ViewVC logotype

Contents of /linuxsampler/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 1847 - (show annotations) (download)
Fri Feb 27 17:12:40 2009 UTC (11 years, 1 month ago) by iliev
File size: 26537 byte(s)
* fixed crash when changing the audio output device of a sampler
  channel with loaded instrument and start playing notes

1 Version CVS HEAD (?)
3 * packaging changes:
4 - autoconf bugfix: the PKG_CONFIG variable wasn't initialized properly,
5 causing e.g. the libgig test to fail when
6 "./configure --disable-jack-driver" was used
7 (patch by Alexis Ballier)
8 - fixed compilation with gcc 4.3
9 - fixes for building on OS X (thanks to Ebrahim Mayat for testing)
10 - fixed configure so it detects x86_64 (#107)
11 - fixes for building with newer MinGW versions
12 - fix for building with bison 2.4 (#111)
13 - fixed building with libgig installed in a non-standard directory
15 * general changes:
16 - bugfix: on some POSIX systems instrument editor plugins refused to
17 load as we used a non-portable Linux specific struct field
18 (fixes bug #70, patch by Ronald Baljeu)
19 - fixed endless loop which occured when loading an instrument editor
20 plugin DLL on Windows systems failed
21 - fixed memory leaks that occurred when liblinuxsampler was unloaded
22 - optimized the SynchronizedConfig class so it doesn't wait
23 unnecessarily long after an update
24 - added support for notifying instrument editors on note-on / note-off
25 events (e.g. to highlight the pressed keys on the virtual keyboard
26 of gigedit)
27 - added support for triggering notes by instrument editors (see above)
28 - be verbose on DLL load errors (on Linux)
29 - fixed di-harmonic triangle LFO implementation (this LFO implementation
30 is only used by very few systems, most pick the int math
31 implementation, which is usually faster)
32 - fixes for audio drivers with varying buffer sizes
33 - experimental support for running LinuxSampler as a DSSI, LV2 and
34 VST plugin
35 - notification events for stream/voice count statistics are now sent
36 only when there are actual changes
37 - added memory ordering constraints to improve stability on
38 multi-core and multi-cpu systems
39 - maximum voices and disk streams can now be altered at runtime
40 - fixed CPU feature detection on x86_64 (maybe fixes #108)
41 - automatic stacktrace mechanism is now turned off by default and can
42 be switched on by command line option "--stacktrace" (the automatic
43 stacktrace mechanism seems to be broken on most systems at the moment)
45 * audio driver:
46 - removed the nonsense audio channel constraint (which was hard coded to
47 max. 100 audio channels) for most audio drivers
48 - JACK audio driver did not offer a device parameter "SAMPLERATE" as
49 opposed to the LSCP specs
50 - bugfix: the SAMPLERATE parameter of some drivers (e.g. JACK)
51 reflected the wrong value
52 - fixed a memory management bug in ASIO driver
53 - Makefile fix: JACK_CFLAGS wasn't used
54 - JACK: use jack_client_open instead of the deprecated
55 jack_client_new
56 - added (experimental) CoreAudio driver
58 * MIDI driver:
59 - added JACK MIDI driver
60 - dispatch bank select as ordinary CC as well, the user might seriously
61 want to (mis)use it for some purpose ("fixed" in all current MIDI
62 input drivers: ALSA, CoreMIDI, JACK, MidiShare, MME)
64 * instruments database:
65 - avoid time consuming samples scanning when adding instruments
66 to the instruments database
67 - added support for handling lost files in the instruments database
68 - Implemented option for adding instruments in separate directories
69 in the instruments database
70 (patch by Chris Cherrett & Andrew Williams, a bit adjusted)
72 * Gigasampler format engine:
73 - fixed a memory leak that could happen when a channel was deleted
74 while notes were playing
75 - made it possible to load an instrument even if the audio thread
76 isn't running
77 - added partial support for the "Controller Triggered" MIDI rule,
78 enough for piano gigs with pedal noise samples
