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

Contents of /linuxsampler/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 2207 - (show annotations) (download)
Fri Jul 15 15:43:49 2011 UTC (12 years, 8 months ago) by iliev
File size: 40400 byte(s)
* simplified the signal unit model
* sf2: initial implementation of modulation LFO and cutoff filter

1 Version CVS HEAD (?)
2
3 * plugin changes:
4 - LV2 Persist + Files support (patch by David Robillard)
5
6 * packaging changes:
7 - fixed building with newer MinGW-w64
8 - Mac OS X: support the new dir for Core Audio SDK
9 - Mac OS X: fixed building outside source directory
10 - made sure all source files for hostplugins are included when
11 doing "make dist"
12 - removed empty directories left from the cvs to svn migration
13 - fixed compilation with gcc 4.6.1
14 - another "make dist" fix, for the LV2 plugin
15 - made --enable-pthread-testcancel default on Mac OS X
16 - minor Makefile fix for building lscpparser when build and source
17 directories are separate
18 - Mac OS X: made it possible to specify plugin installation dir to
19 configure
20
21 * general changes:
22 - Refactoring: moved the independent code from
23 the Gigasampler format engine to base classes
24 - command line option '--profile' is currently disabled, since the
25 respective profiling code is currently broken
26 - Introduced support for optional environment variable
27 "LINUXSAMPLER_PLUGIN_DIR" which allows to override the directory
28 where the sampler shall look for instrument editor plugins
29 (patch by Luis Garrido, slightly modified).
30 - implemented Roland GS NRPN 1ArrH which allows to set volume per note
31 - implemented Roland GS NRPN 1CrrH which allows to set pan per note
32 - implemented Roland GS NRPN 1DrrH which allows to set reverb send per
33 note (in this implementation of the sampler its simply hard routed to
34 the 1st effect send of the sampler channel, no matter what the actual
35 effect type is)
36 - implemented Roland GS NRPN 1ErrH which allows to set chorus send per
37 note (in this implementation of the sampler its simply hard routed to
38 the 2nd effect send of the sampler channel, no matter what the actual
39 effect type is)
40 - implemented support for internal LADSPA effects
41 - raised limit of program change queue from 100 to 512
42 (as suggested by Alex Stone)
43
44 * Gigasampler format engine:
45 - implemented the "round robin keyboard" dimension
46 - fixed round robin and random dimensions for cases when number of
47 dimension zones is not a power of two
48 - made round robin use a counter for each region instead of each
49 key
50
51 * SFZ format engine:
52 - Initial implementation (not usable yet)
53 - added support for v2 multiple stage envelope generators
54 - added a fine-tuned v1 envelope generator instead of using the
55 one from the gig engine
56 - fixed memory leak and memory handling errors
57 - added support for trigger=first, trigger=legato and sw_previous
58 - allow non-numerical key values ("C#4" for example)
59 - "key" opcode now sets pitch_keycenter too
60 - fixed error when unloading instrument with same sample used by
61 multiple regions
62 - added some opcode aliases, like loopmode for loop_mode, to be
63 more compatible
64 - added support for trigger=release and rt_decay
65 - added support for off_mode=normal
66 - added support for random, seq_position, seq_length and volume
67 - added v1 LFO opcodes to parser (no support in engine yet)
68 - added support for amp_veltrack and amp_velcurve_N
69 - fine-tuned the default velocity curve
70 - added support for transpose
71 - fixed crash when using sw_down/up
72 - improved logic for sw_lokey/hikey/up/down/last
73 - added more v1 aliases to parser
74 - reduced memory usage for sfz data
75 - RT-safeness: avoid malloc in audio thread
76 - fixed a bug that could cause voice stealing to fail
77 - optimized sample lookup
78 - improved support for exclusive groups (group, off_by and
79 off_mode)
80 - added support for controller triggered regions (on_locc/on_hicc)
81 - added support for loop_mode=one_shot
82 - fixed playback of 16 bit wav files on big endian CPUs
83 - added support for Ogg Vorbis sample files
84 - use loop markers from sample file if loop_start and loop_end are
85 not set in sfz file
86 - implemeted filters. Filter types: lowpass, bandpass, bandreject
87 and highpass. 1, 2, 4 and 6 pole filters. Opcodes: fil_type,
88 cutoff, resonance, fil_veltrack, fil_keytrack, fil_keycenter,
89 cutoff_cc, cutoff_chanaft.
