implemented scheduler for delayed MIDI events and for suspended scripts.
39 - Built-in script function "wait()": implemented support for function's
40 "duration-us" argument, thus scripts using this function are now
41 correctly resumed after the requested amount of microseconds.
42 - Built-in script function "play_note()": implemented support for
43 function's "duration-us" argument, thus notes triggered with this
44 argument are now correctly released after the requested amount of
45 microseconds.
46 - Fix: script events were not cleared when engine channel was reset,
47 potentially causing undefined behavior.
48 - Fixed crash which happened when trying to reference an undeclared
49 variable.
50 - Built-in script function "play_note()": Added support for passing
51 special value -1 for "duration-us" argument, which will cause the
52 triggered note to be released once the original note was released.
53 - Instrument script classes now exported with the liblinuxsampler C++ API.
54 - Added new API method ScriptVM::syntaxHighlighting() which provides
55 a convenient syntax highlighting backend for external instrument
56 script editor applications.
57 - Added new C++ API class "ScriptVMFactory".
58 - Extended parser issues to provide not only first line and first
59 column, but also last line and last column of issue (thus marking
60 the precise span of the issue within the source code).
61 - Reload script automatically after being modified by an instrument
62 editor.
63 - NKSP language grammar correction: allow empty event handler bodies
64 like "on note end on".
65 - Implemented built-in script function "change_vol()".
66 - Implemented built-in script function "change_tune()".
67 - Implemented built-in script function "change_pan()".
68 - Implemented built-in script function "change_cutoff()".
69 - Implemented built-in script function "change_reso()".
70 - Implemented built-in script function "event_status()".
71 - Added built-in script constants "$EVENT_STATUS_INACTIVE" and
72 "$EVENT_STATUS_NOTE_QUEUE" both for being used as flags for
73 "event_status()" function.
74 - NKSP language: Added support for bitwise operators ".or.", ".and."
75 and ".not.".
76 - NKSP language scanner: Fixed IDs matching to require at least one
77 character (i.e. when matching function names or variable names).
78 - NKSP language scanner: disabled unusued rules.
79 - Fixed behavior of play_note() and note_off() functions which must
80 be distinguished engine internally from "real" MIDI note on/off
81 events in order to avoid misbehaviors like hanging notes.
82 - Implemented built-in script variable "$KSP_TIMER".
83 - Implemented built-in script variable "$NKSP_REAL_TIMER".
84 - Implemented built-in script variable "$NKSP_PERF_TIMER".
85 - Implemented built-in script variable "$ENGINE_UPTIME".
86 - Implemented built-in script function "inc()".
87 - Implemented built-in script function "dec()".
88 - NKSP language fix: division expressions were evaluated too often.
89 - NKSP language fix: string concatenation operator was right
90 associative instead of left (to right).
91 - NKSP language correction: allow empty statements in entire language
92 (i.e. "if end if").
93 - Implemented built-in script function "stop_wait()".
94 - Implemented built-in script variable "$NI_CALLBACK_ID".
95 - Implemented built-in script variable "$NI_CALLBACK_TYPE".
96 - Implemented built-in script variable "$NKSP_IGNORE_WAIT".
97 - Added support for read-only built-in variables (respectively
98 handled by the script parser).
99 - Added built-in script constant "$NI_CB_TYPE_INIT".
100 - Added built-in script constant "$NI_CB_TYPE_NOTE".
101 - Added built-in script constant "$NI_CB_TYPE_RELEASE".
102 - Added built-in script constant "$NI_CB_TYPE_CONTROLLER".
103 - NKSP Language: Added support for user defined script functions.
104 - Implemented built-in script function "change_attack()".
105 - Implemented built-in script function "change_decay()".
106 - Implemented built-in script function "change_release()".
107 - Fixed all change_*() built-in script functions to apply their
108 synthesis parameter changes immediately in case the respective note
109 was triggered at the same time, instead of scheduling the parameter
110 change, especially because it would cause some parameter types's
111 changes either to be ramped (i.e. change_vol()) or other types even
112 to have not effect at all (i.e. change_attack()).
113 - Implemented built-in script function "sh_left()".
114 - Implemented built-in script function "sh_right()".
115 - Implemented built-in script function "min()".
116 - Implemented built-in script function "max()".
117 - NKSP Fix: "init" event handler was not always executed when a script
118 was loaded.
119 - NKSP built-in wait() script function: abort script execution if
120 a negative or zero wait time was passed as argument (since this is
121 a common indication of a bug either of the script or even of the
122 engine, which could lead to RT instability or even worse).
123 - ScriptVM: Implemented automatic suspension of RT safety
124 threatening scripts.
125
126 Version 2.0.0 (15 July 2015)
127
128 * packaging changes:
129 - fixed building with newer MinGW-w64
130 - Mac OS X: support the new dir for Core Audio SDK
131 - Mac OS X: fixed building outside source directory
132 - made sure all source files for hostplugins are included when
133 doing "make dist"
134 - removed empty directories left from the cvs to svn migration
135 - fixed compilation with gcc 4.6.1
136 - another "make dist" fix, for the LV2 plugin
137 - made --enable-pthread-testcancel default on Mac OS X
138 - minor Makefile fix for building lscpparser when build and source
139 directories are separate
140 - Mac OS X: made it possible to specify plugin installation dir to
141 configure
142 - Mac OS X: Makefile fix for the install-strip target
143 - fixed compilation with gcc 4.7
144 - fixed configure script error with old autoconf versions
145 - lsatomic.h: use gcc provided atomic functions if building with
146 gcc 4.7 and C++11
147 - modernized configure script
148 - fixed linkage error when building with
149 LDFLAGS="-Wl,--no-undefined" (#190)
150 - fixed compilation with Clang 3.2
151 - removed usage of deprecated Automake variable INCLUDES
152 - fixed building with C++11
153 - build fix: ChangeFlagRelaxed.h was missing in makefile
154 - build fix: libsndfile compiler flags were missing in some
155 makefiles
156 - fix for building with bison 3.0 (#202)
157 - Mac OS X: added temporary hack allowing to spawn gigedit as callback
158 on the process's main thread
159 - fixed build error on newer MinGW
160 - support building with older jack versions
161 - support building with spaces in vst sdk path
162 - enabled automake 'subdir-objects' option and moved external
163 source references (vst, au, asio) from makefiles to cpp files,
164 in order to get rid of warnings from automake 1.14
165
166 * general changes:
167 - Refactoring: moved the independent code from
168 the Gigasampler format engine to base classes
169 - command line option '--profile' is currently disabled, since the
170 respective profiling code is currently broken
171 - Introduced support for optional environment variable
172 "LINUXSAMPLER_PLUGIN_DIR" which allows to override the directory
173 where the sampler shall look for instrument editor plugins
174 (patch by Luis Garrido, slightly modified).
175 - implemented Roland GS NRPN 1ArrH which allows to set volume per note
176 - implemented Roland GS NRPN 1CrrH which allows to set pan per note
177 - implemented Roland GS NRPN 1DrrH which allows to set reverb send per
178 note (in this implementation of the sampler its simply hard routed to
179 the 1st effect send of the sampler channel, no matter what the actual
180 effect type is)
181 - implemented Roland GS NRPN 1ErrH which allows to set chorus send per
182 note (in this implementation of the sampler its simply hard routed to
183 the 2nd effect send of the sampler channel, no matter what the actual
184 effect type is)
185 - implemented support for internal LADSPA effects
186 - raised limit of program change queue from 100 to 512
187 (as suggested by Alex Stone)
188 - implemented sine LFO, pulse LFO and saw LFO
189 - experimental support for per voice equalization
190 - added command line option --exec-after-init
191 - Introduced new C++ API method:
192 EngineChannel::InstrumentFileName(int index)
193 allowing to retrieve the whole list of files used for the loaded
194 instrument on an engine channel (a.k.a. part). Some GigaStudio
195 instruments for example are splitted over several files like
196 "Foo.gig", "Foo.gx01", "Foo.gx02", ...
197 - Added new C++ API method Sampler::GetGlobalMaxVoices().
198 - Added new C++ API method Sampler::GetGlobalMaxStreams().
199 - Added new C++ API method Sampler::SetGlobalMaxVoices().
200 - Added new C++ API method Sampler::SetGlobalMaxStreams().
201 - Various "const" and "restrict" optimizations.
