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

Contents of /linuxsampler/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 3100 - (show annotations) (download)
Sat Feb 4 16:53:20 2017 UTC (7 years, 1 month ago) by schoenebeck
File size: 69820 byte(s)
* SFZ: Fixed unintended volume fade-in of voices under certain
  conditions.
* Bumped version (2.0.0.svn42).

1 Version SVN trunk (?)
2
3 * SFZ format engine:
4 - added support for <global>, <master> and #define (patch by Alby M)
5 - Removed code duplication in SFZ file loading code.
6 - Added support for sfz extension opcode 'script' which may be used to
7 load real-time instrument script file (NKSP script language).
8 - Implemented opcode set_ccN (initial patch by Giovanni Senatore).
9 - Fixed unintended volume fade-in of voices under certain conditions.
10
11 * Gigasampler format engine:
12 - Fixed clicks and pumping noise with Lowpass Turbo filter on very low
13 cutoff settings.
14
15 * general changes:
16 - fixed printf type errors (mostly in debug messages)
17 - use unique_ptr instead of auto_ptr when building with C++11
18 - Added RTAVLTree class which is a real-time safe ordered multi-map, thus
19 allowing to sort data efficiently in real-time safe manner.
20 - RTList class: added methods for moving/inserting elements to arbitrary
21 position within a list.
22 - RTList class: added method fromPtr() for creating an Iterator object from
23 a raw element pointer.
24 - Attempt to partly fix resetting engine channels vs. resetting engine, an
25 overall cleanup of the Reset*(), ConnectAudioDevice(),
26 DisconnectAudioDevice() API methods would still be desirable though,
27 because the current situation is still inconsistent and error prone.
28 - Active voices are now internally grouped to "Note" objects, instead of
29 being directly assigned to a keyboard key. This allows more fine graded
30 processing of voices, which is i.e. required for certain instrument
31 script features.
32 - Fix: Release trigger voices were not spawned on sustain pedal up
33 (CC #64) events.
34 - Fix: Release trigger voices can now also distinguish correctly between
35 note-off and sustain pedal up events.
36 - All engines: Increased ramp speed of volume smoother and pan smoother
37 (while slow rate of crossfade smoother is preserved) to allow quick
38 volume and pan changes by instrument scripts for instance.
39 - gig/sf2/sfz: Fixed aftertouch channel pressure events being processed
40 properly and thus were ignored for certain uses.
41 - Added new C++ API method
42 VirtualMidiDevice::SendChannelPressureToSampler().
43 - windows, 32-bit: fixed potential crashes by making sure the stack in
44 sub threads is 16-byte aligned
45 - fixed numerous compiler warnings
46
47 * packaging changes:
48 - removed unnecessary dependency to libuuid
49 (originated by libgig's usage of it)
50 - Automake: set environment variable GCC_COLORS=auto to allow GCC to
51 auto detect whether it (sh/c)ould output its messages in color.
52
53 * Real-time instrument scripts:
54 - Implemented scheduler for delayed MIDI events and for suspended scripts.
55 - Built-in script function "wait()": implemented support for function's
56 "duration-us" argument, thus scripts using this function are now
57 correctly resumed after the requested amount of microseconds.
58 - Built-in script function "play_note()": implemented support for
59 function's "duration-us" argument, thus notes triggered with this
60 argument are now correctly released after the requested amount of
61 microseconds.
62 - Fix: script events were not cleared when engine channel was reset,
63 potentially causing undefined behavior.
64 - Fixed crash which happened when trying to reference an undeclared
65 variable.
66 - Built-in script function "play_note()": Added support for passing
67 special value -1 for "duration-us" argument, which will cause the
68 triggered note to be released once the original note was released.
69 - Instrument script classes now exported with the liblinuxsampler C++ API.
70 - Added new API method ScriptVM::syntaxHighlighting() which provides
71 a convenient syntax highlighting backend for external instrument
72 script editor applications.
73 - Added new C++ API class "ScriptVMFactory".
74 - Extended parser issues to provide not only first line and first
75 column, but also last line and last column of issue (thus marking
76 the precise span of the issue within the source code).
77 - Reload script automatically after being modified by an instrument
78 editor.
79 - NKSP language grammar correction: allow empty event handler bodies
80 like "on note end on".
81 - Implemented built-in script function "change_vol()".
82 - Implemented built-in script function "change_tune()".
83 - Implemented built-in script function "change_pan()".
84 - Implemented built-in script function "change_cutoff()".
85 - Implemented built-in script function "change_reso()".
86 - Implemented built-in script function "event_status()".
87 - Added built-in script constants "$EVENT_STATUS_INACTIVE" and
88 "$EVENT_STATUS_NOTE_QUEUE" both for being used as flags for
89 "event_status()" function.
90 - NKSP language: Added support for bitwise operators ".or.", ".and."
91 and ".not.".
92 - NKSP language scanner: Fixed IDs matching to require at least one
93 character (i.e. when matching function names or variable names).
94 - NKSP language scanner: disabled unusued rules.
95 - Fixed behavior of play_note() and note_off() functions which must
96 be distinguished engine internally from "real" MIDI note on/off
97 events in order to avoid misbehaviors like hanging notes.
98 - Implemented built-in script variable "$KSP_TIMER".
99 - Implemented built-in script variable "$NKSP_REAL_TIMER".
100 - Implemented built-in script variable "$NKSP_PERF_TIMER".
101 - Implemented built-in script variable "$ENGINE_UPTIME".
102 - Implemented built-in script function "inc()".
103 - Implemented built-in script function "dec()".
104 - NKSP language fix: division expressions were evaluated too often.
105 - NKSP language fix: string concatenation operator was right
106 associative instead of left (to right).
107 - NKSP language correction: allow empty statements in entire language
108 (i.e. "if end if").
109 - Implemented built-in script function "stop_wait()".
110 - Implemented built-in script variable "$NI_CALLBACK_ID".
111 - Implemented built-in script variable "$NI_CALLBACK_TYPE".
112 - Implemented built-in script variable "$NKSP_IGNORE_WAIT".
113 - Added support for read-only built-in variables (respectively
114 handled by the script parser).
115 - Added built-in script constant "$NI_CB_TYPE_INIT".
116 - Added built-in script constant "$NI_CB_TYPE_NOTE".
117 - Added built-in script constant "$NI_CB_TYPE_RELEASE".
118 - Added built-in script constant "$NI_CB_TYPE_CONTROLLER".
119 - NKSP Language: Added support for user defined script functions.
120 - Implemented built-in script function "change_attack()".
121 - Implemented built-in script function "change_decay()".
122 - Implemented built-in script function "change_release()".
123 - Fixed all change_*() built-in script functions to apply their
124 synthesis parameter changes immediately in case the respective note
125 was triggered at the same time, instead of scheduling the parameter
126 change, especially because it would cause some parameter types's
127 changes either to be ramped (i.e. change_vol()) or other types even
128 to have not effect at all (i.e. change_attack()).
129 - Implemented built-in script function "sh_left()".
130 - Implemented built-in script function "sh_right()".
131 - Implemented built-in script function "min()".
132 - Implemented built-in script function "max()".
133 - NKSP Fix: "init" event handler was not always executed when a script
134 was loaded.
135 - NKSP built-in wait() script function: abort script execution if
136 a negative or zero wait time was passed as argument (since this is
137 a common indication of a bug either of the script or even of the
138 engine, which could lead to RT instability or even worse).
139 - ScriptVM: Implemented automatic suspension of RT safety
140 threatening scripts.
141 - Provide more user friendly error messages on syntax errors.
142 - Fixed NKSP parser warning "Not a statement" when assigning an
143 initializer list to an array variable.
144 - Implemented built-in script array variable "%ALL_EVENTS".
145 - Implemented built-in script function "in_range()".
146
147 * Instruments DB:
148 - Fixed memory access bug of general DB access code which lead to
149 undefined behavior.
150 - Cleanup of instruments DB file creation and opening code.
151 - The instrument DB path of linuxsampler's --create-instruments-db argument
152 is now optional, if it is missing, then a default location is used.
153 - Added support for scanning SFZ (.sfz) files.
154 - Added support for scanning Sound Font (.sf2) files.