90 - bugfix: zero ampeg_sustain didn't work
91 - added support for velocity effect on amplifier envelope time
92 (ampeg_vel2attack, ampeg_vel2decay, ampeg_vel2sustain and
93 ampeg_vel2release)
94 - added support for EG hold (ampeg_hold)
95
96 * SoundFont format engine:
97 - Initial implementation (not usable yet)
98 - RT-safeness: avoid malloc in audio thread
99 - fixed a bug that could cause voice stealing to fail
100 - fine-tuned amplitude EG (by switching from gig to sfz EG)
101 - initial implementation of Vibrato LFO and Modulation LFO
102 - initial implementation of cutoff filter
103
104 * Gigasampler format engine:
105 - bugfix: pitch LFO controller "internal+aftertouch" was broken
106 - bugfix: filter keyboard tracking was broken
107 - filter performance fix (an unnecessary copy was made of the
108 filter parameters in each sub fragment)
109
110 * Host plugins:
111 - AU bugfix: failed to destroy its audio/MIDI devices
112 - Listen to all interfaces on Mac OS X (INADDR_ANY)
113 - VST bugfix: If the host called resume() before and after
114 changing sample rate or block size, the number of channels was
115 incorrectly set to two. This caused silence in Cubase 5.
116 - save engine type (gig, sfz or sf2) in plugin state
117 - VST: when opening Fantasia, look for both 32 and 64 bit Java on
118 64 bit Windows
119 - AU: changed number of output channels from one stereo to 16
120 stereo
121 - VST: made it possible to build the VST plugin for Mac
122 - AU: link AU plugin dynamically if --disable-shared isn't
123 specified
124
125 * MIDI driver:
126 - ALSA MIDI driver supports now "NAME" device parameter, for overriding
127 the ALSA sequencer client name
128 - removed limit of maximum amount of MIDI ports per MIDI device, since
129 there is no reason for this limit
130
131 * audio driver:
132 - ASIO driver fixes for newer gcc versions (fix from PortAudio)
133
134 * LSCP server:
135 - added support for sending MIDI CC messages via LSCP command
136 "SEND CHANNEL MIDI_DATA CC <sampler-chan> <ctrl> <val>"
137 - added LSCP command "GET AVAILABLE_EFFECTS"
138 - added LSCP command "LIST AVAILABLE_EFFECTS"
139 - added LSCP command "GET EFFECT INFO <effect-index>"
140 - added LSCP command "CREATE EFFECT_INSTANCE <effect-index>"
141 - added LSCP command
142 "CREATE EFFECT_INSTANCE <effect-system> <module> <effect-name>"
143 - added LSCP command "DESTROY EFFECT_INSTANCE <effect-instance>"
144 - added LSCP command "GET EFFECT_INSTANCES"
145 - added LSCP command "LIST EFFECT_INSTANCES"
146 - added LSCP command "GET EFFECT_INSTANCE INFO <effect-instance>"
147 - added LSCP command
148 "GET EFFECT_INSTANCE_INPUT_CONTROL INFO <effect-instance> <input-control>"
149 - added LSCP command "SET EFFECT_INSTANCE_INPUT_CONTROL VALUE
150 <effect-instance> <input-control> <value>"
151 - added LSCP command "GET SEND_EFFECT_CHAINS <audio-device>"
152 - added LSCP command "LIST SEND_EFFECT_CHAINS <audio-device>"
153 - added LSCP command "ADD SEND_EFFECT_CHAIN <audio-device>"
154 - added LSCP command
155 "REMOVE SEND_EFFECT_CHAIN <audio-device> <effect-chain>"
156 - added LSCP command
157 "GET SEND_EFFECT_CHAIN INFO <audio-device> <effect-chain>"
158 - added LSCP command "APPEND SEND_EFFECT_CHAIN EFFECT <audio-device>
159 <effect-chain> <effect-instance>"
160 - added LSCP command "INSERT SEND_EFFECT_CHAIN EFFECT <audio-device>
161 <effect-chain> <effect-chain-pos> <effect-instance>"
162 - added LSCP command "REMOVE SEND_EFFECT_CHAIN EFFECT <audio-device>
163 <effect-chain> <chain-pos>"
164 - added LSCP command "SET FX_SEND EFFECT <sampler_channel>
165 <fx_send_id> <effect_chain> <chain_pos>"
166 - added LSCP command "REMOVE FX_SEND EFFECT <sampler_channel> <fx_send_id>"
167 - added LSCP commands "SUBSCRIBE EFFECT_INSTANCE_COUNT",
168 "SUBSCRIBE EFFECT_INSTANCE_INFO", "SUBSCRIBE SEND_EFFECT_CHAIN_COUNT",
169 "SUBSCRIBE SEND_EFFECT_CHAIN_INFO"
170
171 * Bug fixes:
172 - Fixed crash which may occur when MIDI key + transpose is out of range
173 - minor valgrind fixes
174 - fixed crash which occurred when changing an already deployed sampler
175 channel to a different engine type
176 - fixed crash when deleting a sampler channel or changing engine
177 type while an instrument load was in progress
178 - bugfix: playing a note while changing the instrument could cause
179 a crash, or give "Handing back unknown region" error messages
180 - bugfix: calling SET VOICES (which Fantasia does on start and
181 refresh) could cause instruments to be unloaded
182 - Mac OS X: fixed name collision of enum in EffectControl and
183 wrong name of destructor in AudioOutputDeviceCoreAudio.cpp
184 - Mac OS X: fixed hanging threads
185 - Windows: disabled the previous thread fix on non-Mac systems, as
186 it caused hanging threads on Windows
187 - Fixed possible crashes due to corrupted MIDI/audio device list
188 after MIDI/audio device creation failure
189
190 Version 1.0.0 (31 July 2009)
191
192 * packaging changes:
193 - autoconf bugfix: the PKG_CONFIG variable wasn't initialized properly,
194 causing e.g. the libgig test to fail when
195 "./configure --disable-jack-driver" was used
196 (patch by Alexis Ballier)
197 - fixed compilation with gcc 4.3
198 - fixes for building on OS X (thanks to Ebrahim Mayat for testing)
199 - fixed configure so it detects x86_64 (#107)
200 - fixes for building with newer MinGW versions
201 - fix for building with bison 2.4 (#111)
202 - fixed building with libgig installed in a non-standard directory