202 - all engines: add pan CC value to instrument pan parameter before
203 applying panning, instead of using two separate pan functions in
204 series (#182)
205 - added a lock guard class for exception safe mutex handling and
206 used it everywhere appropriate
207 - Immediately apply scale tuning changes to active voices.
208 - Exposed scale tuning to C++ API (along to the already existing standard
209 SysEx way).
210 - lsatomic.h fixes: seq_cst load and store were suboptimal for x86
211 and broken for ppc64. (Seq_cst loads and stores are actually not
212 used in LS, so the bug wasn't noticable.)
213 - lsatomic.h: added ARMv7 support
214 - Added support for multiple MIDI input ports per sampler channel (and
215 added various new C++ methods for this new feature / design change, old
216 C++ API methods for managing SamplerChannel's MIDI inputs are now marked
217 as deprecated but are still there and should provide full behavior
218 backward compatibility).
219 - AbstractEngine::GSChecksum(): don't allocate memory on the stack (was
220 unsafe and caused compilation error with clang 2.x).
221 - Bugfix: only process the latest MIDI program change event.
222 - Introducing the LSCP shell, which provides convenient control of the
223 sampler from the command line by providing LSCP aware features.
224 - VirtualMidiDevice: Added support for program change.
225 - VirtualMidiDevice: Added support for bank select (MSB & LSB).
226 - VirtualMidiDevice: Added support for pitch bend.
227 - Aftertouch: extended API to explicitly handle channel pressure and
228 polyphonic key pressure events (so far polyphonic pressure was not
229 supported at all, and channel pressure was rerouted as CC128 but not
230 used so far).
231 - Added initial support for real-time instrument scripts. The script VM
232 code is shared by all sampler engine implemementations, however only the
233 gig file format currently provides support for storing instrument scripts
234 (as LinuxSampler extension to the original GigaStudio 4 format).
235
236 * Gigasampler format engine:
237 - implemented the "round robin keyboard" dimension
238 - fixed round robin and random dimensions for cases when number of
239 dimension zones is not a power of two
240 - made round robin use a counter for each region instead of each
241 key
242 - bugfix: pitch LFO controller "internal+aftertouch" was broken
243 - bugfix: filter keyboard tracking was broken
244 - filter performance fix (an unnecessary copy was made of the
245 filter parameters in each sub fragment)
246 - handle special case when pan parameter in gig file has max or
247 min value
248 - Exclusive Groups: don't ever stop voices of the same note,
249 doesn't sound naturally with a drumkit
250 - fixed EG1 modulation when attack or release is zero
251 - Fixed support for 'aftertouch' attenuation controller.
252 - Fixed crash that happened with velocity split sounds under certain
253 conditions (see also previous commit on libgig).
254 - fixed behaviour of filter LFO
255
256 * SFZ format engine:
257 - Initial implementation (not usable yet)
258 - added support for v2 multiple stage envelope generators
259 - added a fine-tuned v1 envelope generator instead of using the
260 one from the gig engine
261 - fixed memory leak and memory handling errors
262 - added support for trigger=first, trigger=legato and sw_previous
263 - allow non-numerical key values ("C#4" for example)
264 - "key" opcode now sets pitch_keycenter too
265 - fixed error when unloading instrument with same sample used by
266 multiple regions
267 - added some opcode aliases, like loopmode for loop_mode, to be
268 more compatible
269 - added support for trigger=release and rt_decay
270 - added support for off_mode=normal
271 - added support for random, seq_position, seq_length and volume
272 - added v1 LFO opcodes to parser (no support in engine yet)
273 - added support for amp_veltrack and amp_velcurve_N
274 - fine-tuned the default velocity curve
275 - added support for transpose
276 - fixed crash when using sw_down/up
277 - improved logic for sw_lokey/hikey/up/down/last
278 - added more v1 aliases to parser
279 - reduced memory usage for sfz data
280 - RT-safeness: avoid malloc in audio thread
281 - fixed a bug that could cause voice stealing to fail
282 - optimized sample lookup
283 - improved support for exclusive groups (group, off_by and
284 off_mode)
285 - added support for controller triggered regions (on_locc/on_hicc)
286 - added support for loop_mode=one_shot
287 - fixed playback of 16 bit wav files on big endian CPUs
288 - added support for Ogg Vorbis sample files
289 - use loop markers from sample file if loop_start and loop_end are
290 not set in sfz file
291 - implemeted filters. Filter types: lowpass, bandpass, bandreject
292 and highpass. 1, 2, 4 and 6 pole filters. Opcodes: fil_type,
293 cutoff, resonance, fil_veltrack, fil_keytrack, fil_keycenter,
294 cutoff_cc, cutoff_chanaft.
295 - bugfix: zero ampeg_sustain didn't work
296 - added support for velocity effect on amplifier envelope time
297 (ampeg_vel2attack, ampeg_vel2decay, ampeg_vel2sustain and
298 ampeg_vel2release)
299 - added support for EG hold (ampeg_hold)
300 - added support for sample offset (offset)
301 - use the newly introduced signal units model
302 - implemented opcodes lfoN_delay,
303 lfoN_freq, lfoN_pan, lfoN_cutoff, lfoN_resonance
304 - implemented opcodes ampeg_delay, ampeg_vel2delay
305 pitcheg_delay, pitcheg_start, pitcheg_attack, pitcheg_hold,
306 pitcheg_decay, pitcheg_sustain, pitcheg_release, pitcheg_vel2delay,
307 pitcheg_vel2attack, pitcheg_vel2hold, pitcheg_vel2decay,
308 pitcheg_vel2sustain, pitcheg_vel2release, pitcheg_depth
309 - implemented opcodes pitchlfo_delay, pitchlfo_freq,
310 pitchlfo_depth, fillfo_delay, fillfo_freq, fillfo_depth,
311 amplfo_delay, amplfo_freq, amplfo_depth
312 - implemented opcodes fileg_delay, fileg_start, fileg_attack,
313 fileg_hold, fileg_decay, fileg_sustain, fileg_release, fileg_vel2delay,
314 fileg_vel2attack, fileg_vel2hold, fileg_vel2decay, fileg_vel2sustain,
315 fileg_vel2release
316 - implemented opcode lfoN_wave
317 - implemented opcode pitchlfo_depthccN
318 - implemented opcodes lfoN_phase, lfoN_phase_onccX,
319 lfoN_pitch, lfoN_pitch_onccX
320 - implemented opcodes pitchlfo_fade,
321 fillfo_fade, amplfo_fade, lfoN_fade, lfoN_fade_onccX
322 - implemented opcodes fillfo_freqccN,
323 pitchlfo_freqccN, amplfo_freqccN, lfoN_freq_onccX
324 - implemented opcodes ampeg_delayccN, ampeg_startccN,
325 ampeg_attackccN, ampeg_holdccN, ampeg_decayccN, ampeg_sustainccN,
326 ampeg_releaseccN, egN_timeX_onccY, egN_levelX_onccY
327 - lfoN_* and egN_* opcodes defined in group sections
328 are now taken into account
329 - implemented curves
330 - implemented opcodes volume_onccN, volume_curveccN
331 - implemented opcode volume_smoothccN
332 - implemented opcodes fillfo_depthccN, amplfo_depthccN,
333 lfoN_volume, lfoN_volume_onccX, lfoN_volume_smoothccX,
334 lfoN_freq_smoothccX, lfoN_pitch_smoothccX, lfoN_pan_onccX,
335 lfoN_pan_smoothccX, lfoN_cutoff_onccX, lfoN_cutoff_smoothccX,
336 lfoN_resonance_onccX, lfoN_resonance_smoothccX, lfoN_delay_onccX
337 - implemented opcode end
338 - implemented opcodes egN_amplitude_onccX,
339 egN_volume, egN_volume_onccX, egN_cutoff, egN_cutoff_onccX,