79 - minor fix: only mark FX sends as being modified if really the
80 respective FX send MIDI controller was used
81 - added support for GM global device volume SysEx message
82 (can be selected at compile time to either apply globally to the
83 whole sampler [default] or only to the sampler channels that are
84 connected to the respective MIDI input port the SysEx message arrived
85 on)
86 - bugfix: notes triggered at position 0 in the audio buffer were
87 sometimes wrongly killed in the same buffer, causing no sound to
88 be played
89 - added support for chromatic / drumkit mode Roland GS Sysex message
90 which is usally used to switch a part between chromatic sounds and
91 drumkit sounds (as opposed to their common default setting of MIDI
92 part 10 being a drumkit part and all other ones chromatic parts), in
93 LS however one can switch between the first 16 MIDI instrument maps
94 defined for the sampler
95 - made it possible to create multiple sample channels even if the
96 audio thread isn't running
97 - fixed crash when removing channel with active voices (#116)
99 * LSCP server:
100 - added new LSCP event "CHANNEL_MIDI" which can be used by frontends to
101 react on MIDI data arriving on certain sampler channels (so far only
102 Note-On and Note-Off events are sent via this LSCP event)
103 - added new LSCP event "DEVICE_MIDI" which can be used by frontends to
104 react on MIDI data arriving on certain MIDI input devices (so far only
105 Note-On and Note-Off events are sent via this LSCP event)
106 - added new LSCP commands: FIND LOST DB_INSTRUMENT_FILES and
108 - added new LSCP command: SEND CHANNEL MIDI_DATA which can be used by
109 frontends to send MIDI messages to specific sampler channel
110 - added two additional fields to GET FILE INSTRUMENT INFO command -
112 - bugfix: the bank number provided by MIDI_INSTRUMENT_INFO notifications
113 was incorrect
114 - Optimized the retrieval of the MIDI instrument mappings
115 - added new LSCP command "SET VOICES" to globally alter the maximum
116 amount of voices, added new LSCP event "GLOBAL_INFO:VOICES" which will
117 be triggered respectively
118 - added new LSCP command "SET STREAMS" to globally alter the maximum
119 amount of disk streams, added new LSCP event "GLOBAL_INFO:STREAMS"
120 which will be triggered respectively
121 - bugfix: retry if "select" returns EINTR (this fixes a crash when
122 a gigedit file dialog is opened)
124 * Bug fixes:
125 - fixed a crash which occurs when removing a sampler channel waiting
126 to start instrument loading after another channel
127 - fixed a crash which occurs when removing a sampler channel with
128 instrument loading in progress (bug #113)
129 - fixed termination caused by uncaught exception when adding MIDI
130 instrument with PERSISTENT load mode
131 - fixed possible iterator invalidations when resetting the sampler
132 - fixed memory leaks when issuing the following LSCP commands:
137 - fixed possible compilation error when sqlite is not present
138 - fixed orphaned pointers when setting maximum voices limit (bug #118)
139 - fixed crash when changing the audio output device of a sampler
140 channel with loaded instrument and start playing notes
143 Version 0.5.1 (6 December 2007)
145 * packaging changes:
146 - added autoconf checks for pthread library
147 - added autoconf check for pthread bug found on certain NPTL-enabled
148 glibc versions (see Gentoo bug report #194076)
149 - added autoconf checks for MS Windows
151 * general changes:
152 - bugfix: the thread used by an editor plugin didn't die when the
153 editor closed
154 - bugfix: Ringbuffer.h: fill_write_space_with_null() did not zero
155 out all the space. operator--() did not apply size_mask after
156 decrementing the read_ptr. DEFAULT_WRAP_ELEMENTS set to 0 in
157 order to avoid problems with the _NonVolatileReader functions.
158 - bugfix: Stream.h: added a missing divide by BytesPerSample in
159 GetWriteSpace(). Since this function is currently only used in
160 the stream's qsort() compare function, it didn't trigger any
161 bugs.
