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

Contents of /linuxsampler/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 3293 - (show annotations) (download)
Tue Jun 27 22:19:19 2017 UTC (2 years, 3 months ago) by schoenebeck
File size: 74649 byte(s)
* NKSP: Added built-in script function "fork()".
* NKSP: Added built-in array variable %NKSP_CALLBACK_CHILD_ID[].
* NKSP: Added built-in variable $NKSP_CALLBACK_PARENT_ID.
* NKSP: Fixed potential crash when accessing dynamic built-in
  array variables.
* Bumped version (2.0.0.svn65).

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 - sfz parser: allow missing space between header and opcode
11
12 * Gigasampler/GigaStudio format engine:
13 - Fixed clicks and pumping noise with Lowpass Turbo filter on very low
14 cutoff settings.
15
16 * general changes:
17 - fixed printf type errors (mostly in debug messages)
18 - use unique_ptr instead of auto_ptr when building with C++11
19 - Added RTAVLTree class which is a real-time safe ordered multi-map, thus
20 allowing to sort data efficiently in real-time safe manner.
21 - RTList class: added methods for moving/inserting elements to arbitrary
22 position within a list.
23 - RTList class: added method fromPtr() for creating an Iterator object from
24 a raw element pointer.
25 - Attempt to partly fix resetting engine channels vs. resetting engine, an
26 overall cleanup of the Reset*(), ConnectAudioDevice(),
27 DisconnectAudioDevice() API methods would still be desirable though,
28 because the current situation is still inconsistent and error prone.
29 - Active voices are now internally grouped to "Note" objects, instead of
30 being directly assigned to a keyboard key. This allows more fine graded
31 processing of voices, which is i.e. required for certain instrument
32 script features.
33 - Fix: Release trigger voices were not spawned on sustain pedal up
34 (CC #64) events.
35 - Fix: Release trigger voices can now also distinguish correctly between
36 note-off and sustain pedal up events.
37 - All engines: Increased ramp speed of volume smoother and pan smoother
38 (while slow rate of crossfade smoother is preserved) to allow quick
39 volume and pan changes by instrument scripts for instance.
40 - gig/sf2/sfz: Fixed aftertouch channel pressure events being processed
41 properly and thus were ignored for certain uses.
42 - Added new C++ API method
43 VirtualMidiDevice::SendChannelPressureToSampler().
44 - windows, 32-bit: fixed potential crashes by making sure the stack in
45 sub threads is 16-byte aligned
46 - fixed numerous compiler warnings
47 - Fixed invalid (note-on) event ID being assigned to new Note objects.
48 - Revised fundamental C++ classes "Thread", "Mutex" and "Condition" which
49 fixes potential undefined behavior.
50
51 * packaging changes:
52 - removed unnecessary dependency to libuuid
53 (originated by libgig's usage of it)
54 - Automake: set environment variable GCC_COLORS=auto to allow GCC to
55 auto detect whether it (sh/c)ould output its messages in color.
56
57 * Real-time instrument scripts:
58 - Implemented scheduler for delayed MIDI events and for suspended scripts.
59 - Built-in script function "wait()": implemented support for function's
60 "duration-us" argument, thus scripts using this function are now
61 correctly resumed after the requested amount of microseconds.
62 - Built-in script function "play_note()": implemented support for
63 function's "duration-us" argument, thus notes triggered with this
64 argument are now correctly released after the requested amount of
65 microseconds.
66 - Fix: script events were not cleared when engine channel was reset,
67 potentially causing undefined behavior.
68 - Fixed crash which happened when trying to reference an undeclared
69 variable.
70 - Built-in script function "play_note()": Added support for passing
71 special value -1 for "duration-us" argument, which will cause the
72 triggered note to be released once the original note was released.
73 - Instrument script classes now exported with the liblinuxsampler C++ API.
74 - Added new API method ScriptVM::syntaxHighlighting() which provides
75 a convenient syntax highlighting backend for external instrument
76 script editor applications.
77 - Added new C++ API class "ScriptVMFactory".
78 - Extended parser issues to provide not only first line and first
79 column, but also last line and last column of issue (thus marking
80 the precise span of the issue within the source code).
81 - Reload script automatically after being modified by an instrument
82 editor.
83 - NKSP language grammar correction: allow empty event handler bodies
84 like "on note end on".
85 - Implemented built-in script function "change_vol()".
86 - Implemented built-in script function "change_tune()".
87 - Implemented built-in script function "change_pan()".
88 - Implemented built-in script function "change_cutoff()".
89 - Implemented built-in script function "change_reso()".
90 - Implemented built-in script function "event_status()".
91 - Added built-in script constants "$EVENT_STATUS_INACTIVE" and
92 "$EVENT_STATUS_NOTE_QUEUE" both for being used as flags for
93 "event_status()" function.
94 - NKSP language: Added support for bitwise operators ".or.", ".and."
95 and ".not.".
96 - NKSP language scanner: Fixed IDs matching to require at least one
97 character (i.e. when matching function names or variable names).
98 - NKSP language scanner: disabled unusued rules.
99 - Fixed behavior of play_note() and note_off() functions which must
100 be distinguished engine internally from "real" MIDI note on/off
101 events in order to avoid misbehaviors like hanging notes.
102 - Implemented built-in script variable "$KSP_TIMER".
103 - Implemented built-in script variable "$NKSP_REAL_TIMER".
104 - Implemented built-in script variable "$NKSP_PERF_TIMER".
105 - Implemented built-in script variable "$ENGINE_UPTIME".
106 - Implemented built-in script function "inc()".
107 - Implemented built-in script function "dec()".
108 - NKSP language fix: division expressions were evaluated too often.
109 - NKSP language fix: string concatenation operator was right
110 associative instead of left (to right).
111 - NKSP language correction: allow empty statements in entire language
112 (i.e. "if end if").
113 - Implemented built-in script function "stop_wait()".
114 - Implemented built-in script variable "$NI_CALLBACK_ID".
115 - Implemented built-in script variable "$NI_CALLBACK_TYPE".
116 - Implemented built-in script variable "$NKSP_IGNORE_WAIT".
117 - Added support for read-only built-in variables (respectively
118 handled by the script parser).
119 - Added built-in script constant "$NI_CB_TYPE_INIT".
120 - Added built-in script constant "$NI_CB_TYPE_NOTE".
121 - Added built-in script constant "$NI_CB_TYPE_RELEASE".
122 - Added built-in script constant "$NI_CB_TYPE_CONTROLLER".
123 - NKSP Language: Added support for user defined script functions.
124 - Implemented built-in script function "change_attack()".
125 - Implemented built-in script function "change_decay()".
126 - Implemented built-in script function "change_release()".
127 - Fixed all change_*() built-in script functions to apply their
128 synthesis parameter changes immediately in case the respective note
129 was triggered at the same time, instead of scheduling the parameter
130 change, especially because it would cause some parameter types's
131 changes either to be ramped (i.e. change_vol()) or other types even
132 to have not effect at all (i.e. change_attack()).
133 - Implemented built-in script function "sh_left()".
134 - Implemented built-in script function "sh_right()".
135 - Implemented built-in script function "min()".
136 - Implemented built-in script function "max()".
137 - NKSP Fix: "init" event handler was not always executed when a script
138 was loaded.
139 - NKSP built-in wait() script function: abort script execution if
140 a negative or zero wait time was passed as argument (since this is
141 a common indication of a bug either of the script or even of the
142 engine, which could lead to RT instability or even worse).
143 - ScriptVM: Implemented automatic suspension of RT safety
144 threatening scripts.
145 - Provide more user friendly error messages on syntax errors.
146 - Fixed NKSP parser warning "Not a statement" when assigning an
147 initializer list to an array variable.
148 - Implemented built-in script array variable "%ALL_EVENTS".
149 - Implemented built-in script function "in_range()".
150 - Implemented built-in script function "change_amp_lfo_depth()".
151 - Implemented built-in script function "change_amp_lfo_freq()".
152 - Implemented built-in script function "change_pitch_lfo_depth()".
153 - Implemented built-in script function "change_pitch_lfo_freq()".
154 - Implemented built-in script function "change_vol_time()".
155 - Implemented built-in script function "change_tune_time()".
156 - Implemented built-in script function "fade_in()".
157 - Implemented built-in script function "fade_out()".
158 - Fixed acceptance of wrong data type of parameters passed to built-in
159 script functions "change_vol()", "change_tune()", "change_pan()",
160 "change_cutoff()", "change_reso()", "change_attack()", "change_decay()",
161 "change_release()", "change_amp_lfo_depth()", "change_amp_lfo_freq()",
162 "change_pitch_lfo_depth()" and "change_pitch_lfo_freq()".
163 - Added built-in script function "get_event_par()" and implemented some
164 of its possible parameter selections.
165 - Added built-in script function "set_event_par()" and implemented some
166 of its possible parameter selections.
167 - Fixed a bunch of scheduler time related bugs.
168 - Fixed polyphonic variables not being reset to zero after usage.
169 - Built-in "ignore_event()" function: argument is now optional, like with
170 built-in function "ignore_controller()".
171 - Implemented built-in script function "change_velo()".
172 - Implemented built-in script function "change_note()".
173 - Adjusted behavior of "change_vol()" and "change_tune()" to a more
174 intuitive behavior if used in combination with "change_vol_time()" or
175 "change_tune_time()" respectively: now tuning/volume changes are only
176 assigned (without delay) immediately to a new note if the respective
177 timing function has not been called before, otherwise the volume/tuning
178 changes are automatically faded (before, only the event's time stamp was
179 relevant).