340 egN_pitch, egN_pitch_onccX, egN_resonance, egN_resonance_onccX
341 - implemented opcodes xfin_lokey, xfin_hikey,
342 xfout_lokey, xfout_hikey, xf_keycurve, xfin_lovel, xfin_hivel,
343 xfout_lovel, xfout_hivel, xf_velcurve, xfin_loccN, xfin_hiccN,
344 xfout_loccN, xfout_hiccN, xf_cccurve
345 - implemented opcodes pan_onccN, pan_smoothccN,
346 pan_curveccN, egN_pan, egN_pan_curve, egN_pan_onccX,
347 egN_pan_curveccX, pitch_veltrack
348 - when failed to parse a sfz file
349 print the line number on which the error occurs
350 - use common pool of CC objects to minimize RAM usage
351 - implemented opcodes amplfo_delay_onccN,
352 amplfo_fade_onccN, fillfo_delay_onccN, fillfo_fade_onccN,
353 pitchlfo_delay_onccN, pitchlfo_fade_onccN
354 - implemented opcodes fileg_delay_onccN,
355 fileg_start_onccN, fileg_attack_onccN, fileg_hold_onccN,
356 fileg_decay_onccN, fileg_sustain_onccN, fileg_release_onccN,
357 fileg_depth_onccN, pitcheg_delay_onccN, pitcheg_start_onccN,
358 pitcheg_attack_onccN, pitcheg_hold_onccN, pitcheg_decay_onccN,
359 pitcheg_sustain_onccN, pitcheg_release_onccN, pitcheg_depth_onccN
360 - implemented automatic aliasing *ccN <-> *_onccN
361 - *lfo_freqccN wasn't working when the respective *lfo_freq
362 was not set or was set to zero
363 - lfoN_freq_onccX wasn't working when lfoN_freq
364 was not set or was set to zero
365 - implemented opcodes resonance_onccN, resonance_smoothccN,
366 resonance_curveccN, cutoff_smoothccN, cutoff_curveccN
367 - implemented opcodes fillfo_depthchanaft,
368 fillfo_freqchanaft, amplfo_depthchanaft, amplfo_freqchanaft,
369 pitchlfo_depthchanaft, pitchlfo_freqchanaft
370 - implemented opcodes pitch_onccN,
371 pitch_curveccN, pitch_smoothccN, pitch_stepccN
372 - implemented opcodes volume_stepccN, pan_stepccN,
373 cutoff_stepccN, resonance_stepccN, lfoN_freq_stepccX,
374 lfoN_volume_stepccX, lfoN_pitch_stepccX, lfoN_pan_stepccX,
375 lfoN_cutoff_stepccX, lfoN_resonance_stepccX
376 - implemented opcodes eq1_freq, eq2_freq, eq3_freq,
377 eq1_freqccN, eq2_freqccN, eq3_freqccN, eq1_bw, eq2_bw, eq3_bw,
378 eq1_bwccN, eq2_bwccN, eq3_bwccN, eq1_gain, eq2_gain, eq3_gain,
379 eq1_gainccN, eq2_gainccN, eq3_gainccN
380 - implemented opcodes delay, delay_onccN, delay_random,
381 delay_samples, delay_samples_onccN
382 - implemented opcodes egN_eq1gain, egN_eq2gain, egN_eq3gain,
383 egN_eq1gain_onccX, egN_eq2gain_onccX, egN_eq3gain_onccX, egN_eq1freq,
384 egN_eq2freq, egN_eq3freq, egN_eq1freq_onccX, egN_eq2freq_onccX,
385 egN_eq3freq_onccX, egN_eq1bw, egN_eq2bw, egN_eq3bw, egN_eq1bw_onccX,
386 egN_eq2bw_onccX, egN_eq3bw_onccX, lfoN_eq1gain, lfoN_eq2gain,
387 lfoN_eq3gain, lfoN_eq1gain_onccX, lfoN_eq2gain_onccX, lfoN_eq3gain_onccX,
388 lfoN_eq1gain_smoothccX, lfoN_eq2gain_smoothccX, lfoN_eq3gain_smoothccX,
389 lfoN_eq1gain_stepccX, lfoN_eq2gain_stepccX, lfoN_eq3gain_stepccX,
390 lfoN_eq1freq, lfoN_eq2freq, lfoN_eq3freq, lfoN_eq1freq_onccX,
391 lfoN_eq2freq_onccX, lfoN_eq3freq_onccX, lfoN_eq1freq_smoothccX,
392 lfoN_eq2freq_smoothccX, lfoN_eq3freq_smoothccX, lfoN_eq1freq_stepccX,
393 lfoN_eq2freq_stepccX, lfoN_eq3freq_stepccX, lfoN_eq1bw, lfoN_eq2bw,
394 lfoN_eq3bw, lfoN_eq1bw_onccX, lfoN_eq2bw_onccX, lfoN_eq3bw_onccX,
395 lfoN_eq1bw_smoothccX, lfoN_eq2bw_smoothccX, lfoN_eq3bw_smoothccX,
396 lfoN_eq1bw_stepccX, lfoN_eq2bw_stepccX, lfoN_eq3bw_stepccX
397 - implemented opcodes eq1_vel2freq, eq2_vel2freq,
398 eq3_vel2freq, eq1_vel2gain, eq2_vel2gain, eq3_vel2gain
399 - sfz parser: allow double spaces in sample filenames
400 - sfz parser: allow absolute paths for sample filenames
401 - use linear decay and release for filter and pitch EG
402 - bugfix: only the first amp_veltrack definition in a file was
403 used
404 - bugfix: looping was disabled if loop_start was set to 0
405 - allow regions with end=-1 to turn off other regions using the
406 group and off_by opcodes (#168)
407 - made end=0 play the whole sample
408 - fixed support for lochan and hichan opcodes (#155)
409 - fixed crash when using lochan/hichan opcodes (#187)
410 - sfz parser: allow -200 to 200 for pan_oncc opcode (#182)
411 - added FLAC support (#191)
412 - sfz parser bugfix: lines starting with whitespace were ignored
413 - added amplitude opcode
414 - added support for "#include" instruction
415 (modified patch which was originally posted by Sergey on LS mailing list)
416 - bugfix: generation of velocity curves etc should not be done
417 after each "#include", only after the main file is parsed
418 - bugfix: line numbers in error messages were wrong after "#include"
419 - added support for float and 32 bit sample files
420
421 * SoundFont format engine:
422 - Initial implementation (not usable yet)
423 - RT-safeness: avoid malloc in audio thread
424 - fixed a bug that could cause voice stealing to fail
425 - fine-tuned amplitude EG (by switching from gig to sfz EG)
426 - initial implementation of Vibrato LFO and Modulation LFO
427 - initial implementation of cutoff filter
428 - use linear decay and release for filter and pitch EG
429
430 * Host plugins (VST, AU, LV2, DSSI):
431 - AU bugfix: failed to destroy its audio/MIDI devices
432 - Listen to all interfaces on Mac OS X (INADDR_ANY)
433 - VST bugfix: If the host called resume() before and after
434 changing sample rate or block size, the number of channels was
435 incorrectly set to two. This caused silence in Cubase 5.
436 - save engine type (gig, sfz or sf2) in plugin state
437 - VST: when opening Fantasia, look for both 32 and 64 bit Java on
438 64 bit Windows
439 - AU: changed number of output channels from one stereo to 16
440 stereo
441 - VST: made it possible to build the VST plugin for Mac
442 - AU: link AU plugin dynamically if --disable-shared isn't
443 specified
444 - LV2 "state" extension support (patch by David Robillard)
445 - VST bugfix: instrument loading hang and crashed the host when
446 the plugin was loaded a second time (#174)
447 - plugin bugfix: instrument loading hang when the plugin was
448 loaded a second time (this time it's for Linux and Mac, previous
449 similar fix was for Windows)
450 - thread safety fixes for the instrument loading thread
451 - LV2: use the new lv2 package if present
452 - VST: try to open Fantasia automatically on Linux and Mac too (on
453 Linux, the Fantasia jar should be placed in <prefix>/share/java)
454 - VST: fixed crashes on Linux Ardour and EnergyXT
455 - DSSI bugfix: it wasn't possible to change engine type. The MIDI
456 port and audio channel routing for DSSI plugins are now visible.
457 - LV2: use urid and atom extensions instead of deprecated uri-map
458 and event
459 - LV2: lv2 package 1.0 is now required to build the LV2 plugin
460 - LV2: changed number of output channels to 16 stereo, just like
461 the VST and AU plugins
462 - LV2: fixed save/restore of SFZ state (patch by David Robillard)
463 - LV2: made LV2 plugin buildable on Windows and Mac
464 - VST: implemented retrieval and switching of programs using the
465 sampler's internal MIDI instrument mapping system
466
467 * Instrument editor interface:
468 - Changed instrument editor plugin interface, providing additional
469 informations like the EngineChannel for which the instrument editor was
470 spawned for. This allows the instrument editors to interact more actively
471 with the sampler.