162 - Resampler.h, Synthesizer.h: cubic interpolation now works in
163 24bit mode too. Faster method to read 24bit words on little
164 endian machines (x86): replaced 3 byte reads + shifts with a 1
165 unaligned 32bit read and shift
166 - experimental support for MS Windows (MIDI input via MME, AUDIO
167 output via ASIO)
168 - made handling of SIGINT signal (Ctrl-C) a bit more robust
169 - support for monitoring the total number of active disk streams
170 (new LSCP commands: GET TOTAL_STREAM_COUNT,
173 * AUDIO driver:
174 - added Windows ASIO low latency audio driver
176 * MIDI driver:
177 - added MME Windows MIDI driver
179 * LSCP server:
180 - added support for Windows style path / filenames, however with
181 forward slash path separators instead of backslash
182 (i.e. "C:/foo/bar.gig")
183 - allow naughty liblscp to send non-string device parameters within
184 apostrophes as well
185 - added new LSCP commands: "GET FILE INSTRUMENTS <file>",
186 "LIST FILE INSTRUMENTS <file>" and
187 "GET FILE INSTRUMENT INFO <file> <index>" for retrieving informations
188 about an arbitrary instrument file on the system where the sampler is
189 running on
191 Version 0.5.0 (15 October 2007)
193 * packaging changes:
194 - config.h is not going to be installed along with liblinuxsampler's
195 API header files anymore
196 - only the API relevant header (and source) files will be exported to
197 the Doxygen API documentation (explicitly listed in Doxyfile.in)
198 - added completely new XCode project files for Mac OSX which is now
199 capable to execute our autoconf environment, thus no need anymore
200 to maintain the compile time configuration file (osx/version.h) for
201 OSX manually (patch by Toshi Nagata)
202 - fixed buggy boolean --enable-foo / --disable-foo configure script
203 parameters
204 - global.h now only covers global definitions that are needed for the
205 C++ API header files, all implementation internal global definitions
206 are now in global_private.h
207 - atomic.h is not exposed to the C++ API anymore
208 - no need to include config.h anymore for using LS's API header files
209 - fixed warnings in API doc generation
211 * general changes:
212 - replaced the old, confusing MIDI program change mechanism by a
213 flexible MIDI instrument mapper which allows to map arbitrary
214 (MIDI bank, MIDI program) pairs with arbitrary
215 (engine type, instrument file, file index) triplets which will be
216 loaded on the respective channel when such MIDI program change
217 messages arrive, beside that, each entry allows to define a life-time
218 strategy for the instrument, so the user can define whether the
219 instrument should i.e. be permanently loaded (to be able to switch
220 quickly among sounds) or i.e. loaded just on demand when the
221 respective program change arrives (to safe RAM space), as well as a
222 global volume factor for each entry, so the user can adjust the volume
223 dynamic of his mapped instrument collection without having to modify
224 the instrument files, also one can manage arbitrary amount of such
225 MIDI instrument maps and assign each sampler channel individually
226 a certain map, so that i.e. separation of normal instruments and
227 drumkits is possible
228 - new notification events for tracking changes to audio/MIDI devices,
229 MIDI instrument maps, MIDI instruments, FX sends, global volume.
230 - sampler was limited to load max. 200 instruments in the background
231 due to a constant size RingBuffer FIFO which is now replaced by a
232 dynamic (unlimited) size std::list FIFO
233 - added FX sends, these allow to route audio signals to arbitrary audio
234 output channels for being processed by external effect processors
235 (i.e. jack-rack), the send levels are controllable via arbitrary MIDI
236 controllers
237 - global (sampler wide) volume can now be controlled at runtime
238 - Implemented new, improved notification system
239 - fixed compilation errors regarding OSX
240 (patch by Toshi Nagata)
241 - implemented instruments database
242 - added support for escape sequences to the instruments database
243 - added highly experimental support for on-the-fly instrument editing
244 within the sampler's process (by using instrument editor plugins),
245 you'll notice the new "Registered instrument editors:" message on
246 startup, added a new LSCP command:
247 "EDIT CHANNEL INSTRUMENT <sampler-channel>"
248 to spawn a matching instrument editor for the instrument on the
249 given sampler channel, the plugin path can be overridden at compile
250 time with ./configure --enable-plugin-dir=/some/dir
251 - added experimental code for synchronizing instrument editors hosted
252 in the sampler's process to safely edit instruments while playing
253 without a crash (hopefully) by either suspending single regions
254 wherever possible or if unavoidable whole engine(s)
255 - fixed several issues in fundamental "Thread" class: set scheduling
256 policy and priority on thread level, set a minimum stack size for
257 thread (TODO: a reasonable value yet to be tested), bugfix: non-RT
258 threads simply inherited properties of starting thread instead of
259 setting their own policy and priority
260 - minor fix in our automatic stack trace mechanism on crashes, the
261 main process did not wait for the stack trace process to finish
262 its output
263 - fixed some minor memory leaks
264 - reenabled assembly features support, at the moment only for
265 enabling a fast denormal FPU mode (x86 platforms supporting SSE2)
266 - minor assembly fix in x86 features detection (don't use the PIC
267 register, to avoid relocations in the text segment at runtime)
268 - POSIX callback functions of Thread.h are hidden
270 * MIDI driver:
271 - dispatch bank select (MSB and LSB) messages
273 * audio driver:
274 - the ALSA audio output driver parameters now reflect the correct
275 parameter value ranges for the respective selected sound card
276 (patch by Till Wimmer, a bit fixed and extended)
278 * Gigasampler format engine:
279 - if a filter is used and EG2 finishes before EG1, let the voice
280 die when EG2 has finished (this fixes a problem with clicks and
281 voice starvation for some gigs)
282 - playback is no longer disabled during instrument loading
283 - all notes playing on a channel that changes its instrument keep
284 playing with the old instrument until they get a note off
285 command
286 - EG fix: a release value of zero could cause noises or crash
287 - handle MIDI coarse tuning messages (MIDI RPN #0 MSB #2 LSB)
288 - EG fine tuning: when attack is zero the EG starts at a level
289 above max sustain level, which means that there is a decay phase
290 even if sustain is 100%
291 - more EG fixes: the level could sometimes go below zero and cause
292 noises or crashes
293 - minor fix of EGDecay (patch by Toshi Nagata)
294 - fixed compiler error when --enable-override-filter-type was
295 supplied to the configure script (fixes #46)
296 - disk thread: queue sizes are now proportional to CONFIG_MAX_STREAMS
297 instead of a fix value
298 - behavior fix: on MIDI CC# 65 (portamento on / off), 126 (mono mode),
299 127 (solo mode) only kill voices if the respective mode really
300 changed
302 * LSCP server:
303 - fixed compile time error for old Bison versions
304 (i.e. v1.28 found on MacOS 10.4, patch by Toshi Nagata)
305 - parser now supports extended ASCII character set
306 (up to ASCII code 255, i.e. includes now umlauts and accents)
307 - filename arguments in LSCP commands now allow to use escape
308 sequences, that is directly literal as one of: \', \", \\, \n, \r,
309 \f, \t, \v, or as octal ASCII code value like \132, or as hex ASCII
310 code value like \xf2) (fixes bug #24)
311 - the following LSCP commands now also support escape sequences for at
312 least one of their text-based parameters (i.e. name, description):
319 - returns verbose syntax errors (line and column where syntax error
320 occured, the unexpected character and the actually expected, possible
321 character(s), the latter only if less than 5 possibilities)
322 - made sure that LSCP syntax is not affected by gigedit locale
323 settings
324 - bugfix regarding strings parameter lists: all comma separated lists
325 of strings were treated as being one string containing commas
326 (fixes #57)
328 * Bug fixes:
329 - fixed crash occurring on certain LSCP scripts (Bug 39)
330 - another thread safety fix for lscp "load engine" and "set
331 channel audio output device"
332 - fixed a crash which occurs when reassigning the same engine
333 on a sampler channel with connected MIDI device
334 - fixed a crash which occurs when changing the number of ports of a MIDI
335 device connected to a sampler channel to number less then or equal
336 to the index of the port to which the sampler channel is connected.