203 - minor fix in configure for mmsystem.h detection on MinGW
204 - Windows: look for editor plugins and Fantasia using base
205 directory of liblinuxsampler dll (look in the same directory and one
206 directory above)
207 - configure script fix: removed unconditional use of SSE
208 - fixed building with sqlite installed in a non-standard directory
209 - when cross-compiling, don't try to create instruments.db
210 - fix for new mingw-w64 version, which has usleep
211
212 * general changes:
213 - bugfix: on some POSIX systems instrument editor plugins refused to
214 load as we used a non-portable Linux specific struct field
215 (fixes bug #70, patch by Ronald Baljeu)
216 - fixed endless loop which occured when loading an instrument editor
217 plugin DLL on Windows systems failed
218 - fixed memory leaks that occurred when liblinuxsampler was unloaded
219 - optimized the SynchronizedConfig class so it doesn't wait
220 unnecessarily long after an update
221 - added support for notifying instrument editors on note-on / note-off
222 events (e.g. to highlight the pressed keys on the virtual keyboard
223 of gigedit)
224 - added support for triggering notes by instrument editors (see above)
225 - be verbose on DLL load errors (on Linux)
226 - fixed di-harmonic triangle LFO implementation (this LFO implementation
227 is only used by very few systems, most pick the int math
228 implementation, which is usually faster)
229 - fixes for audio drivers with varying buffer sizes
230 - experimental support for running LinuxSampler as a DSSI, LV2 and
231 VST plugin
232 - notification events for stream/voice count statistics are now sent
233 only when there are actual changes
234 - added memory ordering constraints to improve stability on
235 multi-core and multi-cpu systems
236 - maximum voices and disk streams can now be altered at runtime
237 - fixed CPU feature detection on x86_64 (maybe fixes #108)
238 - automatic stacktrace mechanism is now turned off by default and can
239 be switched on by command line option "--stacktrace" (the automatic
240 stacktrace mechanism seems to be broken on most systems at the moment)
241 - C++ API method InstrumentManager::LaunchInstrumentEditor() now returns
242 a pointer to the launched InstrumentEditor object
243 - added optional 3rd party user data parameter for following
244 liblinuxsampler C++ API methods: InstrumentEditor::Main(),
245 InstrumentEditor::Launch(),
246 InstrumentManager::LaunchInstrumentEditor()
247 - theoretical fix: made SynchronizedConfig follow C++0x memory
248 model more strictly
249 - fixes for using large audio device buffers
250 - Windows: add the installation directory to the DLL search path
251 when loading an editor plugin (solves problems with VST and
252 gigedit on systems with other GTK versions installed)
253
254 * audio driver:
255 - removed the nonsense audio channel constraint (which was hard coded to
256 max. 100 audio channels) for most audio drivers
257 - JACK audio driver did not offer a device parameter "SAMPLERATE" as
258 opposed to the LSCP specs
259 - bugfix: the SAMPLERATE parameter of some drivers (e.g. JACK)
260 reflected the wrong value
261 - fixed a memory management bug in ASIO driver
262 - Makefile fix: JACK_CFLAGS wasn't used
263 - JACK: use jack_client_open instead of the deprecated
264 jack_client_new
265 - added (experimental) CoreAudio driver
266 - applied old fixes to the ASIO driver that were included in the
267 previous binary release but accidentally never committed to CVS
268 (fixes #117)
269 - fixes for ASIO on mingw-w64 (iasio wrapper is not needed on
270 win64)
271 - VST: added support for sample rate and buffer size changes
272 - VST: close editor (Fantasia) when the VST is removed
273 - VST: avoid opening Fantasia more than once for each VST instance
274 - VST: export main function as "main" on Linux too (fix for
275 energyXT)
276 - VST: changed number of output channels from one stereo to 16
277 stereo
278 - added channel routing, fxsends and midi maps to the settings
279 stored in the plugin state
280 - performance optimization of AudioChannel::MixTo() and
281 AudioChannel::CopyTo() methods using GCC vector exensions
282 (if available)
283 - ASIO fixes: avoid initializing the device twice, avoid throwing
284 exception when getting parameters from a disconnected device
285
286 * MIDI driver:
287 - added JACK MIDI driver
288 - dispatch bank select as ordinary CC as well, the user might seriously
289 want to (mis)use it for some purpose ("fixed" in all current MIDI
290 input drivers: ALSA, CoreMIDI, JACK, MidiShare, MME)
291 - bugfix: pitch bend wasn't working with JackMidi, VST, LV2, MME,
292 CoreMidi or AU
293 - fixed mingw-w64 compilation error in MME driver
294 - made program change handling in MIDI thread real-time safe by
295 moving the logic to a non-RT thread
296 - fixed minor memory leak in ALSA driver
297
298 * instruments database:
299 - avoid time consuming samples scanning when adding instruments
300 to the instruments database
301 - added support for handling lost files in the instruments database
302 - Implemented option for adding instruments in separate directories
303 in the instruments database
304 (patch by Chris Cherrett & Andrew Williams, a bit adjusted)
305 - work-around for missing fnmatch function on Windows to make
306 instrument database compilable
307 - added instrument database support on Windows
308 works with both standalone mode and VST plugin.