472
473 * MIDI driver:
474 - ALSA MIDI driver supports now "NAME" device parameter, for overriding
475 the ALSA sequencer client name
476 - removed limit of maximum amount of MIDI ports per MIDI device, since
477 there is no reason for this limit
478 - MME: fixed memory handling bug found with cppcheck
479 - MME: removed compiler warning
480 - CoreMIDI: implemented driver specific port parameter "CORE_MIDI_BINDINGS",
481 which allows to retrieve the list of CoreMIDI clients / ports and to connect
482 to them a la JACK, via the usual sampler APIs
483 - CoreMIDI: added driver specific port parameter "AUTO_BIND", if enabled
484 the driver will automatically connect to other CoreMIDI clients' ports
485 (e.g. external MIDI devices being attached to the Mac)
486 - added support for MIDI note on velocity filter
487 - CoreMIDI: fixed memory deallocation error
488 - Fixed variable underflow in VirtualMidiDevice, which caused graphical
489 virtual keyboards in frontends / instrument editors being stuck.
490 - Bugfix in VirtualMidiDevice: process note on with velocity 0 as note off.
491 - Implemented missing handling of MIDI "running status".
492 - CoreMIDI fix: a MIDIPacket can contain more than one event per packet.
493 - MME bugfix: driver wasn't closed properly
494 - CoreMIDI: automatically connect to all input sources by default (driver
495 parameter "AUTO_BIND").
496 - CoreMIDI: fixed auto bind feature to CoreMIDI ports that go online
497 - CoreMIDI: fixed minor memory leak
498
499 * audio driver:
500 - ASIO driver fixes for newer gcc versions (fix from PortAudio)
501 - JACK audio: react on sample rate changes.
502 - JACK audio: react on buffer size changes.
503 - JACK audio: jack_port_get_buffer() was cached and called outside
504 RT context.
505 - ASIO driver: removed compiler warnings
506 - CoreAudio: fixed minor error handling bug
507 - ASIO driver: be more verbose when no ASIO card could be found (fixes #203)
508 - JACK audio: return the JACK server's current sample rate as default value
509 for audio device parameter "SAMPLERATE" (fixes #166).
510
511 * LSCP server:
512 - added support for sending MIDI CC messages via LSCP command
513 "SEND CHANNEL MIDI_DATA CC <sampler-chan> <ctrl> <val>"
514 - added LSCP command "GET AVAILABLE_EFFECTS"
515 - added LSCP command "LIST AVAILABLE_EFFECTS"
516 - added LSCP command "GET EFFECT INFO <effect-index>"
517 - added LSCP command "CREATE EFFECT_INSTANCE <effect-index>"
518 - added LSCP command
519 "CREATE EFFECT_INSTANCE <effect-system> <module> <effect-name>"
520 - added LSCP command "DESTROY EFFECT_INSTANCE <effect-instance>"
521 - added LSCP command "GET EFFECT_INSTANCES"
522 - added LSCP command "LIST EFFECT_INSTANCES"
523 - added LSCP command "GET EFFECT_INSTANCE INFO <effect-instance>"
524 - added LSCP command
525 "GET EFFECT_INSTANCE_INPUT_CONTROL INFO <effect-instance> <input-control>"
526 - added LSCP command "SET EFFECT_INSTANCE_INPUT_CONTROL VALUE
527 <effect-instance> <input-control> <value>"
528 - added LSCP command "GET SEND_EFFECT_CHAINS <audio-device>"
529 - added LSCP command "LIST SEND_EFFECT_CHAINS <audio-device>"
530 - added LSCP command "ADD SEND_EFFECT_CHAIN <audio-device>"
531 - added LSCP command
532 "REMOVE SEND_EFFECT_CHAIN <audio-device> <effect-chain>"
533 - added LSCP command
534 "GET SEND_EFFECT_CHAIN INFO <audio-device> <effect-chain>"
535 - added LSCP command "APPEND SEND_EFFECT_CHAIN EFFECT <audio-device>
536 <effect-chain> <effect-instance>"
537 - added LSCP command "INSERT SEND_EFFECT_CHAIN EFFECT <audio-device>
538 <effect-chain> <effect-chain-pos> <effect-instance>"
539 - added LSCP command "REMOVE SEND_EFFECT_CHAIN EFFECT <audio-device>
540 <effect-chain> <chain-pos>"
541 - added LSCP command "SET FX_SEND EFFECT <sampler_channel>
542 <fx_send_id> <effect_chain> <chain_pos>"
543 - added LSCP command "REMOVE FX_SEND EFFECT <sampler_channel> <fx_send_id>"
544 - added LSCP commands "SUBSCRIBE EFFECT_INSTANCE_COUNT",
545 "SUBSCRIBE EFFECT_INSTANCE_INFO", "SUBSCRIBE SEND_EFFECT_CHAIN_COUNT",
546 "SUBSCRIBE SEND_EFFECT_CHAIN_INFO"
547 - provide comprehensive error messages on LSCP syntax errors
548 (suggesting expected next non-terminal symbols)
549
550 * LSCP shell:
551 - Added support for auto correction of obvious and trivial syntax mistakes.
552 - Added support for auto completion by tab key.
553 - Show currently available auto completion while typing.
554 - Added support for browsing command history with up / down keys.
555 - Show all possible next symbols immediately right to the current command
556 line while typing (no double tab required for this feature, as it would
557 be the case in other shells).
558 - Added support for moving cursor left/right with arrow keys.
559 - Added support for built-in LSCP reference documentation, which will
560 automatically show the relevant LSCP reference section on screen as soon
561 as one specific LSCP command was detected while typing on the command
562 line.
563
564 * Real-time instrument scripts:
565 - Implemented built-in script array variable %CC.
566 - Implemented built-in script int variable $CC_NUM.
567 - Implemented built-in script int variable $EVENT_NOTE.
568 - Implemented built-in script int variable $EVENT_VELOCITY.
569 - Implemented built-in script constant variable $VCC_MONO_AT.
570 - Implemented built-in script constant variable $VCC_PITCH_BEND.
571 - Implemented execution of script event handler "init".
572 - Implemented execution of script event handler "controller".
573 - Implemented execution of script event handler "note".
574 - Implemented execution of script event handler "release".
575 - Implemented built-in script function "play_note()" (only two of the
576 max. four function arguments are currently implemented yet though).
577 - Implemented built-in script int variable $EVENT_ID.
578 - Implemented built-in script function "ignore_event()"
579 - Implemented built-in script function "ignore_controller()" (may have one
580 or no argument).
581 - Implemented built-in script function "set_controller()".
582 - Added extended script VM for the Gigasampler/GigaStudio format sampler
583 engine, which extends the general instrument script VM with Giga format
584 specific variables and functions.
585 - Giga format scripts: added built-in script int constant variables
586 $GIG_DIM_CHANNEL, $GIG_DIM_LAYER, $GIG_DIM_VELOCITY, $GIG_DIM_AFTERTOUCH,
587 $GIG_DIM_RELEASE, $GIG_DIM_KEYBOARD, $GIG_DIM_ROUNDROBIN, $GIG_DIM_RANDOM,
588 $GIG_DIM_SMARTMIDI, $GIG_DIM_ROUNDROBINKEY, $GIG_DIM_MODWHEEL,
589 $GIG_DIM_BREATH, $GIG_DIM_FOOT, $GIG_DIM_PORTAMENTOTIME, $GIG_DIM_EFFECT1,
590 $GIG_DIM_EFFECT2, $GIG_DIM_GENPURPOSE1, $GIG_DIM_GENPURPOSE2,
591 $GIG_DIM_GENPURPOSE3, $GIG_DIM_GENPURPOSE4, $GIG_DIM_SUSTAIN,
592 $GIG_DIM_PORTAMENTO, $GIG_DIM_SOSTENUTO, $GIG_DIM_SOFT,
593 $GIG_DIM_GENPURPOSE5, $GIG_DIM_GENPURPOSE6, $GIG_DIM_GENPURPOSE7,
594 $GIG_DIM_GENPURPOSE8, $GIG_DIM_EFFECT1DEPTH, $GIG_DIM_EFFECT2DEPTH,
595 $GIG_DIM_EFFECT3DEPTH, $GIG_DIM_EFFECT4DEPTH, $GIG_DIM_EFFECT5DEPTH.
596 - Giga format scripts: Implemented built-in script function
597 "gig_set_dim_zone(event_id, dimension, zone)".
598 - Implemented built-in script int array variable $KEY_DOWN.
599 - Implemented built-in script function "abs()".