180 - Implemented built-in script function "array_equal()".
181 - Implemented built-in script function "search()".
182 - Implemented built-in script function "sort()".
183 - NKSP Fix: Never suspend "init" event handlers.
184 - Implemented built-in script function "same_region()" (currently only
185 available for gig format engine).
186 - Added built-in script constant "$NKSP_LINEAR".
187 - Added built-in script constant "$NKSP_EASE_IN_EASE_OUT".
188 - Implemented built-in script function "change_vol_curve()".
189 - Implemented built-in script function "change_tune_curve()".
190 - built-in "play_note()" function now supports a sample playback start
191 offset with argument 3, where special value -1 means to use the regular
192 sample offset as defined by the instrument file.
193 - Built-in array variable %KEY_DOWN[] is now a read-only variable.
194 - Built-in variable $EVENT_NOTE is now a read-only variable.
195 - Built-in variable $EVENT_VELOCITY is now a read-only variable.
196 - built-in "play_note()" function now accepts -2 for its fourth argument
197 (note duration) which means the life time of the note shall be sticked
198 to the requested note number of argument 1.
199 - Fix: built-in "play_note()" function now returns 0 as result value if
200 -1 was passed for its fourth argument (note duration) and the respective
201 parent note is already gone.
202 - Implemented built-in script function "change_play_pos()".
203 - NKSP language: Added support for user declared const array variables.
204 - NKSP language: Raise parser warning if array variable is accessed with
205 an index that exceeds the array's size.
206 - NKSP language fix: Unknown characters were not handled correctly.
207 - NKSP language: Added support for "synchronized .. end synchronized"
208 code blocks.
209 - Implemented built-in script function "abort()" which allows to abort
210 another script handler by passing its callback ID.
211 - Fixed potential memory access bug and potential undefined behavior of
212 "init" event handlers.
213 - Print a time stamp along to each call of built-in function "message()".
214 - ScriptVM API: Added VMParserContext::preprocessorComments() which allows
215 to retrieve all code blocks filtered out by the preprocessor.
216 - Added built-in script function "fork()".
217 - Added built-in array variable %NKSP_CALLBACK_CHILD_ID[].
218 - Added built-in variable $NKSP_CALLBACK_PARENT_ID.
219 - Fixed potential crash when accessing dynamic built-in array variables.
220
221 * Instruments DB:
222 - Fixed memory access bug of general DB access code which lead to
223 undefined behavior.
224 - Cleanup of instruments DB file creation and opening code.
225 - The instrument DB path of linuxsampler's --create-instruments-db argument
226 is now optional, if it is missing, then a default location is used.
227 - Added support for scanning SFZ (.sfz) files.
228 - Added support for scanning Sound Font (.sf2) files.
229 - Fixed undefined DB transaction behavior.
230
231 Version 2.0.0 (15 July 2015)
232
233 * packaging changes:
234 - fixed building with newer MinGW-w64
235 - Mac OS X: support the new dir for Core Audio SDK
236 - Mac OS X: fixed building outside source directory
237 - made sure all source files for hostplugins are included when
238 doing "make dist"
239 - removed empty directories left from the cvs to svn migration
240 - fixed compilation with gcc 4.6.1
241 - another "make dist" fix, for the LV2 plugin
242 - made --enable-pthread-testcancel default on Mac OS X
243 - minor Makefile fix for building lscpparser when build and source
244 directories are separate
245 - Mac OS X: made it possible to specify plugin installation dir to
246 configure
247 - Mac OS X: Makefile fix for the install-strip target
248 - fixed compilation with gcc 4.7
249 - fixed configure script error with old autoconf versions
250 - lsatomic.h: use gcc provided atomic functions if building with
251 gcc 4.7 and C++11
252 - modernized configure script
253 - fixed linkage error when building with
254 LDFLAGS="-Wl,--no-undefined" (#190)
255 - fixed compilation with Clang 3.2
256 - removed usage of deprecated Automake variable INCLUDES
257 - fixed building with C++11
258 - build fix: ChangeFlagRelaxed.h was missing in makefile
259 - build fix: libsndfile compiler flags were missing in some
260 makefiles
261 - fix for building with bison 3.0 (#202)
262 - Mac OS X: added temporary hack allowing to spawn gigedit as callback
263 on the process's main thread
264 - fixed build error on newer MinGW
265 - support building with older jack versions
266 - support building with spaces in vst sdk path
267 - enabled automake 'subdir-objects' option and moved external
268 source references (vst, au, asio) from makefiles to cpp files,
269 in order to get rid of warnings from automake 1.14
270
271 * general changes:
272 - Refactoring: moved the independent code from
273 the Gigasampler format engine to base classes
274 - command line option '--profile' is currently disabled, since the
275 respective profiling code is currently broken
276 - Introduced support for optional environment variable
277 "LINUXSAMPLER_PLUGIN_DIR" which allows to override the directory
278 where the sampler shall look for instrument editor plugins
279 (patch by Luis Garrido, slightly modified).
280 - implemented Roland GS NRPN 1ArrH which allows to set volume per note
281 - implemented Roland GS NRPN 1CrrH which allows to set pan per note
282 - implemented Roland GS NRPN 1DrrH which allows to set reverb send per
283 note (in this implementation of the sampler its simply hard routed to
284 the 1st effect send of the sampler channel, no matter what the actual
285 effect type is)
286 - implemented Roland GS NRPN 1ErrH which allows to set chorus send per
287 note (in this implementation of the sampler its simply hard routed to
288 the 2nd effect send of the sampler channel, no matter what the actual
289 effect type is)
290 - implemented support for internal LADSPA effects
291 - raised limit of program change queue from 100 to 512
292 (as suggested by Alex Stone)
293 - implemented sine LFO, pulse LFO and saw LFO
294 - experimental support for per voice equalization
295 - added command line option --exec-after-init
296 - Introduced new C++ API method:
297 EngineChannel::InstrumentFileName(int index)
298 allowing to retrieve the whole list of files used for the loaded
299 instrument on an engine channel (a.k.a. part). Some GigaStudio
300 instruments for example are splitted over several files like
301 "Foo.gig", "Foo.gx01", "Foo.gx02", ...
302 - Added new C++ API method Sampler::GetGlobalMaxVoices().
303 - Added new C++ API method Sampler::GetGlobalMaxStreams().
304 - Added new C++ API method Sampler::SetGlobalMaxVoices().
305 - Added new C++ API method Sampler::SetGlobalMaxStreams().
306 - Various "const" and "restrict" optimizations.
307 - all engines: add pan CC value to instrument pan parameter before
308 applying panning, instead of using two separate pan functions in
309 series (#182)
310 - added a lock guard class for exception safe mutex handling and
311 used it everywhere appropriate
312 - Immediately apply scale tuning changes to active voices.
313 - Exposed scale tuning to C++ API (along to the already existing standard
314 SysEx way).
315 - lsatomic.h fixes: seq_cst load and store were suboptimal for x86
316 and broken for ppc64. (Seq_cst loads and stores are actually not
317 used in LS, so the bug wasn't noticable.)
318 - lsatomic.h: added ARMv7 support
319 - Added support for multiple MIDI input ports per sampler channel (and
320 added various new C++ methods for this new feature / design change, old
321 C++ API methods for managing SamplerChannel's MIDI inputs are now marked
322 as deprecated but are still there and should provide full behavior
323 backward compatibility).
324 - AbstractEngine::GSChecksum(): don't allocate memory on the stack (was
325 unsafe and caused compilation error with clang 2.x).
326 - Bugfix: only process the latest MIDI program change event.
327 - Introducing the LSCP shell, which provides convenient control of the
328 sampler from the command line by providing LSCP aware features.
329 - VirtualMidiDevice: Added support for program change.
330 - VirtualMidiDevice: Added support for bank select (MSB & LSB).
331 - VirtualMidiDevice: Added support for pitch bend.
332 - Aftertouch: extended API to explicitly handle channel pressure and
333 polyphonic key pressure events (so far polyphonic pressure was not
334 supported at all, and channel pressure was rerouted as CC128 but not
335 used so far).
336 - Added initial support for real-time instrument scripts. The script VM
337 code is shared by all sampler engine implemementations, however only the
338 gig file format currently provides support for storing instrument scripts
339 (as LinuxSampler extension to the original GigaStudio 4 format).
340
341 * Gigasampler format engine:
342 - implemented the "round robin keyboard" dimension
343 - fixed round robin and random dimensions for cases when number of
344 dimension zones is not a power of two
345 - made round robin use a counter for each region instead of each
346 key
347 - bugfix: pitch LFO controller "internal+aftertouch" was broken
348 - bugfix: filter keyboard tracking was broken
349 - filter performance fix (an unnecessary copy was made of the
350 filter parameters in each sub fragment)
351 - handle special case when pan parameter in gig file has max or
352 min value
353 - Exclusive Groups: don't ever stop voices of the same note,
354 doesn't sound naturally with a drumkit
355 - fixed EG1 modulation when attack or release is zero
356 - Fixed support for 'aftertouch' attenuation controller.
357 - Fixed crash that happened with velocity split sounds under certain
358 conditions (see also previous commit on libgig).