155
156 Version 2.0.0 (15 July 2015)
157
158 * packaging changes:
159 - fixed building with newer MinGW-w64
160 - Mac OS X: support the new dir for Core Audio SDK
161 - Mac OS X: fixed building outside source directory
162 - made sure all source files for hostplugins are included when
163 doing "make dist"
164 - removed empty directories left from the cvs to svn migration
165 - fixed compilation with gcc 4.6.1
166 - another "make dist" fix, for the LV2 plugin
167 - made --enable-pthread-testcancel default on Mac OS X
168 - minor Makefile fix for building lscpparser when build and source
169 directories are separate
170 - Mac OS X: made it possible to specify plugin installation dir to
171 configure
172 - Mac OS X: Makefile fix for the install-strip target
173 - fixed compilation with gcc 4.7
174 - fixed configure script error with old autoconf versions
175 - lsatomic.h: use gcc provided atomic functions if building with
176 gcc 4.7 and C++11
177 - modernized configure script
178 - fixed linkage error when building with
179 LDFLAGS="-Wl,--no-undefined" (#190)
180 - fixed compilation with Clang 3.2
181 - removed usage of deprecated Automake variable INCLUDES
182 - fixed building with C++11
183 - build fix: ChangeFlagRelaxed.h was missing in makefile
184 - build fix: libsndfile compiler flags were missing in some
185 makefiles
186 - fix for building with bison 3.0 (#202)
187 - Mac OS X: added temporary hack allowing to spawn gigedit as callback
188 on the process's main thread
189 - fixed build error on newer MinGW
190 - support building with older jack versions
191 - support building with spaces in vst sdk path
192 - enabled automake 'subdir-objects' option and moved external
193 source references (vst, au, asio) from makefiles to cpp files,
194 in order to get rid of warnings from automake 1.14
195
196 * general changes:
197 - Refactoring: moved the independent code from
198 the Gigasampler format engine to base classes
199 - command line option '--profile' is currently disabled, since the
200 respective profiling code is currently broken
201 - Introduced support for optional environment variable
202 "LINUXSAMPLER_PLUGIN_DIR" which allows to override the directory
203 where the sampler shall look for instrument editor plugins
204 (patch by Luis Garrido, slightly modified).
205 - implemented Roland GS NRPN 1ArrH which allows to set volume per note
206 - implemented Roland GS NRPN 1CrrH which allows to set pan per note
207 - implemented Roland GS NRPN 1DrrH which allows to set reverb send per
208 note (in this implementation of the sampler its simply hard routed to
209 the 1st effect send of the sampler channel, no matter what the actual
210 effect type is)
211 - implemented Roland GS NRPN 1ErrH which allows to set chorus send per
212 note (in this implementation of the sampler its simply hard routed to
213 the 2nd effect send of the sampler channel, no matter what the actual
214 effect type is)
215 - implemented support for internal LADSPA effects
216 - raised limit of program change queue from 100 to 512
217 (as suggested by Alex Stone)
218 - implemented sine LFO, pulse LFO and saw LFO
219 - experimental support for per voice equalization
220 - added command line option --exec-after-init
221 - Introduced new C++ API method:
222 EngineChannel::InstrumentFileName(int index)
223 allowing to retrieve the whole list of files used for the loaded
224 instrument on an engine channel (a.k.a. part). Some GigaStudio
225 instruments for example are splitted over several files like
226 "Foo.gig", "Foo.gx01", "Foo.gx02", ...
227 - Added new C++ API method Sampler::GetGlobalMaxVoices().
228 - Added new C++ API method Sampler::GetGlobalMaxStreams().
229 - Added new C++ API method Sampler::SetGlobalMaxVoices().
230 - Added new C++ API method Sampler::SetGlobalMaxStreams().
231 - Various "const" and "restrict" optimizations.
232 - all engines: add pan CC value to instrument pan parameter before
233 applying panning, instead of using two separate pan functions in
234 series (#182)
235 - added a lock guard class for exception safe mutex handling and
236 used it everywhere appropriate
237 - Immediately apply scale tuning changes to active voices.
238 - Exposed scale tuning to C++ API (along to the already existing standard
239 SysEx way).
240 - lsatomic.h fixes: seq_cst load and store were suboptimal for x86
241 and broken for ppc64. (Seq_cst loads and stores are actually not
242 used in LS, so the bug wasn't noticable.)
243 - lsatomic.h: added ARMv7 support
244 - Added support for multiple MIDI input ports per sampler channel (and
245 added various new C++ methods for this new feature / design change, old
246 C++ API methods for managing SamplerChannel's MIDI inputs are now marked
247 as deprecated but are still there and should provide full behavior
248 backward compatibility).
249 - AbstractEngine::GSChecksum(): don't allocate memory on the stack (was
250 unsafe and caused compilation error with clang 2.x).
251 - Bugfix: only process the latest MIDI program change event.
252 - Introducing the LSCP shell, which provides convenient control of the
253 sampler from the command line by providing LSCP aware features.
254 - VirtualMidiDevice: Added support for program change.
255 - VirtualMidiDevice: Added support for bank select (MSB & LSB).
256 - VirtualMidiDevice: Added support for pitch bend.
257 - Aftertouch: extended API to explicitly handle channel pressure and
258 polyphonic key pressure events (so far polyphonic pressure was not
259 supported at all, and channel pressure was rerouted as CC128 but not
260 used so far).
261 - Added initial support for real-time instrument scripts. The script VM
262 code is shared by all sampler engine implemementations, however only the
263 gig file format currently provides support for storing instrument scripts
264 (as LinuxSampler extension to the original GigaStudio 4 format).
265
266 * Gigasampler format engine:
267 - implemented the "round robin keyboard" dimension
268 - fixed round robin and random dimensions for cases when number of
269 dimension zones is not a power of two
270 - made round robin use a counter for each region instead of each
271 key
272 - bugfix: pitch LFO controller "internal+aftertouch" was broken
273 - bugfix: filter keyboard tracking was broken
274 - filter performance fix (an unnecessary copy was made of the
275 filter parameters in each sub fragment)
276 - handle special case when pan parameter in gig file has max or
277 min value
278 - Exclusive Groups: don't ever stop voices of the same note,
279 doesn't sound naturally with a drumkit
280 - fixed EG1 modulation when attack or release is zero
281 - Fixed support for 'aftertouch' attenuation controller.
282 - Fixed crash that happened with velocity split sounds under certain
283 conditions (see also previous commit on libgig).
284 - fixed behaviour of filter LFO
285
286 * SFZ format engine:
287 - Initial implementation (not usable yet)
288 - added support for v2 multiple stage envelope generators
289 - added a fine-tuned v1 envelope generator instead of using the
290 one from the gig engine
291 - fixed memory leak and memory handling errors
292 - added support for trigger=first, trigger=legato and sw_previous
293 - allow non-numerical key values ("C#4" for example)
294 - "key" opcode now sets pitch_keycenter too
295 - fixed error when unloading instrument with same sample used by
296 multiple regions
297 - added some opcode aliases, like loopmode for loop_mode, to be
298 more compatible
299 - added support for trigger=release and rt_decay
300 - added support for off_mode=normal
301 - added support for random, seq_position, seq_length and volume
302 - added v1 LFO opcodes to parser (no support in engine yet)
303 - added support for amp_veltrack and amp_velcurve_N
304 - fine-tuned the default velocity curve
305 - added support for transpose
306 - fixed crash when using sw_down/up
307 - improved logic for sw_lokey/hikey/up/down/last
308 - added more v1 aliases to parser
309 - reduced memory usage for sfz data
310 - RT-safeness: avoid malloc in audio thread
311 - fixed a bug that could cause voice stealing to fail
312 - optimized sample lookup
313 - improved support for exclusive groups (group, off_by and
314 off_mode)
315 - added support for controller triggered regions (on_locc/on_hicc)
316 - added support for loop_mode=one_shot
317 - fixed playback of 16 bit wav files on big endian CPUs
318 - added support for Ogg Vorbis sample files
319 - use loop markers from sample file if loop_start and loop_end are
320 not set in sfz file
321 - implemeted filters. Filter types: lowpass, bandpass, bandreject
322 and highpass. 1, 2, 4 and 6 pole filters. Opcodes: fil_type,
323 cutoff, resonance, fil_veltrack, fil_keytrack, fil_keycenter,
324 cutoff_cc, cutoff_chanaft.