337 - The previous bindings were not been disconnected when altering
338 the ALSA_SEQ_BINDINGS parameter. Introduced a NONE keyword for
339 unsubscribing from all bindings (e.g. ALSA_SEQ_BINDINGS=NONE).
340 - The active stream/voice count statistic was incorrect.
341 - notification events were not been sent for some sampler
342 channel changes
343 - added default min and max values to restrict the number of allowed
344 audio output channels and MIDI input ports
345 - the connection to the PCM interface is now closed when destroying
346 an audio output device
347 - files with slash in their path or filename could not be loaded
349 * test cases:
350 - updated and fixed (haven't been touched in a while)
352 Version 0.4.0 (24 November 2006)
354 * packaging changes:
355 - changed deprecated copyright attribute to license; added ldconfig
356 to post-(un)install steps; added devel package for liblinuxsampler;
357 to linuxsampler.spec (RPM)
358 - install necessary development header files for allowing 3rd party
359 applications to link against liblinuxsampler
360 - liblinuxsampler's API documentation can be generated with 'make docs'
361 (Doxygen required)
362 - added benchmark to automatically detect the best triangle LFO
363 implementation (currently either an integer math solution or a
364 di-harmonic approximation), automatic detection can be overriden
365 with --enable-signed-triang-algo=x and --enable-unsigned-triang-algo=x
366 configure script argument though (mandatory for cross-compilation)
367 - do not automatically pick optimized gcc flags if the user already
368 provided some on his own (as CXXFLAGS)
369 - added compile time option to disable processing of All-Notes-Off MIDI
370 messages
371 - added compile time options to allow disabling the various audio and
372 MIDI drivers
373 - fixed automatic GCC CPU switch detection on PPC
374 (patch by Ebrahim Mayat)
376 * Gigasampler format engine:
377 - extensive synthesis optimization
378 (reimplementation of EGs and LFO(s), removed synthesis parameter
379 prerendering and the synthesis parameter matrix in general, splitting
380 each audio fragment into subfragments now where each subfragment uses
381 constant pitch and filter coefficients. The volume coefficient is
382 linearly interpolated inside a subfragment, unless
383 --disable-interpolate-volume is set.)
384 - fine tuning of the EG modulation parameters
385 - improved filter cutoff calculation by adding support for the
386 following gig parameters: Cutoff freq (used when no cutoff
387 controller is defined), Control invert, Minimum cutoff, Velocity
388 curve and Velocity range. The keyboard tracking now scales
389 cutoff frequency instead of resonance.
390 - added support for gig parameter Resonance.
391 - fixed bug in sysex handling (patch by Juan Linietsky)
392 - added global volume attenuation of -9 dB (0.35f) to prevent clipping
393 which can be overridden with --enable-global-attenuation
394 - EG fixes: made the length of "attack hold" stage more
395 accurate. Release stage can now start before attack stage
396 ends. Cancel release didn't work when sustain was zero. Attack
397 time now has a minimal value to prevent clicks.
398 - fixed pitch changes larger than one octave
399 - fixed EG3 (pitch envelope) synthesis which was neutral all the time
400 - implemented portamento mode and solo mode (a.k.a 'mono mode'):
401 all modes can be altered via standard GM messages, that is CC5 for
402 altering portamento time, CC65 for enabling / disabling portamento
403 mode, CC126 for enabling solo mode and CC127 for disabling solo mode
404 - fine tuning of the curves for volume (CC7), pan (CC10 and gig
405 parameter) and crossfade
406 - added support for the "attenuation controller threshold" gig
407 parameter
408 - added smoothing of volume changes caused by control change
409 messages
410 - sample loop parameters are now taken from the DimensionRegion
411 instead of the wave chunk
412 - fixed keyswitching for v3 gigs with a number of keyswitch splits
413 not equal to a power of two
414 - reimplementation of the filter algorithm. The new filter is more
415 accurate and supports all gig filter types, including bandreject
416 and lowpass turbo.