309 the instruments DB file is located in
310 %USERPROFILE%\.linuxsampler\instruments.db which allows different
311 databases for each windows user
312 if no DB is present it automatically creates the .linuxsampler subdir
313 and then creates an empty DB
314 - fixed recursive import, which was broken on Windows
315
316 * Gigasampler format engine:
317 - fixed a memory leak that could happen when a channel was deleted
318 while notes were playing
319 - made it possible to load an instrument even if the audio thread
320 isn't running
321 - added partial support for the "Controller Triggered" MIDI rule,
322 enough for piano gigs with pedal noise samples
323 - minor fix: only mark FX sends as being modified if really the
324 respective FX send MIDI controller was used
325 - added support for GM global device volume SysEx message
326 (can be selected at compile time to either apply globally to the
327 whole sampler [default] or only to the sampler channels that are
328 connected to the respective MIDI input port the SysEx message arrived
329 on)
330 - bugfix: notes triggered at position 0 in the audio buffer were
331 sometimes wrongly killed in the same buffer, causing no sound to
332 be played
333 - added support for chromatic / drumkit mode Roland GS Sysex message
334 which is usally used to switch a part between chromatic sounds and
335 drumkit sounds (as opposed to their common default setting of MIDI
336 part 10 being a drumkit part and all other ones chromatic parts), in
337 LS however one can switch between the first 16 MIDI instrument maps
338 defined for the sampler
339 - made it possible to create multiple sample channels even if the
340 audio thread isn't running
341 - fixed crash when removing channel with active voices (#116)
342 - bugfix: on sample reference changes (instrument editor), only
343 un-cache the respective sample if it's really not used by any
344 sampler engine anymore
345 - re-cache samples in case they were changed by an instrument editor,
346 e.g. when a sample was added while playing (#82)
347 - fixed hanging notes which occured when note-off event had the exact
348 same time stamp as the voice's note-on event and occured both in the
349 same audio fragment cycle (fixes bug #112)
350 - added support for the "fine tune" and "pitch bend range"
351 instrument-level gig parameters
352 - fixed minor artifacts in pitch bend handling
353 - added support for GS Reset SysEx message
354 - allow gig files to use unlimited downward pitch shifting
355 - added a limit check for upward pitch shifting
356 - bugfix: sometimes, when playing a note twice fast, the second
357 note was silent
358 - fixed crash happening when a pitch bend event arrived at the
359 same time a new instrument was loading
360
361 * LSCP server:
362 - added new LSCP event "CHANNEL_MIDI" which can be used by frontends to
363 react on MIDI data arriving on certain sampler channels (so far only
364 Note-On and Note-Off events are sent via this LSCP event)
365 - added new LSCP event "DEVICE_MIDI" which can be used by frontends to
366 react on MIDI data arriving on certain MIDI input devices (so far only
367 Note-On and Note-Off events are sent via this LSCP event)
368 - added new LSCP commands: FIND LOST DB_INSTRUMENT_FILES and
369 SET DB_INSTRUMENT FILE_PATH
370 - added new LSCP command: SEND CHANNEL MIDI_DATA which can be used by
371 frontends to send MIDI messages to specific sampler channel
372 - added two additional fields to GET FILE INSTRUMENT INFO command -
373 KEY_BINDINGS and KEYSWITCH_BINDINGS
374 - bugfix: the bank number provided by MIDI_INSTRUMENT_INFO notifications
375 was incorrect
376 - Optimized the retrieval of the MIDI instrument mappings
377 - added new LSCP command "SET VOICES" to globally alter the maximum
378 amount of voices, added new LSCP event "GLOBAL_INFO:VOICES" which will
379 be triggered respectively
380 - added new LSCP command "SET STREAMS" to globally alter the maximum
381 amount of disk streams, added new LSCP event "GLOBAL_INFO:STREAMS"
382 which will be triggered respectively
383 - bugfix: retry if "select" returns EINTR (this fixes a crash when
384 a gigedit file dialog is opened)
385 - close all connections when LSCPServer is deleted
386 - hard close of all sockets on exit
387 - bugfix: SET CHANNEL MIDI_INPUT_TYPE didn't work with the MME
388 driver
389
390 * Bug fixes:
391 - fixed a crash which occurs when removing a sampler channel waiting