325 - bugfix: zero ampeg_sustain didn't work
326 - added support for velocity effect on amplifier envelope time
327 (ampeg_vel2attack, ampeg_vel2decay, ampeg_vel2sustain and
328 ampeg_vel2release)
329 - added support for EG hold (ampeg_hold)
330 - added support for sample offset (offset)
331 - use the newly introduced signal units model
332 - implemented opcodes lfoN_delay,
333 lfoN_freq, lfoN_pan, lfoN_cutoff, lfoN_resonance
334 - implemented opcodes ampeg_delay, ampeg_vel2delay
335 pitcheg_delay, pitcheg_start, pitcheg_attack, pitcheg_hold,
336 pitcheg_decay, pitcheg_sustain, pitcheg_release, pitcheg_vel2delay,
337 pitcheg_vel2attack, pitcheg_vel2hold, pitcheg_vel2decay,
338 pitcheg_vel2sustain, pitcheg_vel2release, pitcheg_depth
339 - implemented opcodes pitchlfo_delay, pitchlfo_freq,
340 pitchlfo_depth, fillfo_delay, fillfo_freq, fillfo_depth,
341 amplfo_delay, amplfo_freq, amplfo_depth
342 - implemented opcodes fileg_delay, fileg_start, fileg_attack,
343 fileg_hold, fileg_decay, fileg_sustain, fileg_release, fileg_vel2delay,
344 fileg_vel2attack, fileg_vel2hold, fileg_vel2decay, fileg_vel2sustain,
345 fileg_vel2release
346 - implemented opcode lfoN_wave
347 - implemented opcode pitchlfo_depthccN
348 - implemented opcodes lfoN_phase, lfoN_phase_onccX,
349 lfoN_pitch, lfoN_pitch_onccX
350 - implemented opcodes pitchlfo_fade,
351 fillfo_fade, amplfo_fade, lfoN_fade, lfoN_fade_onccX
352 - implemented opcodes fillfo_freqccN,
353 pitchlfo_freqccN, amplfo_freqccN, lfoN_freq_onccX
354 - implemented opcodes ampeg_delayccN, ampeg_startccN,
355 ampeg_attackccN, ampeg_holdccN, ampeg_decayccN, ampeg_sustainccN,
356 ampeg_releaseccN, egN_timeX_onccY, egN_levelX_onccY
357 - lfoN_* and egN_* opcodes defined in group sections
358 are now taken into account
359 - implemented curves
360 - implemented opcodes volume_onccN, volume_curveccN
361 - implemented opcode volume_smoothccN
362 - implemented opcodes fillfo_depthccN, amplfo_depthccN,
363 lfoN_volume, lfoN_volume_onccX, lfoN_volume_smoothccX,
364 lfoN_freq_smoothccX, lfoN_pitch_smoothccX, lfoN_pan_onccX,
365 lfoN_pan_smoothccX, lfoN_cutoff_onccX, lfoN_cutoff_smoothccX,
366 lfoN_resonance_onccX, lfoN_resonance_smoothccX, lfoN_delay_onccX
367 - implemented opcode end
368 - implemented opcodes egN_amplitude_onccX,
369 egN_volume, egN_volume_onccX, egN_cutoff, egN_cutoff_onccX,
370 egN_pitch, egN_pitch_onccX, egN_resonance, egN_resonance_onccX
371 - implemented opcodes xfin_lokey, xfin_hikey,
372 xfout_lokey, xfout_hikey, xf_keycurve, xfin_lovel, xfin_hivel,
373 xfout_lovel, xfout_hivel, xf_velcurve, xfin_loccN, xfin_hiccN,
374 xfout_loccN, xfout_hiccN, xf_cccurve
375 - implemented opcodes pan_onccN, pan_smoothccN,
376 pan_curveccN, egN_pan, egN_pan_curve, egN_pan_onccX,
377 egN_pan_curveccX, pitch_veltrack
378 - when failed to parse a sfz file
379 print the line number on which the error occurs
380 - use common pool of CC objects to minimize RAM usage
381 - implemented opcodes amplfo_delay_onccN,
382 amplfo_fade_onccN, fillfo_delay_onccN, fillfo_fade_onccN,
383 pitchlfo_delay_onccN, pitchlfo_fade_onccN
384 - implemented opcodes fileg_delay_onccN,
385 fileg_start_onccN, fileg_attack_onccN, fileg_hold_onccN,
386 fileg_decay_onccN, fileg_sustain_onccN, fileg_release_onccN,
387 fileg_depth_onccN, pitcheg_delay_onccN, pitcheg_start_onccN,
388 pitcheg_attack_onccN, pitcheg_hold_onccN, pitcheg_decay_onccN,
389 pitcheg_sustain_onccN, pitcheg_release_onccN, pitcheg_depth_onccN
390 - implemented automatic aliasing *ccN <-> *_onccN
391 - *lfo_freqccN wasn't working when the respective *lfo_freq
392 was not set or was set to zero
393 - lfoN_freq_onccX wasn't working when lfoN_freq
394 was not set or was set to zero
395 - implemented opcodes resonance_onccN, resonance_smoothccN,
396 resonance_curveccN, cutoff_smoothccN, cutoff_curveccN
397 - implemented opcodes fillfo_depthchanaft,
398 fillfo_freqchanaft, amplfo_depthchanaft, amplfo_freqchanaft,
399 pitchlfo_depthchanaft, pitchlfo_freqchanaft
400 - implemented opcodes pitch_onccN,
401 pitch_curveccN, pitch_smoothccN, pitch_stepccN
402 - implemented opcodes volume_stepccN, pan_stepccN,
403 cutoff_stepccN, resonance_stepccN, lfoN_freq_stepccX,
404 lfoN_volume_stepccX, lfoN_pitch_stepccX, lfoN_pan_stepccX,
405 lfoN_cutoff_stepccX, lfoN_resonance_stepccX
406 - implemented opcodes eq1_freq, eq2_freq, eq3_freq,
407 eq1_freqccN, eq2_freqccN, eq3_freqccN, eq1_bw, eq2_bw, eq3_bw,
408 eq1_bwccN, eq2_bwccN, eq3_bwccN, eq1_gain, eq2_gain, eq3_gain,
409 eq1_gainccN, eq2_gainccN, eq3_gainccN
410 - implemented opcodes delay, delay_onccN, delay_random,
411 delay_samples, delay_samples_onccN
412 - implemented opcodes egN_eq1gain, egN_eq2gain, egN_eq3gain,
413 egN_eq1gain_onccX, egN_eq2gain_onccX, egN_eq3gain_onccX, egN_eq1freq,
414 egN_eq2freq, egN_eq3freq, egN_eq1freq_onccX, egN_eq2freq_onccX,
415 egN_eq3freq_onccX, egN_eq1bw, egN_eq2bw, egN_eq3bw, egN_eq1bw_onccX,
416 egN_eq2bw_onccX, egN_eq3bw_onccX, lfoN_eq1gain, lfoN_eq2gain,
417 lfoN_eq3gain, lfoN_eq1gain_onccX, lfoN_eq2gain_onccX, lfoN_eq3gain_onccX,
418 lfoN_eq1gain_smoothccX, lfoN_eq2gain_smoothccX, lfoN_eq3gain_smoothccX,
419 lfoN_eq1gain_stepccX, lfoN_eq2gain_stepccX, lfoN_eq3gain_stepccX,
420 lfoN_eq1freq, lfoN_eq2freq, lfoN_eq3freq, lfoN_eq1freq_onccX,
421 lfoN_eq2freq_onccX, lfoN_eq3freq_onccX, lfoN_eq1freq_smoothccX,
422 lfoN_eq2freq_smoothccX, lfoN_eq3freq_smoothccX, lfoN_eq1freq_stepccX,
423 lfoN_eq2freq_stepccX, lfoN_eq3freq_stepccX, lfoN_eq1bw, lfoN_eq2bw,
424 lfoN_eq3bw, lfoN_eq1bw_onccX, lfoN_eq2bw_onccX, lfoN_eq3bw_onccX,
425 lfoN_eq1bw_smoothccX, lfoN_eq2bw_smoothccX, lfoN_eq3bw_smoothccX,
426 lfoN_eq1bw_stepccX, lfoN_eq2bw_stepccX, lfoN_eq3bw_stepccX
427 - implemented opcodes eq1_vel2freq, eq2_vel2freq,
428 eq3_vel2freq, eq1_vel2gain, eq2_vel2gain, eq3_vel2gain
429 - sfz parser: allow double spaces in sample filenames
430 - sfz parser: allow absolute paths for sample filenames
431 - use linear decay and release for filter and pitch EG
432 - bugfix: only the first amp_veltrack definition in a file was
433 used
434 - bugfix: looping was disabled if loop_start was set to 0
435 - allow regions with end=-1 to turn off other regions using the
436 group and off_by opcodes (#168)
437 - made end=0 play the whole sample
438 - fixed support for lochan and hichan opcodes (#155)
439 - fixed crash when using lochan/hichan opcodes (#187)
440 - sfz parser: allow -200 to 200 for pan_oncc opcode (#182)
441 - added FLAC support (#191)
442 - sfz parser bugfix: lines starting with whitespace were ignored
443 - added amplitude opcode
444 - added support for "#include" instruction
445 (modified patch which was originally posted by Sergey on LS mailing list)
446 - bugfix: generation of velocity curves etc should not be done
447 after each "#include", only after the main file is parsed
448 - bugfix: line numbers in error messages were wrong after "#include"
449 - added support for float and 32 bit sample files
450
451 * SoundFont format engine:
452 - Initial implementation (not usable yet)
453 - RT-safeness: avoid malloc in audio thread
454 - fixed a bug that could cause voice stealing to fail
455 - fine-tuned amplitude EG (by switching from gig to sfz EG)
456 - initial implementation of Vibrato LFO and Modulation LFO
457 - initial implementation of cutoff filter
458 - use linear decay and release for filter and pitch EG
459
460 * Host plugins (VST, AU, LV2, DSSI):
461 - AU bugfix: failed to destroy its audio/MIDI devices
462 - Listen to all interfaces on Mac OS X (INADDR_ANY)
463 - VST bugfix: If the host called resume() before and after
464 changing sample rate or block size, the number of channels was
465 incorrectly set to two. This caused silence in Cubase 5.