600 - Implemented built-in script function "random()".
601 - Implemented built-in script function "num_elements()".
602 - Implemented built-in script function "note_off()".
603 - Implemented built-in script function "set_event_mark()".
604 - Implemented built-in script function "delete_event_mark()".
605 - Implemented built-in script function "by_marks()".
606 - Added built-in script int const variables $MARK_1 to $MARK_28.
607 - Built-in script functions "ignore_event()", "note_off()" and
608 "gig_set_dim_zone()" now also accept an array of event IDs as argument
609 (i.e. return value of new script function "by_marks()").
610 - Pass/preserve polyphonic variable data from respective "note" event
611 handler to "release" event handler.
612
613 * Bug fixes:
614 - Fixed crash which may occur when MIDI key + transpose is out of range
615 - minor valgrind fixes
616 - fixed crash which occurred when changing an already deployed sampler
617 channel to a different engine type
618 - fixed crash when deleting a sampler channel or changing engine
619 type while an instrument load was in progress
620 - bugfix: playing a note while changing the instrument could cause
621 a crash, or give "Handing back unknown region" error messages
622 - bugfix: calling SET VOICES (which Fantasia does on start and
623 refresh) could cause instruments to be unloaded
624 - Mac OS X: fixed name collision of enum in EffectControl and
625 wrong name of destructor in AudioOutputDeviceCoreAudio.cpp
626 - Mac OS X: fixed hanging threads
627 - Windows: disabled the previous thread fix on non-Mac systems, as
628 it caused hanging threads on Windows
629 - Fixed possible crashes due to corrupted MIDI/audio device list
630 after MIDI/audio device creation failure
631 - When creating MIDI instrument map entries with "PERSISTENT" type, the
632 instruments were uselessly precached with zero samples, however it still
633 took the full preloading time and on 1st program change the respective
634 instrument was completely reloaded again.
635 - fixed handling of rapid bank select and program change messages
636 sent to the same sampler channel (patch from the Open Octave
637 project, slightly adjusted)
638 - fixed crash when trying to create an effect instance with controls
639 which min and/or max values depend on the sample rate
640 - fixed bug #162
641 - bugfix: LADSPA_PATH was not evaluated correctly when containing
642 multiple paths (#165)
643 - thread safety fixes for the instrument loading thread
644 - bugfix: instrument loading crashed for sfz and sf2 in Ardour
645 (#176)
646 - more thread safety fixes for the instrument loading thread
647 - sfz/sf2 engine: fixed crash when using small audio fragment size
648 - Mac OS X: fixed crash when unloading plugin on 10.7 and later
649 - Mac OS X: fixed process hang when unloading 32-bit plugin (bug
650 introduced in previous fix)
651 - fixed crash when a channel received a program change while
652 playing a note in a key group
653 - fixed erroneous error message piping in VoiceBase.h
654 (labelled "Disk stream not available in time")
655 - Update effects on sample rate & period size changes (to avoid
656 crashes and noise on such transitions).
657 - ignore missing LADSPA paths without ignoring valid LADSPA paths
658 (fixes #208)
659 - Fixed bug in LSCP grammar definition which caused a statement like
660 "GET SERVER INFOasdf\n" to be accepted as valid statement (was so far
661 practically irrelevant, however it caused problems with the new LSCP
662 shell's auto completion feature).
663 - Fixed MIDI program change messages being ignored if quickly executed
664 after each other (fixes #231).
665
666 Version 1.0.0 (31 July 2009)
667
668 * packaging changes:
669 - autoconf bugfix: the PKG_CONFIG variable wasn't initialized properly,
670 causing e.g. the libgig test to fail when
671 "./configure --disable-jack-driver" was used
672 (patch by Alexis Ballier)
673 - fixed compilation with gcc 4.3
674 - fixes for building on OS X (thanks to Ebrahim Mayat for testing)
675 - fixed configure so it detects x86_64 (#107)
676 - fixes for building with newer MinGW versions
677 - fix for building with bison 2.4 (#111)
678 - fixed building with libgig installed in a non-standard directory
679 - minor fix in configure for mmsystem.h detection on MinGW
680 - Windows: look for editor plugins and Fantasia using base
681 directory of liblinuxsampler dll (look in the same directory and one
682 directory above)
683 - configure script fix: removed unconditional use of SSE
684 - fixed building with sqlite installed in a non-standard directory
685 - when cross-compiling, don't try to create instruments.db
686 - fix for new mingw-w64 version, which has usleep
687
688 * general changes:
689 - bugfix: on some POSIX systems instrument editor plugins refused to
690 load as we used a non-portable Linux specific struct field
691 (fixes bug #70, patch by Ronald Baljeu)
692 - fixed endless loop which occured when loading an instrument editor
693 plugin DLL on Windows systems failed
694 - fixed memory leaks that occurred when liblinuxsampler was unloaded
695 - optimized the SynchronizedConfig class so it doesn't wait
696 unnecessarily long after an update
697 - added support for notifying instrument editors on note-on / note-off
698 events (e.g. to highlight the pressed keys on the virtual keyboard
699 of gigedit)
700 - added support for triggering notes by instrument editors (see above)
701 - be verbose on DLL load errors (on Linux)
702 - fixed di-harmonic triangle LFO implementation (this LFO implementation
703 is only used by very few systems, most pick the int math
704 implementation, which is usually faster)
705 - fixes for audio drivers with varying buffer sizes
706 - experimental support for running LinuxSampler as a DSSI, LV2 and
707 VST plugin
708 - notification events for stream/voice count statistics are now sent
709 only when there are actual changes
710 - added memory ordering constraints to improve stability on
711 multi-core and multi-cpu systems
712 - maximum voices and disk streams can now be altered at runtime
713 - fixed CPU feature detection on x86_64 (maybe fixes #108)
714 - automatic stacktrace mechanism is now turned off by default and can
715 be switched on by command line option "--stacktrace" (the automatic
716 stacktrace mechanism seems to be broken on most systems at the moment)
717 - C++ API method InstrumentManager::LaunchInstrumentEditor() now returns
718 a pointer to the launched InstrumentEditor object
719 - added optional 3rd party user data parameter for following
720 liblinuxsampler C++ API methods: InstrumentEditor::Main(),
721 InstrumentEditor::Launch(),
722 InstrumentManager::LaunchInstrumentEditor()
723 - theoretical fix: made SynchronizedConfig follow C++0x memory
724 model more strictly
725 - fixes for using large audio device buffers
726 - Windows: add the installation directory to the DLL search path
727 when loading an editor plugin (solves problems with VST and
728 gigedit on systems with other GTK versions installed)
729 - updated linuxsampler man page
730
731 * audio driver:
732 - removed the nonsense audio channel constraint (which was hard coded to
733 max. 100 audio channels) for most audio drivers
734 - JACK audio driver did not offer a device parameter "SAMPLERATE" as
735 opposed to the LSCP specs
736 - bugfix: the SAMPLERATE parameter of some drivers (e.g. JACK)
737 reflected the wrong value
738 - fixed a memory management bug in ASIO driver
739 - Makefile fix: JACK_CFLAGS wasn't used
740 - JACK: use jack_client_open instead of the deprecated
741 jack_client_new
742 - added (experimental) CoreAudio driver
743 - applied old fixes to the ASIO driver that were included in the
744 previous binary release but accidentally never committed to CVS
745 (fixes #117)
746 - fixes for ASIO on mingw-w64 (iasio wrapper is not needed on
747 win64)
748 - VST: added support for sample rate and buffer size changes
749 - VST: close editor (Fantasia) when the VST is removed
750 - VST: avoid opening Fantasia more than once for each VST instance
751 - VST: export main function as "main" on Linux too (fix for
752 energyXT)
753 - VST: changed number of output channels from one stereo to 16
754 stereo
755 - added channel routing, fxsends and midi maps to the settings
756 stored in the plugin state
757 - performance optimization of AudioChannel::MixTo() and
758 AudioChannel::CopyTo() methods using GCC vector exensions
759 (if available)
760 - ASIO fixes: avoid initializing the device twice, avoid throwing
761 exception when getting parameters from a disconnected device
762
763 * MIDI driver:
764 - added JACK MIDI driver
765 - dispatch bank select as ordinary CC as well, the user might seriously
766 want to (mis)use it for some purpose ("fixed" in all current MIDI
767 input drivers: ALSA, CoreMIDI, JACK, MidiShare, MME)
768 - bugfix: pitch bend wasn't working with JackMidi, VST, LV2, MME,
769 CoreMidi or AU
770 - fixed mingw-w64 compilation error in MME driver
771 - made program change handling in MIDI thread real-time safe by
772 moving the logic to a non-RT thread
773 - fixed minor memory leak in ALSA driver
774
775 * instruments database:
776 - avoid time consuming samples scanning when adding instruments
777 to the instruments database
778 - added support for handling lost files in the instruments database
779 - Implemented option for adding instruments in separate directories
780 in the instruments database
781 (patch by Chris Cherrett & Andrew Williams, a bit adjusted)
782 - work-around for missing fnmatch function on Windows to make
783 instrument database compilable
784 - added instrument database support on Windows
785 works with both standalone mode and VST plugin.