392 to start instrument loading after another channel
393 - fixed a crash which occurs when removing a sampler channel with
394 instrument loading in progress (bug #113)
395 - fixed termination caused by uncaught exception when adding MIDI
396 instrument with PERSISTENT load mode
397 - fixed possible iterator invalidations when resetting the sampler
398 - fixed memory leaks when issuing the following LSCP commands:
399 GET AUDIO_OUTPUT_DRIVER INFO
400 GET AUDIO_OUTPUT_DRIVER_PARAMETER INFO
401 GET MIDI_INPUT_DRIVER INFO
402 GET MIDI_INPUT_DRIVER_PARAMETER INFO
403 - fixed possible compilation error when sqlite is not present
404 - fixed orphaned pointers when setting maximum voices limit (bug #118)
405 - fixed crash when changing the audio output device of a sampler
406 channel with loaded instrument and start playing notes
407 - fixed endless loop in Engine::SuspendAll() (bug #120)
408 - fixed a low-level atomic load function that was broken on 64-bit
409 PowerPC, which probably could cause crashes on that platform
410 - fixed a memory management error which could cause a crash when a
411 plugin was unloaded
412 - bugfix: two private structs had the same name, which could cause
413 problems if the linker chose the wrong constructor
414 - fixed low-level ConditionServer usage bug that caused lockups on
415 Windows
416
417
418 Version 0.5.1 (6 December 2007)
419
420 * packaging changes:
421 - added autoconf checks for pthread library
422 - added autoconf check for pthread bug found on certain NPTL-enabled
423 glibc versions (see Gentoo bug report #194076)
424 - added autoconf checks for MS Windows
425
426 * general changes:
427 - bugfix: the thread used by an editor plugin didn't die when the
428 editor closed
429 - bugfix: Ringbuffer.h: fill_write_space_with_null() did not zero
430 out all the space. operator--() did not apply size_mask after
431 decrementing the read_ptr. DEFAULT_WRAP_ELEMENTS set to 0 in
432 order to avoid problems with the _NonVolatileReader functions.
433 - bugfix: Stream.h: added a missing divide by BytesPerSample in
434 GetWriteSpace(). Since this function is currently only used in
435 the stream's qsort() compare function, it didn't trigger any
436 bugs.
437 - Resampler.h, Synthesizer.h: cubic interpolation now works in
438 24bit mode too. Faster method to read 24bit words on little
439 endian machines (x86): replaced 3 byte reads + shifts with a 1
440 unaligned 32bit read and shift
441 - experimental support for MS Windows (MIDI input via MME, AUDIO
442 output via ASIO)
443 - made handling of SIGINT signal (Ctrl-C) a bit more robust
444 - support for monitoring the total number of active disk streams
445 (new LSCP commands: GET TOTAL_STREAM_COUNT,
446 SUBSCRIBE TOTAL_STREAM_COUNT, UNSUBSCRIBE TOTAL_STREAM_COUNT)
447
448 * AUDIO driver:
449 - added Windows ASIO low latency audio driver
450
451 * MIDI driver:
452 - added MME Windows MIDI driver
453
454 * LSCP server:
455 - added support for Windows style path / filenames, however with
456 forward slash path separators instead of backslash
457 (i.e. "C:/foo/bar.gig")
458 - allow naughty liblscp to send non-string device parameters within
459 apostrophes as well
460 - added new LSCP commands: "GET FILE INSTRUMENTS <file>",
461 "LIST FILE INSTRUMENTS <file>" and
462 "GET FILE INSTRUMENT INFO <file> <index>" for retrieving informations
463 about an arbitrary instrument file on the system where the sampler is
464 running on
465
466 Version 0.5.0 (15 October 2007)
467
468 * packaging changes:
469 - config.h is not going to be installed along with liblinuxsampler's
470 API header files anymore
471 - only the API relevant header (and source) files will be exported to
472 the Doxygen API documentation (explicitly listed in Doxyfile.in)
473 - added completely new XCode project files for Mac OSX which is now
474 capable to execute our autoconf environment, thus no need anymore
475 to maintain the compile time configuration file (osx/version.h) for
476 OSX manually (patch by Toshi Nagata)
477 - fixed buggy boolean --enable-foo / --disable-foo configure script
478 parameters
479 - global.h now only covers global definitions that are needed for the
480 C++ API header files, all implementation internal global definitions
481 are now in global_private.h
482 - atomic.h is not exposed to the C++ API anymore
483 - no need to include config.h anymore for using LS's API header files
484 - fixed warnings in API doc generation
485
486 * general changes:
487 - replaced the old, confusing MIDI program change mechanism by a
488 flexible MIDI instrument mapper which allows to map arbitrary
489 (MIDI bank, MIDI program) pairs with arbitrary
490 (engine type, instrument file, file index) triplets which will be
491 loaded on the respective channel when such MIDI program change
492 messages arrive, beside that, each entry allows to define a life-time
493 strategy for the instrument, so the user can define whether the
494 instrument should i.e. be permanently loaded (to be able to switch
495 quickly among sounds) or i.e. loaded just on demand when the
496 respective program change arrives (to safe RAM space), as well as a
497 global volume factor for each entry, so the user can adjust the volume
498 dynamic of his mapped instrument collection without having to modify
499 the instrument files, also one can manage arbitrary amount of such
500 MIDI instrument maps and assign each sampler channel individually
501 a certain map, so that i.e. separation of normal instruments and
502 drumkits is possible
503 - new notification events for tracking changes to audio/MIDI devices,
504 MIDI instrument maps, MIDI instruments, FX sends, global volume.
505 - sampler was limited to load max. 200 instruments in the background
506 due to a constant size RingBuffer FIFO which is now replaced by a
507 dynamic (unlimited) size std::list FIFO
508 - added FX sends, these allow to route audio signals to arbitrary audio
509 output channels for being processed by external effect processors
510 (i.e. jack-rack), the send levels are controllable via arbitrary MIDI
511 controllers
512 - global (sampler wide) volume can now be controlled at runtime
513 - Implemented new, improved notification system
514 - fixed compilation errors regarding OSX
515 (patch by Toshi Nagata)
516 - implemented instruments database
517 - added support for escape sequences to the instruments database
518 - added highly experimental support for on-the-fly instrument editing
519 within the sampler's process (by using instrument editor plugins),
520 you'll notice the new "Registered instrument editors:" message on
521 startup, added a new LSCP command:
522 "EDIT CHANNEL INSTRUMENT <sampler-channel>"
523 to spawn a matching instrument editor for the instrument on the
524 given sampler channel, the plugin path can be overridden at compile
525 time with ./configure --enable-plugin-dir=/some/dir
526 - added experimental code for synchronizing instrument editors hosted
527 in the sampler's process to safely edit instruments while playing
528 without a crash (hopefully) by either suspending single regions
529 wherever possible or if unavoidable whole engine(s)
530 - fixed several issues in fundamental "Thread" class: set scheduling
531 policy and priority on thread level, set a minimum stack size for
532 thread (TODO: a reasonable value yet to be tested), bugfix: non-RT
533 threads simply inherited properties of starting thread instead of
534 setting their own policy and priority
535 - minor fix in our automatic stack trace mechanism on crashes, the
536 main process did not wait for the stack trace process to finish
537 its output
538 - fixed some minor memory leaks
539 - reenabled assembly features support, at the moment only for
540 enabling a fast denormal FPU mode (x86 platforms supporting SSE2)
541 - minor assembly fix in x86 features detection (don't use the PIC
542 register, to avoid relocations in the text segment at runtime)
543 - POSIX callback functions of Thread.h are hidden
544
545 * MIDI driver:
546 - dispatch bank select (MSB and LSB) messages
547
548 * audio driver:
549 - the ALSA audio output driver parameters now reflect the correct
550 parameter value ranges for the respective selected sound card
551 (patch by Till Wimmer, a bit fixed and extended)
552
553 * Gigasampler format engine:
554 - if a filter is used and EG2 finishes before EG1, let the voice
555 die when EG2 has finished (this fixes a problem with clicks and
556 voice starvation for some gigs)
557 - playback is no longer disabled during instrument loading
558 - all notes playing on a channel that changes its instrument keep
559 playing with the old instrument until they get a note off
560 command
561 - EG fix: a release value of zero could cause noises or crash
562 - handle MIDI coarse tuning messages (MIDI RPN #0 MSB #2 LSB)
563 - EG fine tuning: when attack is zero the EG starts at a level
564 above max sustain level, which means that there is a decay phase
565 even if sustain is 100%
566 - more EG fixes: the level could sometimes go below zero and cause
567 noises or crashes
568 - minor fix of EGDecay (patch by Toshi Nagata)
569 - fixed compiler error when --enable-override-filter-type was
570 supplied to the configure script (fixes #46)
571 - disk thread: queue sizes are now proportional to CONFIG_MAX_STREAMS
572 instead of a fix value
573 - behavior fix: on MIDI CC# 65 (portamento on / off), 126 (mono mode),
574 127 (solo mode) only kill voices if the respective mode really
575 changed
576
577 * LSCP server:
578 - fixed compile time error for old Bison versions
579 (i.e. v1.28 found on MacOS 10.4, patch by Toshi Nagata)
580 - parser now supports extended ASCII character set
581 (up to ASCII code 255, i.e. includes now umlauts and accents)
582 - filename arguments in LSCP commands now allow to use escape
583 sequences, that is directly literal as one of: \', \", \\, \n, \r,
584 \f, \t, \v, or as octal ASCII code value like \132, or as hex ASCII
585 code value like \xf2) (fixes bug #24)
586 - the following LSCP commands now also support escape sequences for at
587 least one of their text-based parameters (i.e. name, description):
588 "ADD MIDI_INSTRUMENT_MAP", "MAP MIDI_INSTRUMENT",
589 "SET MIDI_INSTRUMENT_MAP NAME", "SET FX_SEND NAME", "CREATE FX_SEND",
590 "SET DB_INSTRUMENT_DIRECTORY NAME",
591 "SET DB_INSTRUMENT_DIRECTORY DESCRIPTION", "SET DB_INSTRUMENT NAME",
592 "SET DB_INSTRUMENT DESCRIPTION", "FIND DB_INSTRUMENTS",
593 "FIND DB_INSTRUMENT_DIRECTORIES"
594 - returns verbose syntax errors (line and column where syntax error
595 occured, the unexpected character and the actually expected, possible
596 character(s), the latter only if less than 5 possibilities)
597 - made sure that LSCP syntax is not affected by gigedit locale
598 settings
599 - bugfix regarding strings parameter lists: all comma separated lists
600 of strings were treated as being one string containing commas
601 (fixes #57)
602
603 * Bug fixes:
604 - fixed crash occurring on certain LSCP scripts (Bug 39)
605 - another thread safety fix for lscp "load engine" and "set
606 channel audio output device"
607 - fixed a crash which occurs when reassigning the same engine
608 on a sampler channel with connected MIDI device
609 - fixed a crash which occurs when changing the number of ports of a MIDI
610 device connected to a sampler channel to number less then or equal
611 to the index of the port to which the sampler channel is connected.