466 - save engine type (gig, sfz or sf2) in plugin state
467 - VST: when opening Fantasia, look for both 32 and 64 bit Java on
468 64 bit Windows
469 - AU: changed number of output channels from one stereo to 16
470 stereo
471 - VST: made it possible to build the VST plugin for Mac
472 - AU: link AU plugin dynamically if --disable-shared isn't
473 specified
474 - LV2 "state" extension support (patch by David Robillard)
475 - VST bugfix: instrument loading hang and crashed the host when
476 the plugin was loaded a second time (#174)
477 - plugin bugfix: instrument loading hang when the plugin was
478 loaded a second time (this time it's for Linux and Mac, previous
479 similar fix was for Windows)
480 - thread safety fixes for the instrument loading thread
481 - LV2: use the new lv2 package if present
482 - VST: try to open Fantasia automatically on Linux and Mac too (on
483 Linux, the Fantasia jar should be placed in <prefix>/share/java)
484 - VST: fixed crashes on Linux Ardour and EnergyXT
485 - DSSI bugfix: it wasn't possible to change engine type. The MIDI
486 port and audio channel routing for DSSI plugins are now visible.
487 - LV2: use urid and atom extensions instead of deprecated uri-map
488 and event
489 - LV2: lv2 package 1.0 is now required to build the LV2 plugin
490 - LV2: changed number of output channels to 16 stereo, just like
491 the VST and AU plugins
492 - LV2: fixed save/restore of SFZ state (patch by David Robillard)
493 - LV2: made LV2 plugin buildable on Windows and Mac
494 - VST: implemented retrieval and switching of programs using the
495 sampler's internal MIDI instrument mapping system
496
497 * Instrument editor interface:
498 - Changed instrument editor plugin interface, providing additional
499 informations like the EngineChannel for which the instrument editor was
500 spawned for. This allows the instrument editors to interact more actively
501 with the sampler.
502
503 * MIDI driver:
504 - ALSA MIDI driver supports now "NAME" device parameter, for overriding
505 the ALSA sequencer client name
506 - removed limit of maximum amount of MIDI ports per MIDI device, since
507 there is no reason for this limit
508 - MME: fixed memory handling bug found with cppcheck
509 - MME: removed compiler warning
510 - CoreMIDI: implemented driver specific port parameter "CORE_MIDI_BINDINGS",
511 which allows to retrieve the list of CoreMIDI clients / ports and to connect
512 to them a la JACK, via the usual sampler APIs
513 - CoreMIDI: added driver specific port parameter "AUTO_BIND", if enabled
514 the driver will automatically connect to other CoreMIDI clients' ports
515 (e.g. external MIDI devices being attached to the Mac)
516 - added support for MIDI note on velocity filter
517 - CoreMIDI: fixed memory deallocation error
518 - Fixed variable underflow in VirtualMidiDevice, which caused graphical
519 virtual keyboards in frontends / instrument editors being stuck.
520 - Bugfix in VirtualMidiDevice: process note on with velocity 0 as note off.
521 - Implemented missing handling of MIDI "running status".
522 - CoreMIDI fix: a MIDIPacket can contain more than one event per packet.
523 - MME bugfix: driver wasn't closed properly
524 - CoreMIDI: automatically connect to all input sources by default (driver
525 parameter "AUTO_BIND").
526 - CoreMIDI: fixed auto bind feature to CoreMIDI ports that go online
527 - CoreMIDI: fixed minor memory leak
528
529 * audio driver:
530 - ASIO driver fixes for newer gcc versions (fix from PortAudio)
531 - JACK audio: react on sample rate changes.
532 - JACK audio: react on buffer size changes.
533 - JACK audio: jack_port_get_buffer() was cached and called outside
534 RT context.
535 - ASIO driver: removed compiler warnings
536 - CoreAudio: fixed minor error handling bug
537 - ASIO driver: be more verbose when no ASIO card could be found (fixes #203)
538 - JACK audio: return the JACK server's current sample rate as default value
539 for audio device parameter "SAMPLERATE" (fixes #166).
540
541 * LSCP server:
542 - added support for sending MIDI CC messages via LSCP command
543 "SEND CHANNEL MIDI_DATA CC <sampler-chan> <ctrl> <val>"
544 - added LSCP command "GET AVAILABLE_EFFECTS"
545 - added LSCP command "LIST AVAILABLE_EFFECTS"
546 - added LSCP command "GET EFFECT INFO <effect-index>"
547 - added LSCP command "CREATE EFFECT_INSTANCE <effect-index>"
548 - added LSCP command
549 "CREATE EFFECT_INSTANCE <effect-system> <module> <effect-name>"
550 - added LSCP command "DESTROY EFFECT_INSTANCE <effect-instance>"
551 - added LSCP command "GET EFFECT_INSTANCES"
552 - added LSCP command "LIST EFFECT_INSTANCES"
553 - added LSCP command "GET EFFECT_INSTANCE INFO <effect-instance>"
554 - added LSCP command
555 "GET EFFECT_INSTANCE_INPUT_CONTROL INFO <effect-instance> <input-control>"
556 - added LSCP command "SET EFFECT_INSTANCE_INPUT_CONTROL VALUE
557 <effect-instance> <input-control> <value>"
558 - added LSCP command "GET SEND_EFFECT_CHAINS <audio-device>"
559 - added LSCP command "LIST SEND_EFFECT_CHAINS <audio-device>"
560 - added LSCP command "ADD SEND_EFFECT_CHAIN <audio-device>"
561 - added LSCP command
562 "REMOVE SEND_EFFECT_CHAIN <audio-device> <effect-chain>"
563 - added LSCP command
564 "GET SEND_EFFECT_CHAIN INFO <audio-device> <effect-chain>"
565 - added LSCP command "APPEND SEND_EFFECT_CHAIN EFFECT <audio-device>
566 <effect-chain> <effect-instance>"
567 - added LSCP command "INSERT SEND_EFFECT_CHAIN EFFECT <audio-device>
568 <effect-chain> <effect-chain-pos> <effect-instance>"
569 - added LSCP command "REMOVE SEND_EFFECT_CHAIN EFFECT <audio-device>
570 <effect-chain> <chain-pos>"
571 - added LSCP command "SET FX_SEND EFFECT <sampler_channel>
572 <fx_send_id> <effect_chain> <chain_pos>"
573 - added LSCP command "REMOVE FX_SEND EFFECT <sampler_channel> <fx_send_id>"
574 - added LSCP commands "SUBSCRIBE EFFECT_INSTANCE_COUNT",
575 "SUBSCRIBE EFFECT_INSTANCE_INFO", "SUBSCRIBE SEND_EFFECT_CHAIN_COUNT",
576 "SUBSCRIBE SEND_EFFECT_CHAIN_INFO"
577 - provide comprehensive error messages on LSCP syntax errors
578 (suggesting expected next non-terminal symbols)
579 - Fixed client connection not being closed after network errors.
580
581 * LSCP shell:
582 - Added support for auto correction of obvious and trivial syntax mistakes.
583 - Added support for auto completion by tab key.
584 - Show currently available auto completion while typing.
585 - Added support for browsing command history with up / down keys.
586 - Show all possible next symbols immediately right to the current command
587 line while typing (no double tab required for this feature, as it would
588 be the case in other shells).
589 - Added support for moving cursor left/right with arrow keys.
590 - Added support for built-in LSCP reference documentation, which will
591 automatically show the relevant LSCP reference section on screen as soon
592 as one specific LSCP command was detected while typing on the command
593 line.
594
595 * Real-time instrument scripts:
596 - Implemented built-in script array variable %CC.
597 - Implemented built-in script int variable $CC_NUM.
598 - Implemented built-in script int variable $EVENT_NOTE.
599 - Implemented built-in script int variable $EVENT_VELOCITY.
600 - Implemented built-in script constant variable $VCC_MONO_AT.
601 - Implemented built-in script constant variable $VCC_PITCH_BEND.
602 - Implemented execution of script event handler "init".
603 - Implemented execution of script event handler "controller".
604 - Implemented execution of script event handler "note".
605 - Implemented execution of script event handler "release".
606 - Implemented built-in script function "play_note()" (only two of the
607 max. four function arguments are currently implemented yet though).
608 - Implemented built-in script int variable $EVENT_ID.
609 - Implemented built-in script function "ignore_event()"
610 - Implemented built-in script function "ignore_controller()" (may have one
611 or no argument).
612 - Implemented built-in script function "set_controller()".
613 - Added extended script VM for the Gigasampler/GigaStudio format sampler
614 engine, which extends the general instrument script VM with Giga format
615 specific variables and functions.