359 - fixed behaviour of filter LFO
360
361 * SFZ format engine:
362 - Initial implementation (not usable yet)
363 - added support for v2 multiple stage envelope generators
364 - added a fine-tuned v1 envelope generator instead of using the
365 one from the gig engine
366 - fixed memory leak and memory handling errors
367 - added support for trigger=first, trigger=legato and sw_previous
368 - allow non-numerical key values ("C#4" for example)
369 - "key" opcode now sets pitch_keycenter too
370 - fixed error when unloading instrument with same sample used by
371 multiple regions
372 - added some opcode aliases, like loopmode for loop_mode, to be
373 more compatible
374 - added support for trigger=release and rt_decay
375 - added support for off_mode=normal
376 - added support for random, seq_position, seq_length and volume
377 - added v1 LFO opcodes to parser (no support in engine yet)
378 - added support for amp_veltrack and amp_velcurve_N
379 - fine-tuned the default velocity curve
380 - added support for transpose
381 - fixed crash when using sw_down/up
382 - improved logic for sw_lokey/hikey/up/down/last
383 - added more v1 aliases to parser
384 - reduced memory usage for sfz data
385 - RT-safeness: avoid malloc in audio thread
386 - fixed a bug that could cause voice stealing to fail
387 - optimized sample lookup
388 - improved support for exclusive groups (group, off_by and
389 off_mode)
390 - added support for controller triggered regions (on_locc/on_hicc)
391 - added support for loop_mode=one_shot
392 - fixed playback of 16 bit wav files on big endian CPUs
393 - added support for Ogg Vorbis sample files
394 - use loop markers from sample file if loop_start and loop_end are
395 not set in sfz file
396 - implemeted filters. Filter types: lowpass, bandpass, bandreject
397 and highpass. 1, 2, 4 and 6 pole filters. Opcodes: fil_type,
398 cutoff, resonance, fil_veltrack, fil_keytrack, fil_keycenter,
399 cutoff_cc, cutoff_chanaft.
400 - bugfix: zero ampeg_sustain didn't work
401 - added support for velocity effect on amplifier envelope time
402 (ampeg_vel2attack, ampeg_vel2decay, ampeg_vel2sustain and
403 ampeg_vel2release)
404 - added support for EG hold (ampeg_hold)
405 - added support for sample offset (offset)
406 - use the newly introduced signal units model
407 - implemented opcodes lfoN_delay,
408 lfoN_freq, lfoN_pan, lfoN_cutoff, lfoN_resonance
409 - implemented opcodes ampeg_delay, ampeg_vel2delay
410 pitcheg_delay, pitcheg_start, pitcheg_attack, pitcheg_hold,
411 pitcheg_decay, pitcheg_sustain, pitcheg_release, pitcheg_vel2delay,
412 pitcheg_vel2attack, pitcheg_vel2hold, pitcheg_vel2decay,
413 pitcheg_vel2sustain, pitcheg_vel2release, pitcheg_depth
414 - implemented opcodes pitchlfo_delay, pitchlfo_freq,
415 pitchlfo_depth, fillfo_delay, fillfo_freq, fillfo_depth,
416 amplfo_delay, amplfo_freq, amplfo_depth
417 - implemented opcodes fileg_delay, fileg_start, fileg_attack,
418 fileg_hold, fileg_decay, fileg_sustain, fileg_release, fileg_vel2delay,
419 fileg_vel2attack, fileg_vel2hold, fileg_vel2decay, fileg_vel2sustain,
420 fileg_vel2release
421 - implemented opcode lfoN_wave
422 - implemented opcode pitchlfo_depthccN
423 - implemented opcodes lfoN_phase, lfoN_phase_onccX,
424 lfoN_pitch, lfoN_pitch_onccX
425 - implemented opcodes pitchlfo_fade,
426 fillfo_fade, amplfo_fade, lfoN_fade, lfoN_fade_onccX
427 - implemented opcodes fillfo_freqccN,
428 pitchlfo_freqccN, amplfo_freqccN, lfoN_freq_onccX
429 - implemented opcodes ampeg_delayccN, ampeg_startccN,
430 ampeg_attackccN, ampeg_holdccN, ampeg_decayccN, ampeg_sustainccN,
431 ampeg_releaseccN, egN_timeX_onccY, egN_levelX_onccY
432 - lfoN_* and egN_* opcodes defined in group sections
433 are now taken into account
434 - implemented curves
435 - implemented opcodes volume_onccN, volume_curveccN
436 - implemented opcode volume_smoothccN
437 - implemented opcodes fillfo_depthccN, amplfo_depthccN,
438 lfoN_volume, lfoN_volume_onccX, lfoN_volume_smoothccX,
439 lfoN_freq_smoothccX, lfoN_pitch_smoothccX, lfoN_pan_onccX,
440 lfoN_pan_smoothccX, lfoN_cutoff_onccX, lfoN_cutoff_smoothccX,
441 lfoN_resonance_onccX, lfoN_resonance_smoothccX, lfoN_delay_onccX
442 - implemented opcode end
443 - implemented opcodes egN_amplitude_onccX,
444 egN_volume, egN_volume_onccX, egN_cutoff, egN_cutoff_onccX,
445 egN_pitch, egN_pitch_onccX, egN_resonance, egN_resonance_onccX
446 - implemented opcodes xfin_lokey, xfin_hikey,
447 xfout_lokey, xfout_hikey, xf_keycurve, xfin_lovel, xfin_hivel,
448 xfout_lovel, xfout_hivel, xf_velcurve, xfin_loccN, xfin_hiccN,
449 xfout_loccN, xfout_hiccN, xf_cccurve
450 - implemented opcodes pan_onccN, pan_smoothccN,
451 pan_curveccN, egN_pan, egN_pan_curve, egN_pan_onccX,
452 egN_pan_curveccX, pitch_veltrack
453 - when failed to parse a sfz file
454 print the line number on which the error occurs
455 - use common pool of CC objects to minimize RAM usage
456 - implemented opcodes amplfo_delay_onccN,
457 amplfo_fade_onccN, fillfo_delay_onccN, fillfo_fade_onccN,
458 pitchlfo_delay_onccN, pitchlfo_fade_onccN
459 - implemented opcodes fileg_delay_onccN,
460 fileg_start_onccN, fileg_attack_onccN, fileg_hold_onccN,
461 fileg_decay_onccN, fileg_sustain_onccN, fileg_release_onccN,
462 fileg_depth_onccN, pitcheg_delay_onccN, pitcheg_start_onccN,
463 pitcheg_attack_onccN, pitcheg_hold_onccN, pitcheg_decay_onccN,
464 pitcheg_sustain_onccN, pitcheg_release_onccN, pitcheg_depth_onccN
465 - implemented automatic aliasing *ccN <-> *_onccN
466 - *lfo_freqccN wasn't working when the respective *lfo_freq
467 was not set or was set to zero
468 - lfoN_freq_onccX wasn't working when lfoN_freq
469 was not set or was set to zero
470 - implemented opcodes resonance_onccN, resonance_smoothccN,
471 resonance_curveccN, cutoff_smoothccN, cutoff_curveccN
472 - implemented opcodes fillfo_depthchanaft,
473 fillfo_freqchanaft, amplfo_depthchanaft, amplfo_freqchanaft,
474 pitchlfo_depthchanaft, pitchlfo_freqchanaft
475 - implemented opcodes pitch_onccN,
476 pitch_curveccN, pitch_smoothccN, pitch_stepccN
477 - implemented opcodes volume_stepccN, pan_stepccN,
478 cutoff_stepccN, resonance_stepccN, lfoN_freq_stepccX,
479 lfoN_volume_stepccX, lfoN_pitch_stepccX, lfoN_pan_stepccX,
480 lfoN_cutoff_stepccX, lfoN_resonance_stepccX
481 - implemented opcodes eq1_freq, eq2_freq, eq3_freq,
482 eq1_freqccN, eq2_freqccN, eq3_freqccN, eq1_bw, eq2_bw, eq3_bw,
483 eq1_bwccN, eq2_bwccN, eq3_bwccN, eq1_gain, eq2_gain, eq3_gain,
484 eq1_gainccN, eq2_gainccN, eq3_gainccN
485 - implemented opcodes delay, delay_onccN, delay_random,
486 delay_samples, delay_samples_onccN
487 - implemented opcodes egN_eq1gain, egN_eq2gain, egN_eq3gain,
488 egN_eq1gain_onccX, egN_eq2gain_onccX, egN_eq3gain_onccX, egN_eq1freq,
489 egN_eq2freq, egN_eq3freq, egN_eq1freq_onccX, egN_eq2freq_onccX,
490 egN_eq3freq_onccX, egN_eq1bw, egN_eq2bw, egN_eq3bw, egN_eq1bw_onccX,
491 egN_eq2bw_onccX, egN_eq3bw_onccX, lfoN_eq1gain, lfoN_eq2gain,
492 lfoN_eq3gain, lfoN_eq1gain_onccX, lfoN_eq2gain_onccX, lfoN_eq3gain_onccX,
493 lfoN_eq1gain_smoothccX, lfoN_eq2gain_smoothccX, lfoN_eq3gain_smoothccX,
494 lfoN_eq1gain_stepccX, lfoN_eq2gain_stepccX, lfoN_eq3gain_stepccX,
495 lfoN_eq1freq, lfoN_eq2freq, lfoN_eq3freq, lfoN_eq1freq_onccX,
496 lfoN_eq2freq_onccX, lfoN_eq3freq_onccX, lfoN_eq1freq_smoothccX,
497 lfoN_eq2freq_smoothccX, lfoN_eq3freq_smoothccX, lfoN_eq1freq_stepccX,
498 lfoN_eq2freq_stepccX, lfoN_eq3freq_stepccX, lfoN_eq1bw, lfoN_eq2bw,
499 lfoN_eq3bw, lfoN_eq1bw_onccX, lfoN_eq2bw_onccX, lfoN_eq3bw_onccX,
500 lfoN_eq1bw_smoothccX, lfoN_eq2bw_smoothccX, lfoN_eq3bw_smoothccX,
501 lfoN_eq1bw_stepccX, lfoN_eq2bw_stepccX, lfoN_eq3bw_stepccX
502 - implemented opcodes eq1_vel2freq, eq2_vel2freq,
503 eq3_vel2freq, eq1_vel2gain, eq2_vel2gain, eq3_vel2gain
504 - sfz parser: allow double spaces in sample filenames
505 - sfz parser: allow absolute paths for sample filenames
506 - use linear decay and release for filter and pitch EG
507 - bugfix: only the first amp_veltrack definition in a file was
508 used
509 - bugfix: looping was disabled if loop_start was set to 0
510 - allow regions with end=-1 to turn off other regions using the
511 group and off_by opcodes (#168)
512 - made end=0 play the whole sample
513 - fixed support for lochan and hichan opcodes (#155)
514 - fixed crash when using lochan/hichan opcodes (#187)
515 - sfz parser: allow -200 to 200 for pan_oncc opcode (#182)
516 - added FLAC support (#191)
517 - sfz parser bugfix: lines starting with whitespace were ignored
518 - added amplitude opcode
519 - added support for "#include" instruction
520 (modified patch which was originally posted by Sergey on LS mailing list)
521 - bugfix: generation of velocity curves etc should not be done
522 after each "#include", only after the main file is parsed
523 - bugfix: line numbers in error messages were wrong after "#include"
524 - added support for float and 32 bit sample files
525
526 * SoundFont format engine:
527 - Initial implementation (not usable yet)
528 - RT-safeness: avoid malloc in audio thread
529 - fixed a bug that could cause voice stealing to fail
530 - fine-tuned amplitude EG (by switching from gig to sfz EG)
531 - initial implementation of Vibrato LFO and Modulation LFO
532 - initial implementation of cutoff filter
533 - use linear decay and release for filter and pitch EG
534
535 * Host plugins (VST, AU, LV2, DSSI):
536 - AU bugfix: failed to destroy its audio/MIDI devices
537 - Listen to all interfaces on Mac OS X (INADDR_ANY)
538 - VST bugfix: If the host called resume() before and after
539 changing sample rate or block size, the number of channels was
540 incorrectly set to two. This caused silence in Cubase 5.