786 the instruments DB file is located in
787 %USERPROFILE%\.linuxsampler\instruments.db which allows different
788 databases for each windows user
789 if no DB is present it automatically creates the .linuxsampler subdir
790 and then creates an empty DB
791 - fixed recursive import, which was broken on Windows
792
793 * Gigasampler format engine:
794 - fixed a memory leak that could happen when a channel was deleted
795 while notes were playing
796 - made it possible to load an instrument even if the audio thread
797 isn't running
798 - added partial support for the "Controller Triggered" MIDI rule,
799 enough for piano gigs with pedal noise samples
800 - minor fix: only mark FX sends as being modified if really the
801 respective FX send MIDI controller was used
802 - added support for GM global device volume SysEx message
803 (can be selected at compile time to either apply globally to the
804 whole sampler [default] or only to the sampler channels that are
805 connected to the respective MIDI input port the SysEx message arrived
806 on)
807 - bugfix: notes triggered at position 0 in the audio buffer were
808 sometimes wrongly killed in the same buffer, causing no sound to
809 be played
810 - added support for chromatic / drumkit mode Roland GS Sysex message
811 which is usally used to switch a part between chromatic sounds and
812 drumkit sounds (as opposed to their common default setting of MIDI
813 part 10 being a drumkit part and all other ones chromatic parts), in
814 LS however one can switch between the first 16 MIDI instrument maps
815 defined for the sampler
816 - made it possible to create multiple sample channels even if the
817 audio thread isn't running
818 - fixed crash when removing channel with active voices (#116)
819 - bugfix: on sample reference changes (instrument editor), only
820 un-cache the respective sample if it's really not used by any
821 sampler engine anymore
822 - re-cache samples in case they were changed by an instrument editor,
823 e.g. when a sample was added while playing (#82)
824 - fixed hanging notes which occured when note-off event had the exact
825 same time stamp as the voice's note-on event and occured both in the
826 same audio fragment cycle (fixes bug #112)
827 - added support for the "fine tune" and "pitch bend range"
828 instrument-level gig parameters
829 - fixed minor artifacts in pitch bend handling
830 - added support for GS Reset SysEx message
831 - allow gig files to use unlimited downward pitch shifting
832 - added a limit check for upward pitch shifting
833 - bugfix: sometimes, when playing a note twice fast, the second
834 note was silent
835 - fixed crash happening when a pitch bend event arrived at the
836 same time a new instrument was loading
837
838 * LSCP server:
839 - added new LSCP event "CHANNEL_MIDI" which can be used by frontends to
840 react on MIDI data arriving on certain sampler channels (so far only
841 Note-On and Note-Off events are sent via this LSCP event)
842 - added new LSCP event "DEVICE_MIDI" which can be used by frontends to
843 react on MIDI data arriving on certain MIDI input devices (so far only
844 Note-On and Note-Off events are sent via this LSCP event)
845 - added new LSCP commands: FIND LOST DB_INSTRUMENT_FILES and
846 SET DB_INSTRUMENT FILE_PATH
847 - added new LSCP command: SEND CHANNEL MIDI_DATA which can be used by
848 frontends to send MIDI messages to specific sampler channel
849 - added two additional fields to GET FILE INSTRUMENT INFO command -
850 KEY_BINDINGS and KEYSWITCH_BINDINGS
851 - bugfix: the bank number provided by MIDI_INSTRUMENT_INFO notifications
852 was incorrect
853 - Optimized the retrieval of the MIDI instrument mappings
854 - added new LSCP command "SET VOICES" to globally alter the maximum
855 amount of voices, added new LSCP event "GLOBAL_INFO:VOICES" which will
856 be triggered respectively
857 - added new LSCP command "SET STREAMS" to globally alter the maximum
858 amount of disk streams, added new LSCP event "GLOBAL_INFO:STREAMS"
859 which will be triggered respectively
860 - bugfix: retry if "select" returns EINTR (this fixes a crash when
861 a gigedit file dialog is opened)
862 - close all connections when LSCPServer is deleted
863 - hard close of all sockets on exit
864 - bugfix: SET CHANNEL MIDI_INPUT_TYPE didn't work with the MME
865 driver
866
867 * Bug fixes:
868 - fixed a crash which occurs when removing a sampler channel waiting
869 to start instrument loading after another channel
870 - fixed a crash which occurs when removing a sampler channel with
871 instrument loading in progress (bug #113)
872 - fixed termination caused by uncaught exception when adding MIDI
873 instrument with PERSISTENT load mode
874 - fixed possible iterator invalidations when resetting the sampler
875 - fixed memory leaks when issuing the following LSCP commands:
876 GET AUDIO_OUTPUT_DRIVER INFO
877 GET AUDIO_OUTPUT_DRIVER_PARAMETER INFO
878 GET MIDI_INPUT_DRIVER INFO
879 GET MIDI_INPUT_DRIVER_PARAMETER INFO
880 - fixed possible compilation error when sqlite is not present
881 - fixed orphaned pointers when setting maximum voices limit (bug #118)
882 - fixed crash when changing the audio output device of a sampler
883 channel with loaded instrument and start playing notes
884 - fixed endless loop in Engine::SuspendAll() (bug #120)
885 - fixed a low-level atomic load function that was broken on 64-bit
886 PowerPC, which probably could cause crashes on that platform
887 - fixed a memory management error which could cause a crash when a
888 plugin was unloaded
889 - bugfix: two private structs had the same name, which could cause
890 problems if the linker chose the wrong constructor
891 - fixed low-level ConditionServer usage bug that caused lockups on
892 Windows
893
894
895 Version 0.5.1 (6 December 2007)
896
897 * packaging changes:
898 - added autoconf checks for pthread library
899 - added autoconf check for pthread bug found on certain NPTL-enabled
900 glibc versions (see Gentoo bug report #194076)
901 - added autoconf checks for MS Windows
902
903 * general changes:
904 - bugfix: the thread used by an editor plugin didn't die when the
905 editor closed
906 - bugfix: Ringbuffer.h: fill_write_space_with_null() did not zero
907 out all the space. operator--() did not apply size_mask after
908 decrementing the read_ptr. DEFAULT_WRAP_ELEMENTS set to 0 in
909 order to avoid problems with the _NonVolatileReader functions.
910 - bugfix: Stream.h: added a missing divide by BytesPerSample in
911 GetWriteSpace(). Since this function is currently only used in
912 the stream's qsort() compare function, it didn't trigger any
913 bugs.