616 - Giga format scripts: added built-in script int constant variables
617 $GIG_DIM_CHANNEL, $GIG_DIM_LAYER, $GIG_DIM_VELOCITY, $GIG_DIM_AFTERTOUCH,
618 $GIG_DIM_RELEASE, $GIG_DIM_KEYBOARD, $GIG_DIM_ROUNDROBIN, $GIG_DIM_RANDOM,
619 $GIG_DIM_SMARTMIDI, $GIG_DIM_ROUNDROBINKEY, $GIG_DIM_MODWHEEL,
620 $GIG_DIM_BREATH, $GIG_DIM_FOOT, $GIG_DIM_PORTAMENTOTIME, $GIG_DIM_EFFECT1,
621 $GIG_DIM_EFFECT2, $GIG_DIM_GENPURPOSE1, $GIG_DIM_GENPURPOSE2,
622 $GIG_DIM_GENPURPOSE3, $GIG_DIM_GENPURPOSE4, $GIG_DIM_SUSTAIN,
623 $GIG_DIM_PORTAMENTO, $GIG_DIM_SOSTENUTO, $GIG_DIM_SOFT,
624 $GIG_DIM_GENPURPOSE5, $GIG_DIM_GENPURPOSE6, $GIG_DIM_GENPURPOSE7,
625 $GIG_DIM_GENPURPOSE8, $GIG_DIM_EFFECT1DEPTH, $GIG_DIM_EFFECT2DEPTH,
626 $GIG_DIM_EFFECT3DEPTH, $GIG_DIM_EFFECT4DEPTH, $GIG_DIM_EFFECT5DEPTH.
627 - Giga format scripts: Implemented built-in script function
628 "gig_set_dim_zone(event_id, dimension, zone)".
629 - Implemented built-in script int array variable $KEY_DOWN.
630 - Implemented built-in script function "abs()".
631 - Implemented built-in script function "random()".
632 - Implemented built-in script function "num_elements()".
633 - Implemented built-in script function "note_off()".
634 - Implemented built-in script function "set_event_mark()".
635 - Implemented built-in script function "delete_event_mark()".
636 - Implemented built-in script function "by_marks()".
637 - Added built-in script int const variables $MARK_1 to $MARK_28.
638 - Built-in script functions "ignore_event()", "note_off()" and
639 "gig_set_dim_zone()" now also accept an array of event IDs as argument
640 (i.e. return value of new script function "by_marks()").
641 - Pass/preserve polyphonic variable data from respective "note" event
642 handler to "release" event handler.
643 - Fixed crash when using built-in script function "by_marks()".
644
645 * Bug fixes:
646 - Fixed crash which may occur when MIDI key + transpose is out of range
647 - minor valgrind fixes
648 - fixed crash which occurred when changing an already deployed sampler
649 channel to a different engine type
650 - fixed crash when deleting a sampler channel or changing engine
651 type while an instrument load was in progress
652 - bugfix: playing a note while changing the instrument could cause
653 a crash, or give "Handing back unknown region" error messages
654 - bugfix: calling SET VOICES (which Fantasia does on start and
655 refresh) could cause instruments to be unloaded
656 - Mac OS X: fixed name collision of enum in EffectControl and
657 wrong name of destructor in AudioOutputDeviceCoreAudio.cpp
658 - Mac OS X: fixed hanging threads
659 - Windows: disabled the previous thread fix on non-Mac systems, as
660 it caused hanging threads on Windows
661 - Fixed possible crashes due to corrupted MIDI/audio device list
662 after MIDI/audio device creation failure
663 - When creating MIDI instrument map entries with "PERSISTENT" type, the
664 instruments were uselessly precached with zero samples, however it still
665 took the full preloading time and on 1st program change the respective
666 instrument was completely reloaded again.
667 - fixed handling of rapid bank select and program change messages
668 sent to the same sampler channel (patch from the Open Octave
669 project, slightly adjusted)
670 - fixed crash when trying to create an effect instance with controls
671 which min and/or max values depend on the sample rate
672 - fixed bug #162
673 - bugfix: LADSPA_PATH was not evaluated correctly when containing
674 multiple paths (#165)
675 - thread safety fixes for the instrument loading thread
676 - bugfix: instrument loading crashed for sfz and sf2 in Ardour
677 (#176)
678 - more thread safety fixes for the instrument loading thread
679 - sfz/sf2 engine: fixed crash when using small audio fragment size
680 - Mac OS X: fixed crash when unloading plugin on 10.7 and later
681 - Mac OS X: fixed process hang when unloading 32-bit plugin (bug
682 introduced in previous fix)
683 - fixed crash when a channel received a program change while
684 playing a note in a key group
685 - fixed erroneous error message piping in VoiceBase.h
686 (labelled "Disk stream not available in time")
687 - Update effects on sample rate & period size changes (to avoid
688 crashes and noise on such transitions).
689 - ignore missing LADSPA paths without ignoring valid LADSPA paths
690 (fixes #208)
691 - Fixed bug in LSCP grammar definition which caused a statement like
692 "GET SERVER INFOasdf\n" to be accepted as valid statement (was so far
693 practically irrelevant, however it caused problems with the new LSCP
694 shell's auto completion feature).
695 - Fixed MIDI program change messages being ignored if quickly executed
696 after each other (fixes #231).
697
698 Version 1.0.0 (31 July 2009)
699
700 * packaging changes:
701 - autoconf bugfix: the PKG_CONFIG variable wasn't initialized properly,
702 causing e.g. the libgig test to fail when
703 "./configure --disable-jack-driver" was used
704 (patch by Alexis Ballier)
705 - fixed compilation with gcc 4.3
706 - fixes for building on OS X (thanks to Ebrahim Mayat for testing)
707 - fixed configure so it detects x86_64 (#107)
708 - fixes for building with newer MinGW versions
709 - fix for building with bison 2.4 (#111)
710 - fixed building with libgig installed in a non-standard directory
711 - minor fix in configure for mmsystem.h detection on MinGW
712 - Windows: look for editor plugins and Fantasia using base
713 directory of liblinuxsampler dll (look in the same directory and one
714 directory above)
715 - configure script fix: removed unconditional use of SSE
716 - fixed building with sqlite installed in a non-standard directory
717 - when cross-compiling, don't try to create instruments.db
718 - fix for new mingw-w64 version, which has usleep
719
720 * general changes:
721 - bugfix: on some POSIX systems instrument editor plugins refused to
722 load as we used a non-portable Linux specific struct field
723 (fixes bug #70, patch by Ronald Baljeu)
724 - fixed endless loop which occured when loading an instrument editor
725 plugin DLL on Windows systems failed
726 - fixed memory leaks that occurred when liblinuxsampler was unloaded
727 - optimized the SynchronizedConfig class so it doesn't wait
728 unnecessarily long after an update
729 - added support for notifying instrument editors on note-on / note-off
730 events (e.g. to highlight the pressed keys on the virtual keyboard
731 of gigedit)
732 - added support for triggering notes by instrument editors (see above)
733 - be verbose on DLL load errors (on Linux)
734 - fixed di-harmonic triangle LFO implementation (this LFO implementation
735 is only used by very few systems, most pick the int math
736 implementation, which is usually faster)
737 - fixes for audio drivers with varying buffer sizes
738 - experimental support for running LinuxSampler as a DSSI, LV2 and
739 VST plugin
740 - notification events for stream/voice count statistics are now sent
741 only when there are actual changes
742 - added memory ordering constraints to improve stability on
743 multi-core and multi-cpu systems
744 - maximum voices and disk streams can now be altered at runtime
745 - fixed CPU feature detection on x86_64 (maybe fixes #108)
746 - automatic stacktrace mechanism is now turned off by default and can
747 be switched on by command line option "--stacktrace" (the automatic
748 stacktrace mechanism seems to be broken on most systems at the moment)
749 - C++ API method InstrumentManager::LaunchInstrumentEditor() now returns
750 a pointer to the launched InstrumentEditor object
751 - added optional 3rd party user data parameter for following
752 liblinuxsampler C++ API methods: InstrumentEditor::Main(),
753 InstrumentEditor::Launch(),
754 InstrumentManager::LaunchInstrumentEditor()
755 - theoretical fix: made SynchronizedConfig follow C++0x memory
756 model more strictly
757 - fixes for using large audio device buffers
758 - Windows: add the installation directory to the DLL search path
759 when loading an editor plugin (solves problems with VST and
760 gigedit on systems with other GTK versions installed)
761 - updated linuxsampler man page
762
763 * audio driver:
764 - removed the nonsense audio channel constraint (which was hard coded to
765 max. 100 audio channels) for most audio drivers
766 - JACK audio driver did not offer a device parameter "SAMPLERATE" as
767 opposed to the LSCP specs
768 - bugfix: the SAMPLERATE parameter of some drivers (e.g. JACK)
769 reflected the wrong value
770 - fixed a memory management bug in ASIO driver
771 - Makefile fix: JACK_CFLAGS wasn't used
772 - JACK: use jack_client_open instead of the deprecated
773 jack_client_new
774 - added (experimental) CoreAudio driver
775 - applied old fixes to the ASIO driver that were included in the
776 previous binary release but accidentally never committed to CVS
777 (fixes #117)
778 - fixes for ASIO on mingw-w64 (iasio wrapper is not needed on
779 win64)
780 - VST: added support for sample rate and buffer size changes
781 - VST: close editor (Fantasia) when the VST is removed
782 - VST: avoid opening Fantasia more than once for each VST instance
783 - VST: export main function as "main" on Linux too (fix for
784 energyXT)
785 - VST: changed number of output channels from one stereo to 16
786 stereo
787 - added channel routing, fxsends and midi maps to the settings
788 stored in the plugin state
789 - performance optimization of AudioChannel::MixTo() and
790 AudioChannel::CopyTo() methods using GCC vector exensions
791 (if available)
792 - ASIO fixes: avoid initializing the device twice, avoid throwing
793 exception when getting parameters from a disconnected device
794
795 * MIDI driver:
796 - added JACK MIDI driver
797 - dispatch bank select as ordinary CC as well, the user might seriously
798 want to (mis)use it for some purpose ("fixed" in all current MIDI
799 input drivers: ALSA, CoreMIDI, JACK, MidiShare, MME)
800 - bugfix: pitch bend wasn't working with JackMidi, VST, LV2, MME,
801 CoreMidi or AU
802 - fixed mingw-w64 compilation error in MME driver
803 - made program change handling in MIDI thread real-time safe by
804 moving the logic to a non-RT thread
805 - fixed minor memory leak in ALSA driver
806
807 * instruments database:
808 - avoid time consuming samples scanning when adding instruments
809 to the instruments database
810 - added support for handling lost files in the instruments database
811 - Implemented option for adding instruments in separate directories
812 in the instruments database
813 (patch by Chris Cherrett & Andrew Williams, a bit adjusted)
814 - work-around for missing fnmatch function on Windows to make
815 instrument database compilable
816 - added instrument database support on Windows
817 works with both standalone mode and VST plugin.