541 - save engine type (gig, sfz or sf2) in plugin state
542 - VST: when opening Fantasia, look for both 32 and 64 bit Java on
543 64 bit Windows
544 - AU: changed number of output channels from one stereo to 16
545 stereo
546 - VST: made it possible to build the VST plugin for Mac
547 - AU: link AU plugin dynamically if --disable-shared isn't
548 specified
549 - LV2 "state" extension support (patch by David Robillard)
550 - VST bugfix: instrument loading hang and crashed the host when
551 the plugin was loaded a second time (#174)
552 - plugin bugfix: instrument loading hang when the plugin was
553 loaded a second time (this time it's for Linux and Mac, previous
554 similar fix was for Windows)
555 - thread safety fixes for the instrument loading thread
556 - LV2: use the new lv2 package if present
557 - VST: try to open Fantasia automatically on Linux and Mac too (on
558 Linux, the Fantasia jar should be placed in <prefix>/share/java)
559 - VST: fixed crashes on Linux Ardour and EnergyXT
560 - DSSI bugfix: it wasn't possible to change engine type. The MIDI
561 port and audio channel routing for DSSI plugins are now visible.
562 - LV2: use urid and atom extensions instead of deprecated uri-map
563 and event
564 - LV2: lv2 package 1.0 is now required to build the LV2 plugin
565 - LV2: changed number of output channels to 16 stereo, just like
566 the VST and AU plugins
567 - LV2: fixed save/restore of SFZ state (patch by David Robillard)
568 - LV2: made LV2 plugin buildable on Windows and Mac
569 - VST: implemented retrieval and switching of programs using the
570 sampler's internal MIDI instrument mapping system
571
572 * Instrument editor interface:
573 - Changed instrument editor plugin interface, providing additional
574 informations like the EngineChannel for which the instrument editor was
575 spawned for. This allows the instrument editors to interact more actively
576 with the sampler.
577
578 * MIDI driver:
579 - ALSA MIDI driver supports now "NAME" device parameter, for overriding
580 the ALSA sequencer client name
581 - removed limit of maximum amount of MIDI ports per MIDI device, since
582 there is no reason for this limit
583 - MME: fixed memory handling bug found with cppcheck
584 - MME: removed compiler warning
585 - CoreMIDI: implemented driver specific port parameter "CORE_MIDI_BINDINGS",
586 which allows to retrieve the list of CoreMIDI clients / ports and to connect
587 to them a la JACK, via the usual sampler APIs
588 - CoreMIDI: added driver specific port parameter "AUTO_BIND", if enabled
589 the driver will automatically connect to other CoreMIDI clients' ports
590 (e.g. external MIDI devices being attached to the Mac)
591 - added support for MIDI note on velocity filter
592 - CoreMIDI: fixed memory deallocation error
593 - Fixed variable underflow in VirtualMidiDevice, which caused graphical
594 virtual keyboards in frontends / instrument editors being stuck.
595 - Bugfix in VirtualMidiDevice: process note on with velocity 0 as note off.
596 - Implemented missing handling of MIDI "running status".
597 - CoreMIDI fix: a MIDIPacket can contain more than one event per packet.
598 - MME bugfix: driver wasn't closed properly
599 - CoreMIDI: automatically connect to all input sources by default (driver
600 parameter "AUTO_BIND").
601 - CoreMIDI: fixed auto bind feature to CoreMIDI ports that go online
602 - CoreMIDI: fixed minor memory leak
603
604 * audio driver:
605 - ASIO driver fixes for newer gcc versions (fix from PortAudio)
606 - JACK audio: react on sample rate changes.
607 - JACK audio: react on buffer size changes.
608 - JACK audio: jack_port_get_buffer() was cached and called outside
609 RT context.
610 - ASIO driver: removed compiler warnings
611 - CoreAudio: fixed minor error handling bug
612 - ASIO driver: be more verbose when no ASIO card could be found (fixes #203)
613 - JACK audio: return the JACK server's current sample rate as default value
614 for audio device parameter "SAMPLERATE" (fixes #166).
615
616 * LSCP server:
617 - added support for sending MIDI CC messages via LSCP command
618 "SEND CHANNEL MIDI_DATA CC <sampler-chan> <ctrl> <val>"
619 - added LSCP command "GET AVAILABLE_EFFECTS"
620 - added LSCP command "LIST AVAILABLE_EFFECTS"
621 - added LSCP command "GET EFFECT INFO <effect-index>"
622 - added LSCP command "CREATE EFFECT_INSTANCE <effect-index>"
623 - added LSCP command
624 "CREATE EFFECT_INSTANCE <effect-system> <module> <effect-name>"
625 - added LSCP command "DESTROY EFFECT_INSTANCE <effect-instance>"
626 - added LSCP command "GET EFFECT_INSTANCES"
627 - added LSCP command "LIST EFFECT_INSTANCES"
628 - added LSCP command "GET EFFECT_INSTANCE INFO <effect-instance>"
629 - added LSCP command
630 "GET EFFECT_INSTANCE_INPUT_CONTROL INFO <effect-instance> <input-control>"
631 - added LSCP command "SET EFFECT_INSTANCE_INPUT_CONTROL VALUE
632 <effect-instance> <input-control> <value>"
633 - added LSCP command "GET SEND_EFFECT_CHAINS <audio-device>"
634 - added LSCP command "LIST SEND_EFFECT_CHAINS <audio-device>"
635 - added LSCP command "ADD SEND_EFFECT_CHAIN <audio-device>"
636 - added LSCP command
637 "REMOVE SEND_EFFECT_CHAIN <audio-device> <effect-chain>"
638 - added LSCP command
639 "GET SEND_EFFECT_CHAIN INFO <audio-device> <effect-chain>"
640 - added LSCP command "APPEND SEND_EFFECT_CHAIN EFFECT <audio-device>
641 <effect-chain> <effect-instance>"
642 - added LSCP command "INSERT SEND_EFFECT_CHAIN EFFECT <audio-device>
643 <effect-chain> <effect-chain-pos> <effect-instance>"
644 - added LSCP command "REMOVE SEND_EFFECT_CHAIN EFFECT <audio-device>
645 <effect-chain> <chain-pos>"
646 - added LSCP command "SET FX_SEND EFFECT <sampler_channel>
647 <fx_send_id> <effect_chain> <chain_pos>"
648 - added LSCP command "REMOVE FX_SEND EFFECT <sampler_channel> <fx_send_id>"
649 - added LSCP commands "SUBSCRIBE EFFECT_INSTANCE_COUNT",
650 "SUBSCRIBE EFFECT_INSTANCE_INFO", "SUBSCRIBE SEND_EFFECT_CHAIN_COUNT",
651 "SUBSCRIBE SEND_EFFECT_CHAIN_INFO"
652 - provide comprehensive error messages on LSCP syntax errors
653 (suggesting expected next non-terminal symbols)
654 - Fixed client connection not being closed after network errors.