914 - Resampler.h, Synthesizer.h: cubic interpolation now works in
915 24bit mode too. Faster method to read 24bit words on little
916 endian machines (x86): replaced 3 byte reads + shifts with a 1
917 unaligned 32bit read and shift
918 - experimental support for MS Windows (MIDI input via MME, AUDIO
919 output via ASIO)
920 - made handling of SIGINT signal (Ctrl-C) a bit more robust
921 - support for monitoring the total number of active disk streams
922 (new LSCP commands: GET TOTAL_STREAM_COUNT,
923 SUBSCRIBE TOTAL_STREAM_COUNT, UNSUBSCRIBE TOTAL_STREAM_COUNT)
924
925 * AUDIO driver:
926 - added Windows ASIO low latency audio driver
927
928 * MIDI driver:
929 - added MME Windows MIDI driver
930
931 * LSCP server:
932 - added support for Windows style path / filenames, however with
933 forward slash path separators instead of backslash
934 (i.e. "C:/foo/bar.gig")
935 - allow naughty liblscp to send non-string device parameters within
936 apostrophes as well
937 - added new LSCP commands: "GET FILE INSTRUMENTS <file>",
938 "LIST FILE INSTRUMENTS <file>" and
939 "GET FILE INSTRUMENT INFO <file> <index>" for retrieving informations
940 about an arbitrary instrument file on the system where the sampler is
941 running on
942
943 Version 0.5.0 (15 October 2007)
944
945 * packaging changes:
946 - config.h is not going to be installed along with liblinuxsampler's
947 API header files anymore
948 - only the API relevant header (and source) files will be exported to
949 the Doxygen API documentation (explicitly listed in Doxyfile.in)
950 - added completely new XCode project files for Mac OSX which is now
951 capable to execute our autoconf environment, thus no need anymore
952 to maintain the compile time configuration file (osx/version.h) for
953 OSX manually (patch by Toshi Nagata)
954 - fixed buggy boolean --enable-foo / --disable-foo configure script
955 parameters
956 - global.h now only covers global definitions that are needed for the
957 C++ API header files, all implementation internal global definitions
958 are now in global_private.h
959 - atomic.h is not exposed to the C++ API anymore
960 - no need to include config.h anymore for using LS's API header files
961 - fixed warnings in API doc generation
962
963 * general changes:
964 - replaced the old, confusing MIDI program change mechanism by a
965 flexible MIDI instrument mapper which allows to map arbitrary
966 (MIDI bank, MIDI program) pairs with arbitrary
967 (engine type, instrument file, file index) triplets which will be
968 loaded on the respective channel when such MIDI program change
969 messages arrive, beside that, each entry allows to define a life-time
970 strategy for the instrument, so the user can define whether the
971 instrument should i.e. be permanently loaded (to be able to switch
972 quickly among sounds) or i.e. loaded just on demand when the
973 respective program change arrives (to safe RAM space), as well as a
974 global volume factor for each entry, so the user can adjust the volume
975 dynamic of his mapped instrument collection without having to modify
976 the instrument files, also one can manage arbitrary amount of such
977 MIDI instrument maps and assign each sampler channel individually
978 a certain map, so that i.e. separation of normal instruments and
979 drumkits is possible
980 - new notification events for tracking changes to audio/MIDI devices,
981 MIDI instrument maps, MIDI instruments, FX sends, global volume.
982 - sampler was limited to load max. 200 instruments in the background
983 due to a constant size RingBuffer FIFO which is now replaced by a
984 dynamic (unlimited) size std::list FIFO
985 - added FX sends, these allow to route audio signals to arbitrary audio
986 output channels for being processed by external effect processors
987 (i.e. jack-rack), the send levels are controllable via arbitrary MIDI
988 controllers
989 - global (sampler wide) volume can now be controlled at runtime
990 - Implemented new, improved notification system
991 - fixed compilation errors regarding OSX
992 (patch by Toshi Nagata)
993 - implemented instruments database
994 - added support for escape sequences to the instruments database
995 - added highly experimental support for on-the-fly instrument editing
996 within the sampler's process (by using instrument editor plugins),
997 you'll notice the new "Registered instrument editors:" message on
998 startup, added a new LSCP command:
999 "EDIT CHANNEL INSTRUMENT <sampler-channel>"
1000 to spawn a matching instrument editor for the instrument on the
1001 given sampler channel, the plugin path can be overridden at compile
1002 time with ./configure --enable-plugin-dir=/some/dir
1003 - added experimental code for synchronizing instrument editors hosted
1004 in the sampler's process to safely edit instruments while playing
1005 without a crash (hopefully) by either suspending single regions
1006 wherever possible or if unavoidable whole engine(s)
1007 - fixed several issues in fundamental "Thread" class: set scheduling
1008 policy and priority on thread level, set a minimum stack size for
1009 thread (TODO: a reasonable value yet to be tested), bugfix: non-RT
1010 threads simply inherited properties of starting thread instead of
1011 setting their own policy and priority
1012 - minor fix in our automatic stack trace mechanism on crashes, the
1013 main process did not wait for the stack trace process to finish
1014 its output
1015 - fixed some minor memory leaks
1016 - reenabled assembly features support, at the moment only for
1017 enabling a fast denormal FPU mode (x86 platforms supporting SSE2)
1018 - minor assembly fix in x86 features detection (don't use the PIC
1019 register, to avoid relocations in the text segment at runtime)
1020 - POSIX callback functions of Thread.h are hidden
1021
1022 * MIDI driver:
1023 - dispatch bank select (MSB and LSB) messages
1024
1025 * audio driver:
1026 - the ALSA audio output driver parameters now reflect the correct
1027 parameter value ranges for the respective selected sound card
1028 (patch by Till Wimmer, a bit fixed and extended)
1029
1030 * Gigasampler format engine:
1031 - if a filter is used and EG2 finishes before EG1, let the voice
1032 die when EG2 has finished (this fixes a problem with clicks and
1033 voice starvation for some gigs)
1034 - playback is no longer disabled during instrument loading
1035 - all notes playing on a channel that changes its instrument keep
1036 playing with the old instrument until they get a note off
1037 command
1038 - EG fix: a release value of zero could cause noises or crash
1039 - handle MIDI coarse tuning messages (MIDI RPN #0 MSB #2 LSB)
1040 - EG fine tuning: when attack is zero the EG starts at a level
1041 above max sustain level, which means that there is a decay phase
1042 even if sustain is 100%
1043 - more EG fixes: the level could sometimes go below zero and cause
1044 noises or crashes
1045 - minor fix of EGDecay (patch by Toshi Nagata)
1046 - fixed compiler error when --enable-override-filter-type was
1047 supplied to the configure script (fixes #46)
1048 - disk thread: queue sizes are now proportional to CONFIG_MAX_STREAMS
1049 instead of a fix value
1050 - behavior fix: on MIDI CC# 65 (portamento on / off), 126 (mono mode),
1051 127 (solo mode) only kill voices if the respective mode really
1052 changed
1053
1054 * LSCP server:
1055 - fixed compile time error for old Bison versions
1056 (i.e. v1.28 found on MacOS 10.4, patch by Toshi Nagata)
1057 - parser now supports extended ASCII character set
1058 (up to ASCII code 255, i.e. includes now umlauts and accents)
1059 - filename arguments in LSCP commands now allow to use escape
1060 sequences, that is directly literal as one of: \', \", \\, \n, \r,
1061 \f, \t, \v, or as octal ASCII code value like \132, or as hex ASCII
1062 code value like \xf2) (fixes bug #24)
1063 - the following LSCP commands now also support escape sequences for at
1064 least one of their text-based parameters (i.e. name, description):
1065 "ADD MIDI_INSTRUMENT_MAP", "MAP MIDI_INSTRUMENT",
1066 "SET MIDI_INSTRUMENT_MAP NAME", "SET FX_SEND NAME", "CREATE FX_SEND",
1067 "SET DB_INSTRUMENT_DIRECTORY NAME",
1068 "SET DB_INSTRUMENT_DIRECTORY DESCRIPTION", "SET DB_INSTRUMENT NAME",
1069 "SET DB_INSTRUMENT DESCRIPTION", "FIND DB_INSTRUMENTS",
1070 "FIND DB_INSTRUMENT_DIRECTORIES"
1071 - returns verbose syntax errors (line and column where syntax error
1072 occured, the unexpected character and the actually expected, possible
1073 character(s), the latter only if less than 5 possibilities)
1074 - made sure that LSCP syntax is not affected by gigedit locale
1075 settings
1076 - bugfix regarding strings parameter lists: all comma separated lists
1077 of strings were treated as being one string containing commas
1078 (fixes #57)
1079
1080 * Bug fixes:
1081 - fixed crash occurring on certain LSCP scripts (Bug 39)
1082 - another thread safety fix for lscp "load engine" and "set
1083 channel audio output device"
1084 - fixed a crash which occurs when reassigning the same engine
1085 on a sampler channel with connected MIDI device
1086 - fixed a crash which occurs when changing the number of ports of a MIDI
1087 device connected to a sampler channel to number less then or equal
1088 to the index of the port to which the sampler channel is connected.
1089 - The previous bindings were not been disconnected when altering
1090 the ALSA_SEQ_BINDINGS parameter. Introduced a NONE keyword for
1091 unsubscribing from all bindings (e.g. ALSA_SEQ_BINDINGS=NONE).
1092 - The active stream/voice count statistic was incorrect.
