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

Contents of /linuxsampler/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1761 - (show annotations) (download)
Fri Aug 29 15:42:06 2008 UTC (15 years, 7 months ago) by iliev
File size: 23382 byte(s)
* fixed a crash which occurs when removing a sampler channel waiting
  to start instrument loading after another channel

1 Version CVS HEAD (?)
2
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
10 * general changes:
11 - bugfix: on some POSIX systems instrument editor plugins refused to
12 load as we used a non-portable Linux specific struct field
13 (fixes bug #70, patch by Ronald Baljeu)
14 - fixed endless loop which occured when loading an instrument editor
15 plugin DLL on Windows systems failed
16 - fixed memory leaks that occurred when liblinuxsampler was unloaded
17 - optimized the SynchronizedConfig class so it doesn't wait
18 unnecessarily long after an update
19 - added support for notifying instrument editors on note-on / note-off
20 events (e.g. to highlight the pressed keys on the virtual keyboard
21 of gigedit)
22 - added support for triggering notes by instrument editors (see above)
23 - be verbose on DLL load errors (on Linux)
24 - fixed di-harmonic triangle LFO implementation (this LFO implementation
25 is only used by very few systems, most pick the int math
26 implementation, which is usually faster)
27 - fixes for audio drivers with varying buffer sizes
28
29 * audio driver:
30 - removed the nonsense audio channel constraint (which was hard coded to
31 max. 100 audio channels) for most audio drivers
32 - JACK audio driver did not offer a device parameter "SAMPLERATE" as
33 opposed to the LSCP specs
34 - bugfix: the SAMPLERATE parameter of some drivers (e.g. JACK)
35 reflected the wrong value
36 - fixed a memory management bug in ASIO driver
37 - Makefile fix: JACK_CFLAGS wasn't used
38 - JACK: use jack_client_open instead of thedeprecated
39 jack_client_new
40
41 * MIDI driver:
42 - added JACK MIDI driver
43 - dispatch bank select as ordinary CC as well, the user might seriously
44 want to (mis)use it for some purpose ("fixed" in all current MIDI
45 input drivers: ALSA, CoreMIDI, JACK, MidiShare, MME)
46
47 * instruments database:
48 - avoid time consuming samples scanning when adding instruments
49 to the instruments database
50 - added support for handling lost files in the instruments database
51
52 * Gigasampler format engine:
53 - fixed a memory leak that could happen when a channel was deleted
54 while notes were playing
55 - made it possible to load an instrument even if the audio thread
56 isn't running
57 - added partial support for the "Controller Triggered" MIDI rule,
58 enough for piano gigs with pedal noise samples
59 - minor fix: only mark FX sends as being modified if really the
60 respective FX send MIDI controller was used
61 - added support for GM global device volume SysEx message
62 - bugfix: notes triggered at position 0 in the audio buffer were
63 sometimes wrongly killed in the same buffer, causing no sound to
64 be played
65 - added support for chromatic / drumkit mode Roland GS Sysex message
66 which is usally used to switch a part between chromatic sounds and
67 drumkit sounds (as opposed to their common default setting of MIDI
68 part 10 being a drumkit part and all other ones chromatic parts), in
69 LS however one can switch between the first 16 MIDI instrument maps
70 defined for the sampler
71
72 * LSCP server:
73 - added new LSCP event "CHANNEL_MIDI" which can be used by frontends to
74 react on MIDI data arriving on certain sampler channels (so far only
75 Note-On and Note-Off events are sent via this LSCP event)
76 - added new LSCP event "DEVICE_MIDI" which can be used by frontends to
77 react on MIDI data arriving on certain MIDI input devices (so far only
78 Note-On and Note-Off events are sent via this LSCP event)
79 - added new LSCP commands: FIND LOST DB_INSTRUMENT_FILES and
80 SET DB_INSTRUMENT FILE_PATH
81 - bugfix: the bank number provided by MIDI_INSTRUMENT_INFO notifications
82 was incorrect
83
84 * Bug fixes:
85 - fixed a crash which occurs when removing a sampler channel waiting
86 to start instrument loading after another channel
87
88
89 Version 0.5.1 (6 December 2007)
90
91 * packaging changes:
92 - added autoconf checks for pthread library
93 - added autoconf check for pthread bug found on certain NPTL-enabled
94 glibc versions (see Gentoo bug report #194076)
95 - added autoconf checks for MS Windows
96
97 * general changes:
98 - bugfix: the thread used by an editor plugin didn't die when the
99 editor closed
100 - bugfix: Ringbuffer.h: fill_write_space_with_null() did not zero
101 out all the space. operator--() did not apply size_mask after
102 decrementing the read_ptr. DEFAULT_WRAP_ELEMENTS set to 0 in
103 order to avoid problems with the _NonVolatileReader functions.