655
656 * LSCP shell:
657 - Added support for auto correction of obvious and trivial syntax mistakes.
658 - Added support for auto completion by tab key.
659 - Show currently available auto completion while typing.
660 - Added support for browsing command history with up / down keys.
661 - Show all possible next symbols immediately right to the current command
662 line while typing (no double tab required for this feature, as it would
663 be the case in other shells).
664 - Added support for moving cursor left/right with arrow keys.
665 - Added support for built-in LSCP reference documentation, which will
666 automatically show the relevant LSCP reference section on screen as soon
667 as one specific LSCP command was detected while typing on the command
668 line.
669
670 * Real-time instrument scripts:
671 - Implemented built-in script array variable %CC.
672 - Implemented built-in script int variable $CC_NUM.
673 - Implemented built-in script int variable $EVENT_NOTE.
674 - Implemented built-in script int variable $EVENT_VELOCITY.
675 - Implemented built-in script constant variable $VCC_MONO_AT.
676 - Implemented built-in script constant variable $VCC_PITCH_BEND.
677 - Implemented execution of script event handler "init".
678 - Implemented execution of script event handler "controller".
679 - Implemented execution of script event handler "note".
680 - Implemented execution of script event handler "release".
681 - Implemented built-in script function "play_note()" (only two of the
682 max. four function arguments are currently implemented yet though).
683 - Implemented built-in script int variable $EVENT_ID.
684 - Implemented built-in script function "ignore_event()"
685 - Implemented built-in script function "ignore_controller()" (may have one
686 or no argument).
687 - Implemented built-in script function "set_controller()".
688 - Added extended script VM for the Gigasampler/GigaStudio format sampler
689 engine, which extends the general instrument script VM with Giga format
690 specific variables and functions.
691 - Giga format scripts: added built-in script int constant variables
692 $GIG_DIM_CHANNEL, $GIG_DIM_LAYER, $GIG_DIM_VELOCITY, $GIG_DIM_AFTERTOUCH,
693 $GIG_DIM_RELEASE, $GIG_DIM_KEYBOARD, $GIG_DIM_ROUNDROBIN, $GIG_DIM_RANDOM,
694 $GIG_DIM_SMARTMIDI, $GIG_DIM_ROUNDROBINKEY, $GIG_DIM_MODWHEEL,
695 $GIG_DIM_BREATH, $GIG_DIM_FOOT, $GIG_DIM_PORTAMENTOTIME, $GIG_DIM_EFFECT1,
696 $GIG_DIM_EFFECT2, $GIG_DIM_GENPURPOSE1, $GIG_DIM_GENPURPOSE2,
697 $GIG_DIM_GENPURPOSE3, $GIG_DIM_GENPURPOSE4, $GIG_DIM_SUSTAIN,
698 $GIG_DIM_PORTAMENTO, $GIG_DIM_SOSTENUTO, $GIG_DIM_SOFT,
699 $GIG_DIM_GENPURPOSE5, $GIG_DIM_GENPURPOSE6, $GIG_DIM_GENPURPOSE7,
700 $GIG_DIM_GENPURPOSE8, $GIG_DIM_EFFECT1DEPTH, $GIG_DIM_EFFECT2DEPTH,
701 $GIG_DIM_EFFECT3DEPTH, $GIG_DIM_EFFECT4DEPTH, $GIG_DIM_EFFECT5DEPTH.
702 - Giga format scripts: Implemented built-in script function
703 "gig_set_dim_zone(event_id, dimension, zone)".
704 - Implemented built-in script int array variable %KEY_DOWN.
705 - Implemented built-in script function "abs()".
706 - Implemented built-in script function "random()".
707 - Implemented built-in script function "num_elements()".
708 - Implemented built-in script function "note_off()".
709 - Implemented built-in script function "set_event_mark()".
710 - Implemented built-in script function "delete_event_mark()".
711 - Implemented built-in script function "by_marks()".
712 - Added built-in script int const variables $MARK_1 to $MARK_28.
713 - Built-in script functions "ignore_event()", "note_off()" and
714 "gig_set_dim_zone()" now also accept an array of event IDs as argument
715 (i.e. return value of new script function "by_marks()").
716 - Pass/preserve polyphonic variable data from respective "note" event
717 handler to "release" event handler.
718 - Fixed crash when using built-in script function "by_marks()".
719
720 * Bug fixes:
721 - Fixed crash which may occur when MIDI key + transpose is out of range
722 - minor valgrind fixes
723 - fixed crash which occurred when changing an already deployed sampler
724 channel to a different engine type
725 - fixed crash when deleting a sampler channel or changing engine
726 type while an instrument load was in progress
727 - bugfix: playing a note while changing the instrument could cause
728 a crash, or give "Handing back unknown region" error messages
729 - bugfix: calling SET VOICES (which Fantasia does on start and
730 refresh) could cause instruments to be unloaded
731 - Mac OS X: fixed name collision of enum in EffectControl and
732 wrong name of destructor in AudioOutputDeviceCoreAudio.cpp
733 - Mac OS X: fixed hanging threads
734 - Windows: disabled the previous thread fix on non-Mac systems, as
735 it caused hanging threads on Windows
736 - Fixed possible crashes due to corrupted MIDI/audio device list
737 after MIDI/audio device creation failure
738 - When creating MIDI instrument map entries with "PERSISTENT" type, the
739 instruments were uselessly precached with zero samples, however it still
740 took the full preloading time and on 1st program change the respective
741 instrument was completely reloaded again.
742 - fixed handling of rapid bank select and program change messages
743 sent to the same sampler channel (patch from the Open Octave
744 project, slightly adjusted)
745 - fixed crash when trying to create an effect instance with controls
746 which min and/or max values depend on the sample rate
747 - fixed bug #162
748 - bugfix: LADSPA_PATH was not evaluated correctly when containing
749 multiple paths (#165)
750 - thread safety fixes for the instrument loading thread
751 - bugfix: instrument loading crashed for sfz and sf2 in Ardour
752 (#176)
753 - more thread safety fixes for the instrument loading thread
754 - sfz/sf2 engine: fixed crash when using small audio fragment size
755 - Mac OS X: fixed crash when unloading plugin on 10.7 and later
756 - Mac OS X: fixed process hang when unloading 32-bit plugin (bug
757 introduced in previous fix)
758 - fixed crash when a channel received a program change while
759 playing a note in a key group
760 - fixed erroneous error message piping in VoiceBase.h
761 (labelled "Disk stream not available in time")
762 - Update effects on sample rate & period size changes (to avoid
763 crashes and noise on such transitions).
764 - ignore missing LADSPA paths without ignoring valid LADSPA paths
765 (fixes #208)
766 - Fixed bug in LSCP grammar definition which caused a statement like
767 "GET SERVER INFOasdf\n" to be accepted as valid statement (was so far
768 practically irrelevant, however it caused problems with the new LSCP
769 shell's auto completion feature).
770 - Fixed MIDI program change messages being ignored if quickly executed
771 after each other (fixes #231).