818 the instruments DB file is located in
819 %USERPROFILE%\.linuxsampler\instruments.db which allows different
820 databases for each windows user
821 if no DB is present it automatically creates the .linuxsampler subdir
822 and then creates an empty DB
823 - fixed recursive import, which was broken on Windows
824
825 * Gigasampler format engine:
826 - fixed a memory leak that could happen when a channel was deleted
827 while notes were playing
828 - made it possible to load an instrument even if the audio thread
829 isn't running
830 - added partial support for the "Controller Triggered" MIDI rule,
831 enough for piano gigs with pedal noise samples
832 - minor fix: only mark FX sends as being modified if really the
833 respective FX send MIDI controller was used
834 - added support for GM global device volume SysEx message
835 (can be selected at compile time to either apply globally to the
836 whole sampler [default] or only to the sampler channels that are
837 connected to the respective MIDI input port the SysEx message arrived
838 on)
839 - bugfix: notes triggered at position 0 in the audio buffer were
840 sometimes wrongly killed in the same buffer, causing no sound to
841 be played
842 - added support for chromatic / drumkit mode Roland GS Sysex message
843 which is usally used to switch a part between chromatic sounds and
844 drumkit sounds (as opposed to their common default setting of MIDI
845 part 10 being a drumkit part and all other ones chromatic parts), in
846 LS however one can switch between the first 16 MIDI instrument maps
847 defined for the sampler
848 - made it possible to create multiple sample channels even if the
849 audio thread isn't running
850 - fixed crash when removing channel with active voices (#116)
851 - bugfix: on sample reference changes (instrument editor), only
852 un-cache the respective sample if it's really not used by any
853 sampler engine anymore
854 - re-cache samples in case they were changed by an instrument editor,
855 e.g. when a sample was added while playing (#82)
856 - fixed hanging notes which occured when note-off event had the exact
857 same time stamp as the voice's note-on event and occured both in the
858 same audio fragment cycle (fixes bug #112)
859 - added support for the "fine tune" and "pitch bend range"
860 instrument-level gig parameters
861 - fixed minor artifacts in pitch bend handling
862 - added support for GS Reset SysEx message
863 - allow gig files to use unlimited downward pitch shifting
864 - added a limit check for upward pitch shifting
865 - bugfix: sometimes, when playing a note twice fast, the second
866 note was silent
867 - fixed crash happening when a pitch bend event arrived at the
868 same time a new instrument was loading
869
870 * LSCP server:
871 - added new LSCP event "CHANNEL_MIDI" which can be used by frontends to
872 react on MIDI data arriving on certain sampler channels (so far only
873 Note-On and Note-Off events are sent via this LSCP event)
874 - added new LSCP event "DEVICE_MIDI" which can be used by frontends to
875 react on MIDI data arriving on certain MIDI input devices (so far only
876 Note-On and Note-Off events are sent via this LSCP event)
877 - added new LSCP commands: FIND LOST DB_INSTRUMENT_FILES and
878 SET DB_INSTRUMENT FILE_PATH
879 - added new LSCP command: SEND CHANNEL MIDI_DATA which can be used by
880 frontends to send MIDI messages to specific sampler channel
881 - added two additional fields to GET FILE INSTRUMENT INFO command -
882 KEY_BINDINGS and KEYSWITCH_BINDINGS
883 - bugfix: the bank number provided by MIDI_INSTRUMENT_INFO notifications
884 was incorrect
885 - Optimized the retrieval of the MIDI instrument mappings
886 - added new LSCP command "SET VOICES" to globally alter the maximum
887 amount of voices, added new LSCP event "GLOBAL_INFO:VOICES" which will
888 be triggered respectively
889 - added new LSCP command "SET STREAMS" to globally alter the maximum
890 amount of disk streams, added new LSCP event "GLOBAL_INFO:STREAMS"
891 which will be triggered respectively
892 - bugfix: retry if "select" returns EINTR (this fixes a crash when
893 a gigedit file dialog is opened)
894 - close all connections when LSCPServer is deleted
895 - hard close of all sockets on exit
896 - bugfix: SET CHANNEL MIDI_INPUT_TYPE didn't work with the MME
897 driver
898
899 * Bug fixes:
900 - fixed a crash which occurs when removing a sampler channel waiting
901 to start instrument loading after another channel
902 - fixed a crash which occurs when removing a sampler channel with
903 instrument loading in progress (bug #113)
904 - fixed termination caused by uncaught exception when adding MIDI
905 instrument with PERSISTENT load mode
906 - fixed possible iterator invalidations when resetting the sampler
907 - fixed memory leaks when issuing the following LSCP commands:
908 GET AUDIO_OUTPUT_DRIVER INFO
909 GET AUDIO_OUTPUT_DRIVER_PARAMETER INFO
910 GET MIDI_INPUT_DRIVER INFO
911 GET MIDI_INPUT_DRIVER_PARAMETER INFO
912 - fixed possible compilation error when sqlite is not present
913 - fixed orphaned pointers when setting maximum voices limit (bug #118)
914 - fixed crash when changing the audio output device of a sampler
915 channel with loaded instrument and start playing notes
916 - fixed endless loop in Engine::SuspendAll() (bug #120)
917 - fixed a low-level atomic load function that was broken on 64-bit
918 PowerPC, which probably could cause crashes on that platform
919 - fixed a memory management error which could cause a crash when a
920 plugin was unloaded
921 - bugfix: two private structs had the same name, which could cause
922 problems if the linker chose the wrong constructor
923 - fixed low-level ConditionServer usage bug that caused lockups on
924 Windows
925
926
927 Version 0.5.1 (6 December 2007)
928
929 * packaging changes:
930 - added autoconf checks for pthread library
931 - added autoconf check for pthread bug found on certain NPTL-enabled
932 glibc versions (see Gentoo bug report #194076)
933 - added autoconf checks for MS Windows
934
935 * general changes:
936 - bugfix: the thread used by an editor plugin didn't die when the
937 editor closed
938 - bugfix: Ringbuffer.h: fill_write_space_with_null() did not zero
939 out all the space. operator--() did not apply size_mask after
940 decrementing the read_ptr. DEFAULT_WRAP_ELEMENTS set to 0 in
941 order to avoid problems with the _NonVolatileReader functions.
942 - bugfix: Stream.h: added a missing divide by BytesPerSample in
943 GetWriteSpace(). Since this function is currently only used in
944 the stream's qsort() compare function, it didn't trigger any
945 bugs.
