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

Contents of /linuxsampler/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 2063 - (show annotations) (download)
Sun Feb 28 07:28:56 2010 UTC (14 years, 1 month ago) by persson
File size: 33498 byte(s)
* sfz engine: added support for off_mode=normal

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

  ViewVC Help
Powered by ViewVC