417 - real support for 24 bit samples - samples are not truncated to
418 16 bits anymore
419 - support for aftertouch (channel pressure, not polyphonic
420 aftertouch)
421 - LFO1 behaviour fixed (dampening from max volume instead of
422 amplifying from 0)
424 * LSCP server:
425 - fixed application exit on broken pipe error (fixes bug #20)
426 - fixed the notification messages delay due to lack of
427 network activity (fixes bug #26)
428 - fixed parser bug which occured on space(s) within device parameters
431 * audio driver:
432 - added aRts audio output driver (by no means RT safe)
434 * MIDI driver:
435 - fixed legacy sysex code which caused dispatching of MIDI SysEx
436 messages several times instead of once
437 - API extension for MIDI drivers which already supply exact time stamps
438 for events (i.e. for offline rendering based MIDI drivers)
439 - added checks for bad MIDI data
441 * linuxsampler application:
442 - show available sampler engine types on startup
444 * general changes:
445 - support for muting sampler channels and solo mode of the same, two new
446 LSCP commands ("SET CHANNEL MUTE" and "SET CHANNEL SOLO") and two new
447 fields ("MUTE" and "SOLO") for command "GET CHANNEL INFO" were
448 introduced for this, the behavior is the same like on a mixer console
449 (patch by Grigor Iliev, a bit adjusted). Also added configure option
450 --enable-process-muted-channels which can be used to enable the
451 processing of muted channels.
452 - support for sostenuto pedal
453 - support for monitoring the total number of active voices
456 - fixed some memory management errors
457 - fixed some concurrency problems that could lead to crashes when
458 LSCP commands were executed
459 - fixed crash when instrument loading failed previously on the same
460 sampler channel (fixes bug #36)
462 Version 0.3.3 (15 July 2005)
464 * packaging changes:
465 - fixed compilation with gcc 4.0
467 * Gigasampler format engine:
468 - LFO preprocessing was performed even though the respective LFO was
469 disabled by the instrument patch (minor efficiency fix)
470 - if period time of chosen audio device is too small (< MIN_RELEASE_TIME)
471 for volume ramp downs in the same fragment (needed for current voice
472 stealing implementation) then simply inform the user with a warning
473 about possible click sounds and reduce the volume ramp down
474 appropriately instead of cancelling the audio device connection
476 Version 0.3.2 (24 June 2005)
478 * packaging changes:
479 - updated autotools build files to compile on OS X
480 (conditional compilation of CoreMIDI and MidiShare drivers)
481 - hand-crafted assembly optimization code can be disabled with
482 './configure --disable-asm' (definitely not recommended)
483 - fixed 'make dist' rule to include all necessary files
484 - require automake (>= 1.5) for 'make -f Makefile.cvs'
485 (due to 'dist-bzip2' automake option)
487 * Gigasampler format engine:
488 - support for the gig parameters for "release velocity response" curves
489 - fine tuning of the EGADSR envelope
490 - volume of release triggered samples now depends on note-on velocity,
491 note length and gig parameter "release trigger decay" instead of
492 note-off velocity.
493 - revised voice stealing
494 (fixes crash and endless loop caused by voice stealing)
495 - don't reset scale tuning on instrument or audio output device change
496 - handle key group conflicts right at the beginning of each fragment
497 instead of when the respective voice is actually launched
498 (fixes undefined behavior if stolen voices belonged to a key group -
499 this case was followed by a "killed voice survived" error message)
500 - fixed minor issue with null/silence samples
501 (those stole voices even though they don't need a voice at all which
502 resulted in "voice stealing didn't work out" messages)
503 - don't reset volume, pan, pitch and MIDI controller values on
504 instrument or audio output device change
506 * LSCP server:
507 - fixed some crashes (patch by Grigor Iliev, fixes #19)
508 - fixed LSCP event "CHANNEL_INFO" notification
509 (e.g. did not notify on volume changes or MIDI program change events)
511 * linuxsampler application:
512 - added command line parameters --lscp-addr and --lscp-port to override
513 default IP address and port of LSCP server
515 Version 0.3.1 (24 May 2005)
517 * initial release

  ViewVC Help
Powered by ViewVC