612 - The previous bindings were not been disconnected when altering
613 the ALSA_SEQ_BINDINGS parameter. Introduced a NONE keyword for
614 unsubscribing from all bindings (e.g. ALSA_SEQ_BINDINGS=NONE).
615 - The active stream/voice count statistic was incorrect.
616 - notification events were not been sent for some sampler
617 channel changes
618 - added default min and max values to restrict the number of allowed
619 audio output channels and MIDI input ports
620 - the connection to the PCM interface is now closed when destroying
621 an audio output device
622 - files with slash in their path or filename could not be loaded
623
624 * test cases:
625 - updated and fixed (haven't been touched in a while)
626
627 Version 0.4.0 (24 November 2006)
628
629 * packaging changes:
630 - changed deprecated copyright attribute to license; added ldconfig
631 to post-(un)install steps; added devel package for liblinuxsampler;
632 to linuxsampler.spec (RPM)
633 - install necessary development header files for allowing 3rd party
634 applications to link against liblinuxsampler
635 - liblinuxsampler's API documentation can be generated with 'make docs'
636 (Doxygen required)
637 - added benchmark to automatically detect the best triangle LFO
638 implementation (currently either an integer math solution or a
639 di-harmonic approximation), automatic detection can be overriden
640 with --enable-signed-triang-algo=x and --enable-unsigned-triang-algo=x
641 configure script argument though (mandatory for cross-compilation)
642 - do not automatically pick optimized gcc flags if the user already
643 provided some on his own (as CXXFLAGS)
644 - added compile time option to disable processing of All-Notes-Off MIDI
645 messages
646 - added compile time options to allow disabling the various audio and
647 MIDI drivers
648 - fixed automatic GCC CPU switch detection on PPC
649 (patch by Ebrahim Mayat)
650
651 * Gigasampler format engine:
652 - extensive synthesis optimization
653 (reimplementation of EGs and LFO(s), removed synthesis parameter
654 prerendering and the synthesis parameter matrix in general, splitting
655 each audio fragment into subfragments now where each subfragment uses
656 constant pitch and filter coefficients. The volume coefficient is
657 linearly interpolated inside a subfragment, unless
658 --disable-interpolate-volume is set.)
659 - fine tuning of the EG modulation parameters
660 - improved filter cutoff calculation by adding support for the
661 following gig parameters: Cutoff freq (used when no cutoff
662 controller is defined), Control invert, Minimum cutoff, Velocity
663 curve and Velocity range. The keyboard tracking now scales
664 cutoff frequency instead of resonance.
665 - added support for gig parameter Resonance.
666 - fixed bug in sysex handling (patch by Juan Linietsky)
667 - added global volume attenuation of -9 dB (0.35f) to prevent clipping
668 which can be overridden with --enable-global-attenuation
669 - EG fixes: made the length of "attack hold" stage more
670 accurate. Release stage can now start before attack stage
671 ends. Cancel release didn't work when sustain was zero. Attack
672 time now has a minimal value to prevent clicks.
673 - fixed pitch changes larger than one octave
674 - fixed EG3 (pitch envelope) synthesis which was neutral all the time
675 - implemented portamento mode and solo mode (a.k.a 'mono mode'):
676 all modes can be altered via standard GM messages, that is CC5 for
677 altering portamento time, CC65 for enabling / disabling portamento
678 mode, CC126 for enabling solo mode and CC127 for disabling solo mode
679 - fine tuning of the curves for volume (CC7), pan (CC10 and gig
680 parameter) and crossfade
681 - added support for the "attenuation controller threshold" gig
682 parameter
683 - added smoothing of volume changes caused by control change
684 messages
685 - sample loop parameters are now taken from the DimensionRegion
686 instead of the wave chunk
687 - fixed keyswitching for v3 gigs with a number of keyswitch splits
688 not equal to a power of two
689 - reimplementation of the filter algorithm. The new filter is more
690 accurate and supports all gig filter types, including bandreject
691 and lowpass turbo.