1093 - notification events were not been sent for some sampler
1094 channel changes
1095 - added default min and max values to restrict the number of allowed
1096 audio output channels and MIDI input ports
1097 - the connection to the PCM interface is now closed when destroying
1098 an audio output device
1099 - files with slash in their path or filename could not be loaded
1100
1101 * test cases:
1102 - updated and fixed (haven't been touched in a while)
1103
1104 Version 0.4.0 (24 November 2006)
1105
1106 * packaging changes:
1107 - changed deprecated copyright attribute to license; added ldconfig
1108 to post-(un)install steps; added devel package for liblinuxsampler;
1109 to linuxsampler.spec (RPM)
1110 - install necessary development header files for allowing 3rd party
1111 applications to link against liblinuxsampler
1112 - liblinuxsampler's API documentation can be generated with 'make docs'
1113 (Doxygen required)
1114 - added benchmark to automatically detect the best triangle LFO
1115 implementation (currently either an integer math solution or a
1116 di-harmonic approximation), automatic detection can be overriden
1117 with --enable-signed-triang-algo=x and --enable-unsigned-triang-algo=x
1118 configure script argument though (mandatory for cross-compilation)
1119 - do not automatically pick optimized gcc flags if the user already
1120 provided some on his own (as CXXFLAGS)
1121 - added compile time option to disable processing of All-Notes-Off MIDI
1122 messages
1123 - added compile time options to allow disabling the various audio and
1124 MIDI drivers
1125 - fixed automatic GCC CPU switch detection on PPC
1126 (patch by Ebrahim Mayat)
1127
1128 * Gigasampler format engine:
1129 - extensive synthesis optimization
1130 (reimplementation of EGs and LFO(s), removed synthesis parameter
1131 prerendering and the synthesis parameter matrix in general, splitting
1132 each audio fragment into subfragments now where each subfragment uses
1133 constant pitch and filter coefficients. The volume coefficient is
1134 linearly interpolated inside a subfragment, unless
1135 --disable-interpolate-volume is set.)
1136 - fine tuning of the EG modulation parameters
1137 - improved filter cutoff calculation by adding support for the
1138 following gig parameters: Cutoff freq (used when no cutoff
1139 controller is defined), Control invert, Minimum cutoff, Velocity
1140 curve and Velocity range. The keyboard tracking now scales
1141 cutoff frequency instead of resonance.
1142 - added support for gig parameter Resonance.
1143 - fixed bug in sysex handling (patch by Juan Linietsky)
1144 - added global volume attenuation of -9 dB (0.35f) to prevent clipping
1145 which can be overridden with --enable-global-attenuation
1146 - EG fixes: made the length of "attack hold" stage more
1147 accurate. Release stage can now start before attack stage
1148 ends. Cancel release didn't work when sustain was zero. Attack
1149 time now has a minimal value to prevent clicks.
1150 - fixed pitch changes larger than one octave
1151 - fixed EG3 (pitch envelope) synthesis which was neutral all the time
1152 - implemented portamento mode and solo mode (a.k.a 'mono mode'):
1153 all modes can be altered via standard GM messages, that is CC5 for
1154 altering portamento time, CC65 for enabling / disabling portamento
1155 mode, CC126 for enabling solo mode and CC127 for disabling solo mode
1156 - fine tuning of the curves for volume (CC7), pan (CC10 and gig
1157 parameter) and crossfade
1158 - added support for the "attenuation controller threshold" gig
1159 parameter
1160 - added smoothing of volume changes caused by control change
1161 messages
1162 - sample loop parameters are now taken from the DimensionRegion
1163 instead of the wave chunk
1164 - fixed keyswitching for v3 gigs with a number of keyswitch splits
1165 not equal to a power of two
1166 - reimplementation of the filter algorithm. The new filter is more
1167 accurate and supports all gig filter types, including bandreject
1168 and lowpass turbo.
1169 - real support for 24 bit samples - samples are not truncated to
1170 16 bits anymore
1171 - support for aftertouch (channel pressure, not polyphonic
1172 aftertouch)
1173 - LFO1 behaviour fixed (dampening from max volume instead of
1174 amplifying from 0)
1175
1176 * LSCP server:
1177 - fixed application exit on broken pipe error (fixes bug #20)
1178 - fixed the notification messages delay due to lack of
1179 network activity (fixes bug #26)
1180 - fixed parser bug which occured on space(s) within device parameters
1181 (e.g. "SET AUDIO_OUTPUT_CHANNEL_PARAMETER 0 0 SOMEPARAM='foo bar'")
1182
1183 * audio driver:
1184 - added aRts audio output driver (by no means RT safe)
1185
1186 * MIDI driver:
1187 - fixed legacy sysex code which caused dispatching of MIDI SysEx
1188 messages several times instead of once
1189 - API extension for MIDI drivers which already supply exact time stamps
1190 for events (i.e. for offline rendering based MIDI drivers)
1191 - added checks for bad MIDI data
1192
1193 * linuxsampler application:
1194 - show available sampler engine types on startup
1195
1196 * general changes:
1197 - support for muting sampler channels and solo mode of the same, two new
1198 LSCP commands ("SET CHANNEL MUTE" and "SET CHANNEL SOLO") and two new
1199 fields ("MUTE" and "SOLO") for command "GET CHANNEL INFO" were
1200 introduced for this, the behavior is the same like on a mixer console
1201 (patch by Grigor Iliev, a bit adjusted). Also added configure option
1202 --enable-process-muted-channels which can be used to enable the
1203 processing of muted channels.
1204 - support for sostenuto pedal
1205 - support for monitoring the total number of active voices
1206 (new LSCP commands: GET TOTAL_VOICE_COUNT, GET TOTAL_VOICE_COUNT_MAX,
1207 SUBSCRIBE TOTAL_VOICE_COUNT, UNSUBSCRIBE TOTAL_VOICE_COUNT)
1208 - fixed some memory management errors
1209 - fixed some concurrency problems that could lead to crashes when
1210 LSCP commands were executed
1211 - fixed crash when instrument loading failed previously on the same
1212 sampler channel (fixes bug #36)
1213
1214 Version 0.3.3 (15 July 2005)
1215
1216 * packaging changes:
1217 - fixed compilation with gcc 4.0
1218
1219 * Gigasampler format engine:
1220 - LFO preprocessing was performed even though the respective LFO was
1221 disabled by the instrument patch (minor efficiency fix)
1222 - if period time of chosen audio device is too small (< MIN_RELEASE_TIME)
1223 for volume ramp downs in the same fragment (needed for current voice
1224 stealing implementation) then simply inform the user with a warning
1225 about possible click sounds and reduce the volume ramp down
1226 appropriately instead of cancelling the audio device connection
1227
1228 Version 0.3.2 (24 June 2005)
1229
1230 * packaging changes:
1231 - updated autotools build files to compile on OS X
1232 (conditional compilation of CoreMIDI and MidiShare drivers)
1233 - hand-crafted assembly optimization code can be disabled with
1234 './configure --disable-asm' (definitely not recommended)
1235 - fixed 'make dist' rule to include all necessary files
1236 - require automake (>= 1.5) for 'make -f Makefile.cvs'
1237 (due to 'dist-bzip2' automake option)
1238
1239 * Gigasampler format engine:
1240 - support for the gig parameters for "release velocity response" curves
1241 - fine tuning of the EGADSR envelope
1242 - volume of release triggered samples now depends on note-on velocity,
1243 note length and gig parameter "release trigger decay" instead of
1244 note-off velocity.
1245 - revised voice stealing
1246 (fixes crash and endless loop caused by voice stealing)
1247 - don't reset scale tuning on instrument or audio output device change
1248 - handle key group conflicts right at the beginning of each fragment
1249 instead of when the respective voice is actually launched
1250 (fixes undefined behavior if stolen voices belonged to a key group -
1251 this case was followed by a "killed voice survived" error message)
1252 - fixed minor issue with null/silence samples
1253 (those stole voices even though they don't need a voice at all which
1254 resulted in "voice stealing didn't work out" messages)
1255 - don't reset volume, pan, pitch and MIDI controller values on
1256 instrument or audio output device change
1257
1258 * LSCP server:
1259 - fixed some crashes (patch by Grigor Iliev, fixes #19)
1260 - fixed LSCP event "CHANNEL_INFO" notification
1261 (e.g. did not notify on volume changes or MIDI program change events)
1262
1263 * linuxsampler application:
1264 - added command line parameters --lscp-addr and --lscp-port to override
1265 default IP address and port of LSCP server
1266
1267 Version 0.3.1 (24 May 2005)
1268
1269 * initial release

  ViewVC Help
Powered by ViewVC