946 - Resampler.h, Synthesizer.h: cubic interpolation now works in
947 24bit mode too. Faster method to read 24bit words on little
948 endian machines (x86): replaced 3 byte reads + shifts with a 1
949 unaligned 32bit read and shift
950 - experimental support for MS Windows (MIDI input via MME, AUDIO
951 output via ASIO)
952 - made handling of SIGINT signal (Ctrl-C) a bit more robust
953 - support for monitoring the total number of active disk streams
954 (new LSCP commands: GET TOTAL_STREAM_COUNT,
955 SUBSCRIBE TOTAL_STREAM_COUNT, UNSUBSCRIBE TOTAL_STREAM_COUNT)
956
957 * AUDIO driver:
958 - added Windows ASIO low latency audio driver
959
960 * MIDI driver:
961 - added MME Windows MIDI driver
962
963 * LSCP server:
964 - added support for Windows style path / filenames, however with
965 forward slash path separators instead of backslash
966 (i.e. "C:/foo/bar.gig")
967 - allow naughty liblscp to send non-string device parameters within
968 apostrophes as well
969 - added new LSCP commands: "GET FILE INSTRUMENTS <file>",
970 "LIST FILE INSTRUMENTS <file>" and
971 "GET FILE INSTRUMENT INFO <file> <index>" for retrieving informations
972 about an arbitrary instrument file on the system where the sampler is
973 running on
974
975 Version 0.5.0 (15 October 2007)
976
977 * packaging changes:
978 - config.h is not going to be installed along with liblinuxsampler's
979 API header files anymore
980 - only the API relevant header (and source) files will be exported to
981 the Doxygen API documentation (explicitly listed in Doxyfile.in)
982 - added completely new XCode project files for Mac OSX which is now
983 capable to execute our autoconf environment, thus no need anymore
984 to maintain the compile time configuration file (osx/version.h) for
985 OSX manually (patch by Toshi Nagata)
986 - fixed buggy boolean --enable-foo / --disable-foo configure script
987 parameters
988 - global.h now only covers global definitions that are needed for the
989 C++ API header files, all implementation internal global definitions
990 are now in global_private.h
991 - atomic.h is not exposed to the C++ API anymore
992 - no need to include config.h anymore for using LS's API header files
993 - fixed warnings in API doc generation
994
995 * general changes:
996 - replaced the old, confusing MIDI program change mechanism by a
997 flexible MIDI instrument mapper which allows to map arbitrary
998 (MIDI bank, MIDI program) pairs with arbitrary
999 (engine type, instrument file, file index) triplets which will be
1000 loaded on the respective channel when such MIDI program change
1001 messages arrive, beside that, each entry allows to define a life-time
1002 strategy for the instrument, so the user can define whether the
1003 instrument should i.e. be permanently loaded (to be able to switch
1004 quickly among sounds) or i.e. loaded just on demand when the
1005 respective program change arrives (to safe RAM space), as well as a
1006 global volume factor for each entry, so the user can adjust the volume
1007 dynamic of his mapped instrument collection without having to modify
1008 the instrument files, also one can manage arbitrary amount of such
1009 MIDI instrument maps and assign each sampler channel individually
1010 a certain map, so that i.e. separation of normal instruments and
1011 drumkits is possible
1012 - new notification events for tracking changes to audio/MIDI devices,
1013 MIDI instrument maps, MIDI instruments, FX sends, global volume.
1014 - sampler was limited to load max. 200 instruments in the background
1015 due to a constant size RingBuffer FIFO which is now replaced by a
1016 dynamic (unlimited) size std::list FIFO
1017 - added FX sends, these allow to route audio signals to arbitrary audio
1018 output channels for being processed by external effect processors
1019 (i.e. jack-rack), the send levels are controllable via arbitrary MIDI
1020 controllers
1021 - global (sampler wide) volume can now be controlled at runtime
1022 - Implemented new, improved notification system
1023 - fixed compilation errors regarding OSX
1024 (patch by Toshi Nagata)
1025 - implemented instruments database
1026 - added support for escape sequences to the instruments database
1027 - added highly experimental support for on-the-fly instrument editing
1028 within the sampler's process (by using instrument editor plugins),
1029 you'll notice the new "Registered instrument editors:" message on
1030 startup, added a new LSCP command:
1031 "EDIT CHANNEL INSTRUMENT <sampler-channel>"
1032 to spawn a matching instrument editor for the instrument on the
1033 given sampler channel, the plugin path can be overridden at compile
1034 time with ./configure --enable-plugin-dir=/some/dir
1035 - added experimental code for synchronizing instrument editors hosted
1036 in the sampler's process to safely edit instruments while playing
1037 without a crash (hopefully) by either suspending single regions
1038 wherever possible or if unavoidable whole engine(s)
1039 - fixed several issues in fundamental "Thread" class: set scheduling
1040 policy and priority on thread level, set a minimum stack size for
1041 thread (TODO: a reasonable value yet to be tested), bugfix: non-RT
1042 threads simply inherited properties of starting thread instead of
1043 setting their own policy and priority
1044 - minor fix in our automatic stack trace mechanism on crashes, the
1045 main process did not wait for the stack trace process to finish
1046 its output
1047 - fixed some minor memory leaks
1048 - reenabled assembly features support, at the moment only for
1049 enabling a fast denormal FPU mode (x86 platforms supporting SSE2)
1050 - minor assembly fix in x86 features detection (don't use the PIC
1051 register, to avoid relocations in the text segment at runtime)
1052 - POSIX callback functions of Thread.h are hidden
1053
1054 * MIDI driver:
1055 - dispatch bank select (MSB and LSB) messages
1056
1057 * audio driver:
1058 - the ALSA audio output driver parameters now reflect the correct
1059 parameter value ranges for the respective selected sound card
1060 (patch by Till Wimmer, a bit fixed and extended)
1061
1062 * Gigasampler format engine:
1063 - if a filter is used and EG2 finishes before EG1, let the voice
1064 die when EG2 has finished (this fixes a problem with clicks and
1065 voice starvation for some gigs)
1066 - playback is no longer disabled during instrument loading
1067 - all notes playing on a channel that changes its instrument keep
1068 playing with the old instrument until they get a note off
1069 command
1070 - EG fix: a release value of zero could cause noises or crash
1071 - handle MIDI coarse tuning messages (MIDI RPN #0 MSB #2 LSB)
1072 - EG fine tuning: when attack is zero the EG starts at a level
1073 above max sustain level, which means that there is a decay phase
1074 even if sustain is 100%
1075 - more EG fixes: the level could sometimes go below zero and cause
1076 noises or crashes
1077 - minor fix of EGDecay (patch by Toshi Nagata)
1078 - fixed compiler error when --enable-override-filter-type was
1079 supplied to the configure script (fixes #46)
1080 - disk thread: queue sizes are now proportional to CONFIG_MAX_STREAMS
1081 instead of a fix value
1082 - behavior fix: on MIDI CC# 65 (portamento on / off), 126 (mono mode),
1083 127 (solo mode) only kill voices if the respective mode really
1084 changed
1085
1086 * LSCP server:
1087 - fixed compile time error for old Bison versions
1088 (i.e. v1.28 found on MacOS 10.4, patch by Toshi Nagata)
1089 - parser now supports extended ASCII character set
1090 (up to ASCII code 255, i.e. includes now umlauts and accents)
1091 - filename arguments in LSCP commands now allow to use escape
1092 sequences, that is directly literal as one of: \', \", \\, \n, \r,
1093 \f, \t, \v, or as octal ASCII code value like \132, or as hex ASCII
1094 code value like \xf2) (fixes bug #24)
1095 - the following LSCP commands now also support escape sequences for at
1096 least one of their text-based parameters (i.e. name, description):
1097 "ADD MIDI_INSTRUMENT_MAP", "MAP MIDI_INSTRUMENT",
1098 "SET MIDI_INSTRUMENT_MAP NAME", "SET FX_SEND NAME", "CREATE FX_SEND",
1099 "SET DB_INSTRUMENT_DIRECTORY NAME",
1100 "SET DB_INSTRUMENT_DIRECTORY DESCRIPTION", "SET DB_INSTRUMENT NAME",
1101 "SET DB_INSTRUMENT DESCRIPTION", "FIND DB_INSTRUMENTS",
1102 "FIND DB_INSTRUMENT_DIRECTORIES"
1103 - returns verbose syntax errors (line and column where syntax error
1104 occured, the unexpected character and the actually expected, possible
1105 character(s), the latter only if less than 5 possibilities)
1106 - made sure that LSCP syntax is not affected by gigedit locale
1107 settings
1108 - bugfix regarding strings parameter lists: all comma separated lists
1109 of strings were treated as being one string containing commas
1110 (fixes #57)
1111
1112 * Bug fixes:
1113 - fixed crash occurring on certain LSCP scripts (Bug 39)
1114 - another thread safety fix for lscp "load engine" and "set
1115 channel audio output device"
1116 - fixed a crash which occurs when reassigning the same engine
1117 on a sampler channel with connected MIDI device
1118 - fixed a crash which occurs when changing the number of ports of a MIDI
1119 device connected to a sampler channel to number less then or equal
1120 to the index of the port to which the sampler channel is connected.