772
773 Version 1.0.0 (31 July 2009)
774
775 * packaging changes:
776 - autoconf bugfix: the PKG_CONFIG variable wasn't initialized properly,
777 causing e.g. the libgig test to fail when
778 "./configure --disable-jack-driver" was used
779 (patch by Alexis Ballier)
780 - fixed compilation with gcc 4.3
781 - fixes for building on OS X (thanks to Ebrahim Mayat for testing)
782 - fixed configure so it detects x86_64 (#107)
783 - fixes for building with newer MinGW versions
784 - fix for building with bison 2.4 (#111)
785 - fixed building with libgig installed in a non-standard directory
786 - minor fix in configure for mmsystem.h detection on MinGW
787 - Windows: look for editor plugins and Fantasia using base
788 directory of liblinuxsampler dll (look in the same directory and one
789 directory above)
790 - configure script fix: removed unconditional use of SSE
791 - fixed building with sqlite installed in a non-standard directory
792 - when cross-compiling, don't try to create instruments.db
793 - fix for new mingw-w64 version, which has usleep
794
795 * general changes:
796 - bugfix: on some POSIX systems instrument editor plugins refused to
797 load as we used a non-portable Linux specific struct field
798 (fixes bug #70, patch by Ronald Baljeu)
799 - fixed endless loop which occured when loading an instrument editor
800 plugin DLL on Windows systems failed
801 - fixed memory leaks that occurred when liblinuxsampler was unloaded
802 - optimized the SynchronizedConfig class so it doesn't wait
803 unnecessarily long after an update
804 - added support for notifying instrument editors on note-on / note-off
805 events (e.g. to highlight the pressed keys on the virtual keyboard
806 of gigedit)
807 - added support for triggering notes by instrument editors (see above)
808 - be verbose on DLL load errors (on Linux)
809 - fixed di-harmonic triangle LFO implementation (this LFO implementation
810 is only used by very few systems, most pick the int math
811 implementation, which is usually faster)
812 - fixes for audio drivers with varying buffer sizes
813 - experimental support for running LinuxSampler as a DSSI, LV2 and
814 VST plugin
815 - notification events for stream/voice count statistics are now sent
816 only when there are actual changes
817 - added memory ordering constraints to improve stability on
818 multi-core and multi-cpu systems
819 - maximum voices and disk streams can now be altered at runtime
820 - fixed CPU feature detection on x86_64 (maybe fixes #108)
821 - automatic stacktrace mechanism is now turned off by default and can
822 be switched on by command line option "--stacktrace" (the automatic
823 stacktrace mechanism seems to be broken on most systems at the moment)
824 - C++ API method InstrumentManager::LaunchInstrumentEditor() now returns
825 a pointer to the launched InstrumentEditor object
826 - added optional 3rd party user data parameter for following
827 liblinuxsampler C++ API methods: InstrumentEditor::Main(),
828 InstrumentEditor::Launch(),
829 InstrumentManager::LaunchInstrumentEditor()
830 - theoretical fix: made SynchronizedConfig follow C++0x memory
831 model more strictly
832 - fixes for using large audio device buffers
833 - Windows: add the installation directory to the DLL search path
834 when loading an editor plugin (solves problems with VST and
835 gigedit on systems with other GTK versions installed)
836 - updated linuxsampler man page
837
838 * audio driver:
839 - removed the nonsense audio channel constraint (which was hard coded to
840 max. 100 audio channels) for most audio drivers
841 - JACK audio driver did not offer a device parameter "SAMPLERATE" as
842 opposed to the LSCP specs
843 - bugfix: the SAMPLERATE parameter of some drivers (e.g. JACK)
844 reflected the wrong value
845 - fixed a memory management bug in ASIO driver
846 - Makefile fix: JACK_CFLAGS wasn't used
847 - JACK: use jack_client_open instead of the deprecated
848 jack_client_new
849 - added (experimental) CoreAudio driver
850 - applied old fixes to the ASIO driver that were included in the
851 previous binary release but accidentally never committed to CVS
852 (fixes #117)
853 - fixes for ASIO on mingw-w64 (iasio wrapper is not needed on
854 win64)
855 - VST: added support for sample rate and buffer size changes
856 - VST: close editor (Fantasia) when the VST is removed
857 - VST: avoid opening Fantasia more than once for each VST instance
858 - VST: export main function as "main" on Linux too (fix for
859 energyXT)
860 - VST: changed number of output channels from one stereo to 16
861 stereo
862 - added channel routing, fxsends and midi maps to the settings
863 stored in the plugin state
864 - performance optimization of AudioChannel::MixTo() and
865 AudioChannel::CopyTo() methods using GCC vector exensions
866 (if available)
867 - ASIO fixes: avoid initializing the device twice, avoid throwing
868 exception when getting parameters from a disconnected device
869
870 * MIDI driver:
871 - added JACK MIDI driver
872 - dispatch bank select as ordinary CC as well, the user might seriously
873 want to (mis)use it for some purpose ("fixed" in all current MIDI
874 input drivers: ALSA, CoreMIDI, JACK, MidiShare, MME)
875 - bugfix: pitch bend wasn't working with JackMidi, VST, LV2, MME,
876 CoreMidi or AU
877 - fixed mingw-w64 compilation error in MME driver
878 - made program change handling in MIDI thread real-time safe by
879 moving the logic to a non-RT thread
880 - fixed minor memory leak in ALSA driver
881
882 * instruments database:
883 - avoid time consuming samples scanning when adding instruments
884 to the instruments database
885 - added support for handling lost files in the instruments database
886 - Implemented option for adding instruments in separate directories
887 in the instruments database
888 (patch by Chris Cherrett & Andrew Williams, a bit adjusted)
889 - work-around for missing fnmatch function on Windows to make
890 instrument database compilable
891 - added instrument database support on Windows
892 works with both standalone mode and VST plugin.
893 the instruments DB file is located in
894 %USERPROFILE%\.linuxsampler\instruments.db which allows different
895 databases for each windows user
896 if no DB is present it automatically creates the .linuxsampler subdir
897 and then creates an empty DB
898 - fixed recursive import, which was broken on Windows
899
900 * Gigasampler format engine:
901 - fixed a memory leak that could happen when a channel was deleted
902 while notes were playing
903 - made it possible to load an instrument even if the audio thread
904 isn't running
905 - added partial support for the "Controller Triggered" MIDI rule,
906 enough for piano gigs with pedal noise samples
907 - minor fix: only mark FX sends as being modified if really the
908 respective FX send MIDI controller was used
909 - added support for GM global device volume SysEx message
910 (can be selected at compile time to either apply globally to the
911 whole sampler [default] or only to the sampler channels that are
912 connected to the respective MIDI input port the SysEx message arrived
913 on)
914 - bugfix: notes triggered at position 0 in the audio buffer were
915 sometimes wrongly killed in the same buffer, causing no sound to
916 be played
917 - added support for chromatic / drumkit mode Roland GS Sysex message
918 which is usally used to switch a part between chromatic sounds and
919 drumkit sounds (as opposed to their common default setting of MIDI
920 part 10 being a drumkit part and all other ones chromatic parts), in
921 LS however one can switch between the first 16 MIDI instrument maps
922 defined for the sampler
923 - made it possible to create multiple sample channels even if the
924 audio thread isn't running
925 - fixed crash when removing channel with active voices (#116)
926 - bugfix: on sample reference changes (instrument editor), only
927 un-cache the respective sample if it's really not used by any
928 sampler engine anymore
929 - re-cache samples in case they were changed by an instrument editor,
930 e.g. when a sample was added while playing (#82)
931 - fixed hanging notes which occured when note-off event had the exact
932 same time stamp as the voice's note-on event and occured both in the
933 same audio fragment cycle (fixes bug #112)
934 - added support for the "fine tune" and "pitch bend range"
935 instrument-level gig parameters
936 - fixed minor artifacts in pitch bend handling
937 - added support for GS Reset SysEx message
938 - allow gig files to use unlimited downward pitch shifting
939 - added a limit check for upward pitch shifting
940 - bugfix: sometimes, when playing a note twice fast, the second
941 note was silent
942 - fixed crash happening when a pitch bend event arrived at the
943 same time a new instrument was loading
944
945 * LSCP server:
946 - added new LSCP event "CHANNEL_MIDI" which can be used by frontends to
947 react on MIDI data arriving on certain sampler channels (so far only
948 Note-On and Note-Off events are sent via this LSCP event)
949 - added new LSCP event "DEVICE_MIDI" which can be used by frontends to
950 react on MIDI data arriving on certain MIDI input devices (so far only
951 Note-On and Note-Off events are sent via this LSCP event)
952 - added new LSCP commands: FIND LOST DB_INSTRUMENT_FILES and
953 SET DB_INSTRUMENT FILE_PATH
954 - added new LSCP command: SEND CHANNEL MIDI_DATA which can be used by
955 frontends to send MIDI messages to specific sampler channel
956 - added two additional fields to GET FILE INSTRUMENT INFO command -
957 KEY_BINDINGS and KEYSWITCH_BINDINGS
958 - bugfix: the bank number provided by MIDI_INSTRUMENT_INFO notifications
959 was incorrect
960 - Optimized the retrieval of the MIDI instrument mappings
961 - added new LSCP command "SET VOICES" to globally alter the maximum
962 amount of voices, added new LSCP event "GLOBAL_INFO:VOICES" which will
963 be triggered respectively
964 - added new LSCP command "SET STREAMS" to globally alter the maximum
965 amount of disk streams, added new LSCP event "GLOBAL_INFO:STREAMS"
966 which will be triggered respectively
967 - bugfix: retry if "select" returns EINTR (this fixes a crash when
968 a gigedit file dialog is opened)
969 - close all connections when LSCPServer is deleted
970 - hard close of all sockets on exit
971 - bugfix: SET CHANNEL MIDI_INPUT_TYPE didn't work with the MME
972 driver
973
974 * Bug fixes:
975 - fixed a crash which occurs when removing a sampler channel waiting
976 to start instrument loading after another channel
977 - fixed a crash which occurs when removing a sampler channel with
978 instrument loading in progress (bug #113)
979 - fixed termination caused by uncaught exception when adding MIDI
980 instrument with PERSISTENT load mode
981 - fixed possible iterator invalidations when resetting the sampler
982 - fixed memory leaks when issuing the following LSCP commands:
983 GET AUDIO_OUTPUT_DRIVER INFO
984 GET AUDIO_OUTPUT_DRIVER_PARAMETER INFO
985 GET MIDI_INPUT_DRIVER INFO
986 GET MIDI_INPUT_DRIVER_PARAMETER INFO
987 - fixed possible compilation error when sqlite is not present
988 - fixed orphaned pointers when setting maximum voices limit (bug #118)
989 - fixed crash when changing the audio output device of a sampler
990 channel with loaded instrument and start playing notes
991 - fixed endless loop in Engine::SuspendAll() (bug #120)
992 - fixed a low-level atomic load function that was broken on 64-bit
993 PowerPC, which probably could cause crashes on that platform
994 - fixed a memory management error which could cause a crash when a
995 plugin was unloaded
996 - bugfix: two private structs had the same name, which could cause
997 problems if the linker chose the wrong constructor
998 - fixed low-level ConditionServer usage bug that caused lockups on
999 Windows
1000
1001
1002 Version 0.5.1 (6 December 2007)
1003
1004 * packaging changes:
1005 - added autoconf checks for pthread library
1006 - added autoconf check for pthread bug found on certain NPTL-enabled
1007 glibc versions (see Gentoo bug report #194076)
1008 - added autoconf checks for MS Windows
1009
1010 * general changes:
1011 - bugfix: the thread used by an editor plugin didn't die when the
1012 editor closed
1013 - bugfix: Ringbuffer.h: fill_write_space_with_null() did not zero
1014 out all the space. operator--() did not apply size_mask after
1015 decrementing the read_ptr. DEFAULT_WRAP_ELEMENTS set to 0 in
1016 order to avoid problems with the _NonVolatileReader functions.