692 - real support for 24 bit samples - samples are not truncated to
693 16 bits anymore
694 - support for aftertouch (channel pressure, not polyphonic
695 aftertouch)
696 - LFO1 behaviour fixed (dampening from max volume instead of
697 amplifying from 0)
698
699 * LSCP server:
700 - fixed application exit on broken pipe error (fixes bug #20)
701 - fixed the notification messages delay due to lack of
702 network activity (fixes bug #26)
703 - fixed parser bug which occured on space(s) within device parameters
704 (e.g. "SET AUDIO_OUTPUT_CHANNEL_PARAMETER 0 0 SOMEPARAM='foo bar'")
705
706 * audio driver:
707 - added aRts audio output driver (by no means RT safe)
708
709 * MIDI driver:
710 - fixed legacy sysex code which caused dispatching of MIDI SysEx
711 messages several times instead of once
712 - API extension for MIDI drivers which already supply exact time stamps
713 for events (i.e. for offline rendering based MIDI drivers)
714 - added checks for bad MIDI data
715
716 * linuxsampler application:
717 - show available sampler engine types on startup
718
719 * general changes:
720 - support for muting sampler channels and solo mode of the same, two new
721 LSCP commands ("SET CHANNEL MUTE" and "SET CHANNEL SOLO") and two new
722 fields ("MUTE" and "SOLO") for command "GET CHANNEL INFO" were
723 introduced for this, the behavior is the same like on a mixer console
724 (patch by Grigor Iliev, a bit adjusted). Also added configure option
725 --enable-process-muted-channels which can be used to enable the
726 processing of muted channels.
727 - support for sostenuto pedal
728 - support for monitoring the total number of active voices
729 (new LSCP commands: GET TOTAL_VOICE_COUNT, GET TOTAL_VOICE_COUNT_MAX,
730 SUBSCRIBE TOTAL_VOICE_COUNT, UNSUBSCRIBE TOTAL_VOICE_COUNT)
731 - fixed some memory management errors
732 - fixed some concurrency problems that could lead to crashes when
733 LSCP commands were executed
734 - fixed crash when instrument loading failed previously on the same
735 sampler channel (fixes bug #36)
736
737 Version 0.3.3 (15 July 2005)
738
739 * packaging changes:
740 - fixed compilation with gcc 4.0
741
742 * Gigasampler format engine:
743 - LFO preprocessing was performed even though the respective LFO was
744 disabled by the instrument patch (minor efficiency fix)
745 - if period time of chosen audio device is too small (< MIN_RELEASE_TIME)
746 for volume ramp downs in the same fragment (needed for current voice
747 stealing implementation) then simply inform the user with a warning
748 about possible click sounds and reduce the volume ramp down
749 appropriately instead of cancelling the audio device connection
750
751 Version 0.3.2 (24 June 2005)
752
753 * packaging changes:
754 - updated autotools build files to compile on OS X
755 (conditional compilation of CoreMIDI and MidiShare drivers)
756 - hand-crafted assembly optimization code can be disabled with
757 './configure --disable-asm' (definitely not recommended)
758 - fixed 'make dist' rule to include all necessary files
759 - require automake (>= 1.5) for 'make -f Makefile.cvs'
760 (due to 'dist-bzip2' automake option)
761
762 * Gigasampler format engine:
763 - support for the gig parameters for "release velocity response" curves
764 - fine tuning of the EGADSR envelope
765 - volume of release triggered samples now depends on note-on velocity,
766 note length and gig parameter "release trigger decay" instead of
767 note-off velocity.
768 - revised voice stealing
769 (fixes crash and endless loop caused by voice stealing)
770 - don't reset scale tuning on instrument or audio output device change
771 - handle key group conflicts right at the beginning of each fragment
772 instead of when the respective voice is actually launched
773 (fixes undefined behavior if stolen voices belonged to a key group -
774 this case was followed by a "killed voice survived" error message)
775 - fixed minor issue with null/silence samples
776 (those stole voices even though they don't need a voice at all which
777 resulted in "voice stealing didn't work out" messages)
778 - don't reset volume, pan, pitch and MIDI controller values on
779 instrument or audio output device change
780
781 * LSCP server:
782 - fixed some crashes (patch by Grigor Iliev, fixes #19)
783 - fixed LSCP event "CHANNEL_INFO" notification
784 (e.g. did not notify on volume changes or MIDI program change events)
785
786 * linuxsampler application:
787 - added command line parameters --lscp-addr and --lscp-port to override
788 default IP address and port of LSCP server
789
790 Version 0.3.1 (24 May 2005)
791
792 * initial release

  ViewVC Help
Powered by ViewVC