1121 - The previous bindings were not been disconnected when altering
1122 the ALSA_SEQ_BINDINGS parameter. Introduced a NONE keyword for
1123 unsubscribing from all bindings (e.g. ALSA_SEQ_BINDINGS=NONE).
1124 - The active stream/voice count statistic was incorrect.
1125 - notification events were not been sent for some sampler
1126 channel changes
1127 - added default min and max values to restrict the number of allowed
1128 audio output channels and MIDI input ports
1129 - the connection to the PCM interface is now closed when destroying
1130 an audio output device
1131 - files with slash in their path or filename could not be loaded
1132
1133 * test cases:
1134 - updated and fixed (haven't been touched in a while)
1135
1136 Version 0.4.0 (24 November 2006)
1137
1138 * packaging changes:
1139 - changed deprecated copyright attribute to license; added ldconfig
1140 to post-(un)install steps; added devel package for liblinuxsampler;
1141 to linuxsampler.spec (RPM)
1142 - install necessary development header files for allowing 3rd party
1143 applications to link against liblinuxsampler
1144 - liblinuxsampler's API documentation can be generated with 'make docs'
1145 (Doxygen required)
1146 - added benchmark to automatically detect the best triangle LFO
1147 implementation (currently either an integer math solution or a
1148 di-harmonic approximation), automatic detection can be overriden
1149 with --enable-signed-triang-algo=x and --enable-unsigned-triang-algo=x
1150 configure script argument though (mandatory for cross-compilation)
1151 - do not automatically pick optimized gcc flags if the user already
1152 provided some on his own (as CXXFLAGS)
1153 - added compile time option to disable processing of All-Notes-Off MIDI
1154 messages
1155 - added compile time options to allow disabling the various audio and
1156 MIDI drivers
1157 - fixed automatic GCC CPU switch detection on PPC
1158 (patch by Ebrahim Mayat)
1159
1160 * Gigasampler format engine:
1161 - extensive synthesis optimization
1162 (reimplementation of EGs and LFO(s), removed synthesis parameter
1163 prerendering and the synthesis parameter matrix in general, splitting
1164 each audio fragment into subfragments now where each subfragment uses
1165 constant pitch and filter coefficients. The volume coefficient is
1166 linearly interpolated inside a subfragment, unless
1167 --disable-interpolate-volume is set.)
1168 - fine tuning of the EG modulation parameters
1169 - improved filter cutoff calculation by adding support for the
1170 following gig parameters: Cutoff freq (used when no cutoff
1171 controller is defined), Control invert, Minimum cutoff, Velocity
1172 curve and Velocity range. The keyboard tracking now scales
1173 cutoff frequency instead of resonance.
1174 - added support for gig parameter Resonance.
1175 - fixed bug in sysex handling (patch by Juan Linietsky)
1176 - added global volume attenuation of -9 dB (0.35f) to prevent clipping
1177 which can be overridden with --enable-global-attenuation
1178 - EG fixes: made the length of "attack hold" stage more
1179 accurate. Release stage can now start before attack stage
1180 ends. Cancel release didn't work when sustain was zero. Attack
1181 time now has a minimal value to prevent clicks.
1182 - fixed pitch changes larger than one octave
1183 - fixed EG3 (pitch envelope) synthesis which was neutral all the time
1184 - implemented portamento mode and solo mode (a.k.a 'mono mode'):
1185 all modes can be altered via standard GM messages, that is CC5 for
1186 altering portamento time, CC65 for enabling / disabling portamento
1187 mode, CC126 for enabling solo mode and CC127 for disabling solo mode
1188 - fine tuning of the curves for volume (CC7), pan (CC10 and gig
1189 parameter) and crossfade
1190 - added support for the "attenuation controller threshold" gig
1191 parameter
1192 - added smoothing of volume changes caused by control change
1193 messages
1194 - sample loop parameters are now taken from the DimensionRegion
1195 instead of the wave chunk
1196 - fixed keyswitching for v3 gigs with a number of keyswitch splits
1197 not equal to a power of two
1198 - reimplementation of the filter algorithm. The new filter is more
1199 accurate and supports all gig filter types, including bandreject
1200 and lowpass turbo.
1201 - real support for 24 bit samples - samples are not truncated to
1202 16 bits anymore
1203 - support for aftertouch (channel pressure, not polyphonic
1204 aftertouch)
1205 - LFO1 behaviour fixed (dampening from max volume instead of
1206 amplifying from 0)
1207
1208 * LSCP server:
1209 - fixed application exit on broken pipe error (fixes bug #20)
1210 - fixed the notification messages delay due to lack of
1211 network activity (fixes bug #26)
1212 - fixed parser bug which occured on space(s) within device parameters
1213 (e.g. "SET AUDIO_OUTPUT_CHANNEL_PARAMETER 0 0 SOMEPARAM='foo bar'")
1214
1215 * audio driver:
1216 - added aRts audio output driver (by no means RT safe)
1217
1218 * MIDI driver:
1219 - fixed legacy sysex code which caused dispatching of MIDI SysEx
1220 messages several times instead of once
1221 - API extension for MIDI drivers which already supply exact time stamps
1222 for events (i.e. for offline rendering based MIDI drivers)
1223 - added checks for bad MIDI data
1224
1225 * linuxsampler application:
1226 - show available sampler engine types on startup
1227
1228 * general changes:
1229 - support for muting sampler channels and solo mode of the same, two new
1230 LSCP commands ("SET CHANNEL MUTE" and "SET CHANNEL SOLO") and two new
1231 fields ("MUTE" and "SOLO") for command "GET CHANNEL INFO" were
1232 introduced for this, the behavior is the same like on a mixer console
1233 (patch by Grigor Iliev, a bit adjusted). Also added configure option
1234 --enable-process-muted-channels which can be used to enable the
1235 processing of muted channels.
1236 - support for sostenuto pedal
1237 - support for monitoring the total number of active voices
1238 (new LSCP commands: GET TOTAL_VOICE_COUNT, GET TOTAL_VOICE_COUNT_MAX,
1239 SUBSCRIBE TOTAL_VOICE_COUNT, UNSUBSCRIBE TOTAL_VOICE_COUNT)
1240 - fixed some memory management errors
1241 - fixed some concurrency problems that could lead to crashes when
1242 LSCP commands were executed
1243 - fixed crash when instrument loading failed previously on the same
1244 sampler channel (fixes bug #36)
1245
1246 Version 0.3.3 (15 July 2005)
1247
1248 * packaging changes:
1249 - fixed compilation with gcc 4.0
1250
1251 * Gigasampler format engine:
1252 - LFO preprocessing was performed even though the respective LFO was
1253 disabled by the instrument patch (minor efficiency fix)
1254 - if period time of chosen audio device is too small (< MIN_RELEASE_TIME)
1255 for volume ramp downs in the same fragment (needed for current voice
1256 stealing implementation) then simply inform the user with a warning
1257 about possible click sounds and reduce the volume ramp down
1258 appropriately instead of cancelling the audio device connection
1259
1260 Version 0.3.2 (24 June 2005)
1261
1262 * packaging changes:
1263 - updated autotools build files to compile on OS X
1264 (conditional compilation of CoreMIDI and MidiShare drivers)
1265 - hand-crafted assembly optimization code can be disabled with
1266 './configure --disable-asm' (definitely not recommended)
1267 - fixed 'make dist' rule to include all necessary files
1268 - require automake (>= 1.5) for 'make -f Makefile.cvs'
1269 (due to 'dist-bzip2' automake option)
1270
1271 * Gigasampler format engine:
1272 - support for the gig parameters for "release velocity response" curves
1273 - fine tuning of the EGADSR envelope
1274 - volume of release triggered samples now depends on note-on velocity,
1275 note length and gig parameter "release trigger decay" instead of
1276 note-off velocity.
1277 - revised voice stealing
1278 (fixes crash and endless loop caused by voice stealing)
1279 - don't reset scale tuning on instrument or audio output device change
1280 - handle key group conflicts right at the beginning of each fragment
1281 instead of when the respective voice is actually launched
1282 (fixes undefined behavior if stolen voices belonged to a key group -
1283 this case was followed by a "killed voice survived" error message)
1284 - fixed minor issue with null/silence samples
1285 (those stole voices even though they don't need a voice at all which
1286 resulted in "voice stealing didn't work out" messages)
1287 - don't reset volume, pan, pitch and MIDI controller values on
1288 instrument or audio output device change
1289
1290 * LSCP server:
1291 - fixed some crashes (patch by Grigor Iliev, fixes #19)
1292 - fixed LSCP event "CHANNEL_INFO" notification
1293 (e.g. did not notify on volume changes or MIDI program change events)
1294
1295 * linuxsampler application:
1296 - added command line parameters --lscp-addr and --lscp-port to override
1297 default IP address and port of LSCP server
1298
1299 Version 0.3.1 (24 May 2005)
1300
1301 * initial release

  ViewVC Help
Powered by ViewVC