1017 - bugfix: Stream.h: added a missing divide by BytesPerSample in
1018 GetWriteSpace(). Since this function is currently only used in
1019 the stream's qsort() compare function, it didn't trigger any
1020 bugs.
1021 - Resampler.h, Synthesizer.h: cubic interpolation now works in
1022 24bit mode too. Faster method to read 24bit words on little
1023 endian machines (x86): replaced 3 byte reads + shifts with a 1
1024 unaligned 32bit read and shift
1025 - experimental support for MS Windows (MIDI input via MME, AUDIO
1026 output via ASIO)
1027 - made handling of SIGINT signal (Ctrl-C) a bit more robust
1028 - support for monitoring the total number of active disk streams
1029 (new LSCP commands: GET TOTAL_STREAM_COUNT,
1030 SUBSCRIBE TOTAL_STREAM_COUNT, UNSUBSCRIBE TOTAL_STREAM_COUNT)
1031
1032 * AUDIO driver:
1033 - added Windows ASIO low latency audio driver
1034
1035 * MIDI driver:
1036 - added MME Windows MIDI driver
1037
1038 * LSCP server:
1039 - added support for Windows style path / filenames, however with
1040 forward slash path separators instead of backslash
1041 (i.e. "C:/foo/bar.gig")
1042 - allow naughty liblscp to send non-string device parameters within
1043 apostrophes as well
1044 - added new LSCP commands: "GET FILE INSTRUMENTS <file>",
1045 "LIST FILE INSTRUMENTS <file>" and
1046 "GET FILE INSTRUMENT INFO <file> <index>" for retrieving informations
1047 about an arbitrary instrument file on the system where the sampler is
1048 running on
1049
1050 Version 0.5.0 (15 October 2007)
1051
1052 * packaging changes:
1053 - config.h is not going to be installed along with liblinuxsampler's
1054 API header files anymore
1055 - only the API relevant header (and source) files will be exported to
1056 the Doxygen API documentation (explicitly listed in Doxyfile.in)
1057 - added completely new XCode project files for Mac OSX which is now
1058 capable to execute our autoconf environment, thus no need anymore
1059 to maintain the compile time configuration file (osx/version.h) for
1060 OSX manually (patch by Toshi Nagata)
1061 - fixed buggy boolean --enable-foo / --disable-foo configure script
1062 parameters
1063 - global.h now only covers global definitions that are needed for the
1064 C++ API header files, all implementation internal global definitions
1065 are now in global_private.h
1066 - atomic.h is not exposed to the C++ API anymore
1067 - no need to include config.h anymore for using LS's API header files
1068 - fixed warnings in API doc generation
1069
1070 * general changes:
1071 - replaced the old, confusing MIDI program change mechanism by a
1072 flexible MIDI instrument mapper which allows to map arbitrary
1073 (MIDI bank, MIDI program) pairs with arbitrary
1074 (engine type, instrument file, file index) triplets which will be
1075 loaded on the respective channel when such MIDI program change
1076 messages arrive, beside that, each entry allows to define a life-time
1077 strategy for the instrument, so the user can define whether the
1078 instrument should i.e. be permanently loaded (to be able to switch
1079 quickly among sounds) or i.e. loaded just on demand when the
1080 respective program change arrives (to safe RAM space), as well as a
1081 global volume factor for each entry, so the user can adjust the volume
1082 dynamic of his mapped instrument collection without having to modify
1083 the instrument files, also one can manage arbitrary amount of such
1084 MIDI instrument maps and assign each sampler channel individually
1085 a certain map, so that i.e. separation of normal instruments and
1086 drumkits is possible
1087 - new notification events for tracking changes to audio/MIDI devices,
1088 MIDI instrument maps, MIDI instruments, FX sends, global volume.
1089 - sampler was limited to load max. 200 instruments in the background
1090 due to a constant size RingBuffer FIFO which is now replaced by a
1091 dynamic (unlimited) size std::list FIFO
1092 - added FX sends, these allow to route audio signals to arbitrary audio
1093 output channels for being processed by external effect processors
1094 (i.e. jack-rack), the send levels are controllable via arbitrary MIDI
1095 controllers
1096 - global (sampler wide) volume can now be controlled at runtime
1097 - Implemented new, improved notification system
1098 - fixed compilation errors regarding OSX
1099 (patch by Toshi Nagata)
1100 - implemented instruments database
1101 - added support for escape sequences to the instruments database
1102 - added highly experimental support for on-the-fly instrument editing
1103 within the sampler's process (by using instrument editor plugins),
1104 you'll notice the new "Registered instrument editors:" message on
1105 startup, added a new LSCP command:
1106 "EDIT CHANNEL INSTRUMENT <sampler-channel>"
1107 to spawn a matching instrument editor for the instrument on the
1108 given sampler channel, the plugin path can be overridden at compile
1109 time with ./configure --enable-plugin-dir=/some/dir
1110 - added experimental code for synchronizing instrument editors hosted
1111 in the sampler's process to safely edit instruments while playing
1112 without a crash (hopefully) by either suspending single regions
1113 wherever possible or if unavoidable whole engine(s)
1114 - fixed several issues in fundamental "Thread" class: set scheduling
1115 policy and priority on thread level, set a minimum stack size for
1116 thread (TODO: a reasonable value yet to be tested), bugfix: non-RT
1117 threads simply inherited properties of starting thread instead of
1118 setting their own policy and priority
1119 - minor fix in our automatic stack trace mechanism on crashes, the
1120 main process did not wait for the stack trace process to finish
1121 its output
1122 - fixed some minor memory leaks
1123 - reenabled assembly features support, at the moment only for
1124 enabling a fast denormal FPU mode (x86 platforms supporting SSE2)
1125 - minor assembly fix in x86 features detection (don't use the PIC
1126 register, to avoid relocations in the text segment at runtime)
1127 - POSIX callback functions of Thread.h are hidden
1128
1129 * MIDI driver:
1130 - dispatch bank select (MSB and LSB) messages
1131
1132 * audio driver:
1133 - the ALSA audio output driver parameters now reflect the correct
1134 parameter value ranges for the respective selected sound card
1135 (patch by Till Wimmer, a bit fixed and extended)
1136
1137 * Gigasampler format engine:
1138 - if a filter is used and EG2 finishes before EG1, let the voice
1139 die when EG2 has finished (this fixes a problem with clicks and
1140 voice starvation for some gigs)
1141 - playback is no longer disabled during instrument loading
1142 - all notes playing on a channel that changes its instrument keep
1143 playing with the old instrument until they get a note off
1144 command
1145 - EG fix: a release value of zero could cause noises or crash
1146 - handle MIDI coarse tuning messages (MIDI RPN #0 MSB #2 LSB)
1147 - EG fine tuning: when attack is zero the EG starts at a level
1148 above max sustain level, which means that there is a decay phase
1149 even if sustain is 100%
1150 - more EG fixes: the level could sometimes go below zero and cause
1151 noises or crashes
1152 - minor fix of EGDecay (patch by Toshi Nagata)
1153 - fixed compiler error when --enable-override-filter-type was
1154 supplied to the configure script (fixes #46)
1155 - disk thread: queue sizes are now proportional to CONFIG_MAX_STREAMS
1156 instead of a fix value
1157 - behavior fix: on MIDI CC# 65 (portamento on / off), 126 (mono mode),
1158 127 (solo mode) only kill voices if the respective mode really
1159 changed
1160
1161 * LSCP server:
1162 - fixed compile time error for old Bison versions
1163 (i.e. v1.28 found on MacOS 10.4, patch by Toshi Nagata)
1164 - parser now supports extended ASCII character set
1165 (up to ASCII code 255, i.e. includes now umlauts and accents)
1166 - filename arguments in LSCP commands now allow to use escape
1167 sequences, that is directly literal as one of: \', \", \\, \n, \r,
1168 \f, \t, \v, or as octal ASCII code value like \132, or as hex ASCII
1169 code value like \xf2) (fixes bug #24)
1170 - the following LSCP commands now also support escape sequences for at
1171 least one of their text-based parameters (i.e. name, description):
1172 "ADD MIDI_INSTRUMENT_MAP", "MAP MIDI_INSTRUMENT",
1173 "SET MIDI_INSTRUMENT_MAP NAME", "SET FX_SEND NAME", "CREATE FX_SEND",
1174 "SET DB_INSTRUMENT_DIRECTORY NAME",
1175 "SET DB_INSTRUMENT_DIRECTORY DESCRIPTION", "SET DB_INSTRUMENT NAME",
1176 "SET DB_INSTRUMENT DESCRIPTION", "FIND DB_INSTRUMENTS",
1177 "FIND DB_INSTRUMENT_DIRECTORIES"
1178 - returns verbose syntax errors (line and column where syntax error
1179 occured, the unexpected character and the actually expected, possible
1180 character(s), the latter only if less than 5 possibilities)
1181 - made sure that LSCP syntax is not affected by gigedit locale
1182 settings
1183 - bugfix regarding strings parameter lists: all comma separated lists
1184 of strings were treated as being one string containing commas
1185 (fixes #57)
1186
1187 * Bug fixes:
1188 - fixed crash occurring on certain LSCP scripts (Bug 39)
1189 - another thread safety fix for lscp "load engine" and "set
1190 channel audio output device"
1191 - fixed a crash which occurs when reassigning the same engine
1192 on a sampler channel with connected MIDI device
1193 - fixed a crash which occurs when changing the number of ports of a MIDI
1194 device connected to a sampler channel to number less then or equal
1195 to the index of the port to which the sampler channel is connected.
1196 - The previous bindings were not been disconnected when altering
1197 the ALSA_SEQ_BINDINGS parameter. Introduced a NONE keyword for
1198 unsubscribing from all bindings (e.g. ALSA_SEQ_BINDINGS=NONE).
1199 - The active stream/voice count statistic was incorrect.
1200 - notification events were not been sent for some sampler
1201 channel changes
1202 - added default min and max values to restrict the number of allowed
1203 audio output channels and MIDI input ports
1204 - the connection to the PCM interface is now closed when destroying
1205 an audio output device
1206 - files with slash in their path or filename could not be loaded
1207
1208 * test cases:
1209 - updated and fixed (haven't been touched in a while)
1210
1211 Version 0.4.0 (24 November 2006)
1212
1213 * packaging changes:
1214 - changed deprecated copyright attribute to license; added ldconfig
1215 to post-(un)install steps; added devel package for liblinuxsampler;
1216 to linuxsampler.spec (RPM)
1217 - install necessary development header files for allowing 3rd party
1218 applications to link against liblinuxsampler
1219 - liblinuxsampler's API documentation can be generated with 'make docs'
1220 (Doxygen required)
1221 - added benchmark to automatically detect the best triangle LFO
1222 implementation (currently either an integer math solution or a
1223 di-harmonic approximation), automatic detection can be overriden
1224 with --enable-signed-triang-algo=x and --enable-unsigned-triang-algo=x
1225 configure script argument though (mandatory for cross-compilation)
1226 - do not automatically pick optimized gcc flags if the user already
1227 provided some on his own (as CXXFLAGS)
1228 - added compile time option to disable processing of All-Notes-Off MIDI
1229 messages
1230 - added compile time options to allow disabling the various audio and
1231 MIDI drivers
1232 - fixed automatic GCC CPU switch detection on PPC
1233 (patch by Ebrahim Mayat)
1234
1235 * Gigasampler format engine:
1236 - extensive synthesis optimization
1237 (reimplementation of EGs and LFO(s), removed synthesis parameter
1238 prerendering and the synthesis parameter matrix in general, splitting
1239 each audio fragment into subfragments now where each subfragment uses
1240 constant pitch and filter coefficients. The volume coefficient is
1241 linearly interpolated inside a subfragment, unless
1242 --disable-interpolate-volume is set.)
1243 - fine tuning of the EG modulation parameters
1244 - improved filter cutoff calculation by adding support for the
1245 following gig parameters: Cutoff freq (used when no cutoff
1246 controller is defined), Control invert, Minimum cutoff, Velocity
1247 curve and Velocity range. The keyboard tracking now scales
1248 cutoff frequency instead of resonance.
1249 - added support for gig parameter Resonance.
1250 - fixed bug in sysex handling (patch by Juan Linietsky)
1251 - added global volume attenuation of -9 dB (0.35f) to prevent clipping
1252 which can be overridden with --enable-global-attenuation
1253 - EG fixes: made the length of "attack hold" stage more
1254 accurate. Release stage can now start before attack stage
1255 ends. Cancel release didn't work when sustain was zero. Attack
1256 time now has a minimal value to prevent clicks.
1257 - fixed pitch changes larger than one octave
1258 - fixed EG3 (pitch envelope) synthesis which was neutral all the time
1259 - implemented portamento mode and solo mode (a.k.a 'mono mode'):
1260 all modes can be altered via standard GM messages, that is CC5 for
1261 altering portamento time, CC65 for enabling / disabling portamento
1262 mode, CC126 for enabling solo mode and CC127 for disabling solo mode
1263 - fine tuning of the curves for volume (CC7), pan (CC10 and gig
1264 parameter) and crossfade
1265 - added support for the "attenuation controller threshold" gig
1266 parameter
1267 - added smoothing of volume changes caused by control change
1268 messages
1269 - sample loop parameters are now taken from the DimensionRegion
1270 instead of the wave chunk
1271 - fixed keyswitching for v3 gigs with a number of keyswitch splits
1272 not equal to a power of two
1273 - reimplementation of the filter algorithm. The new filter is more
1274 accurate and supports all gig filter types, including bandreject
1275 and lowpass turbo.
1276 - real support for 24 bit samples - samples are not truncated to
1277 16 bits anymore
1278 - support for aftertouch (channel pressure, not polyphonic
1279 aftertouch)
1280 - LFO1 behaviour fixed (dampening from max volume instead of
1281 amplifying from 0)
1282
1283 * LSCP server:
1284 - fixed application exit on broken pipe error (fixes bug #20)
1285 - fixed the notification messages delay due to lack of
1286 network activity (fixes bug #26)
1287 - fixed parser bug which occured on space(s) within device parameters
1288 (e.g. "SET AUDIO_OUTPUT_CHANNEL_PARAMETER 0 0 SOMEPARAM='foo bar'")
1289
1290 * audio driver:
1291 - added aRts audio output driver (by no means RT safe)
1292
1293 * MIDI driver:
1294 - fixed legacy sysex code which caused dispatching of MIDI SysEx
1295 messages several times instead of once
1296 - API extension for MIDI drivers which already supply exact time stamps
1297 for events (i.e. for offline rendering based MIDI drivers)
1298 - added checks for bad MIDI data
1299
1300 * linuxsampler application:
1301 - show available sampler engine types on startup
1302
1303 * general changes:
1304 - support for muting sampler channels and solo mode of the same, two new
1305 LSCP commands ("SET CHANNEL MUTE" and "SET CHANNEL SOLO") and two new
1306 fields ("MUTE" and "SOLO") for command "GET CHANNEL INFO" were
1307 introduced for this, the behavior is the same like on a mixer console
1308 (patch by Grigor Iliev, a bit adjusted). Also added configure option
1309 --enable-process-muted-channels which can be used to enable the
1310 processing of muted channels.
1311 - support for sostenuto pedal
1312 - support for monitoring the total number of active voices
1313 (new LSCP commands: GET TOTAL_VOICE_COUNT, GET TOTAL_VOICE_COUNT_MAX,
1314 SUBSCRIBE TOTAL_VOICE_COUNT, UNSUBSCRIBE TOTAL_VOICE_COUNT)
1315 - fixed some memory management errors
1316 - fixed some concurrency problems that could lead to crashes when
1317 LSCP commands were executed
1318 - fixed crash when instrument loading failed previously on the same
1319 sampler channel (fixes bug #36)
1320
1321 Version 0.3.3 (15 July 2005)
1322
1323 * packaging changes:
1324 - fixed compilation with gcc 4.0
1325
1326 * Gigasampler format engine:
1327 - LFO preprocessing was performed even though the respective LFO was
1328 disabled by the instrument patch (minor efficiency fix)
1329 - if period time of chosen audio device is too small (< MIN_RELEASE_TIME)
1330 for volume ramp downs in the same fragment (needed for current voice
1331 stealing implementation) then simply inform the user with a warning
1332 about possible click sounds and reduce the volume ramp down
1333 appropriately instead of cancelling the audio device connection
1334
1335 Version 0.3.2 (24 June 2005)
1336
1337 * packaging changes:
1338 - updated autotools build files to compile on OS X
1339 (conditional compilation of CoreMIDI and MidiShare drivers)
1340 - hand-crafted assembly optimization code can be disabled with
1341 './configure --disable-asm' (definitely not recommended)
1342 - fixed 'make dist' rule to include all necessary files
1343 - require automake (>= 1.5) for 'make -f Makefile.cvs'
1344 (due to 'dist-bzip2' automake option)
1345
1346 * Gigasampler format engine:
1347 - support for the gig parameters for "release velocity response" curves
1348 - fine tuning of the EGADSR envelope
1349 - volume of release triggered samples now depends on note-on velocity,
1350 note length and gig parameter "release trigger decay" instead of
1351 note-off velocity.
1352 - revised voice stealing
1353 (fixes crash and endless loop caused by voice stealing)
1354 - don't reset scale tuning on instrument or audio output device change
1355 - handle key group conflicts right at the beginning of each fragment
1356 instead of when the respective voice is actually launched
1357 (fixes undefined behavior if stolen voices belonged to a key group -
1358 this case was followed by a "killed voice survived" error message)
1359 - fixed minor issue with null/silence samples
1360 (those stole voices even though they don't need a voice at all which
1361 resulted in "voice stealing didn't work out" messages)
1362 - don't reset volume, pan, pitch and MIDI controller values on
1363 instrument or audio output device change
1364
1365 * LSCP server:
1366 - fixed some crashes (patch by Grigor Iliev, fixes #19)
1367 - fixed LSCP event "CHANNEL_INFO" notification
1368 (e.g. did not notify on volume changes or MIDI program change events)
1369
1370 * linuxsampler application:
1371 - added command line parameters --lscp-addr and --lscp-port to override
1372 default IP address and port of LSCP server
1373
1374 Version 0.3.1 (24 May 2005)
1375
1376 * initial release

  ViewVC Help
Powered by ViewVC