104 - bugfix: Stream.h: added a missing divide by BytesPerSample in
105 GetWriteSpace(). Since this function is currently only used in
106 the stream's qsort() compare function, it didn't trigger any
107 bugs.
108 - Resampler.h, Synthesizer.h: cubic interpolation now works in
109 24bit mode too. Faster method to read 24bit words on little
110 endian machines (x86): replaced 3 byte reads + shifts with a 1
111 unaligned 32bit read and shift
112 - experimental support for MS Windows (MIDI input via MME, AUDIO
113 output via ASIO)
114 - made handling of SIGINT signal (Ctrl-C) a bit more robust
115 - support for monitoring the total number of active disk streams
116 (new LSCP commands: GET TOTAL_STREAM_COUNT,
117 SUBSCRIBE TOTAL_STREAM_COUNT, UNSUBSCRIBE TOTAL_STREAM_COUNT)
118
119 * AUDIO driver:
120 - added Windows ASIO low latency audio driver
121
122 * MIDI driver:
123 - added MME Windows MIDI driver
124
125 * LSCP server:
126 - added support for Windows style path / filenames, however with
127 forward slash path separators instead of backslash
128 (i.e. "C:/foo/bar.gig")
129 - allow naughty liblscp to send non-string device parameters within
130 apostrophes as well
131 - added new LSCP commands: "GET FILE INSTRUMENTS <file>",
132 "LIST FILE INSTRUMENTS <file>" and
133 "GET FILE INSTRUMENT INFO <file> <index>" for retrieving informations
134 about an arbitrary instrument file on the system where the sampler is
135 running on
136
137 Version 0.5.0 (15 October 2007)
138
139 * packaging changes:
140 - config.h is not going to be installed along with liblinuxsampler's
141 API header files anymore
142 - only the API relevant header (and source) files will be exported to
143 the Doxygen API documentation (explicitly listed in Doxyfile.in)
144 - added completely new XCode project files for Mac OSX which is now
145 capable to execute our autoconf environment, thus no need anymore
146 to maintain the compile time configuration file (osx/version.h) for
147 OSX manually (patch by Toshi Nagata)
148 - fixed buggy boolean --enable-foo / --disable-foo configure script
149 parameters
150 - global.h now only covers global definitions that are needed for the
151 C++ API header files, all implementation internal global definitions
152 are now in global_private.h
153 - atomic.h is not exposed to the C++ API anymore
154 - no need to include config.h anymore for using LS's API header files
155 - fixed warnings in API doc generation
156
157 * general changes:
158 - replaced the old, confusing MIDI program change mechanism by a
159 flexible MIDI instrument mapper which allows to map arbitrary
160 (MIDI bank, MIDI program) pairs with arbitrary
161 (engine type, instrument file, file index) triplets which will be
162 loaded on the respective channel when such MIDI program change
163 messages arrive, beside that, each entry allows to define a life-time
164 strategy for the instrument, so the user can define whether the
165 instrument should i.e. be permanently loaded (to be able to switch
166 quickly among sounds) or i.e. loaded just on demand when the
167 respective program change arrives (to safe RAM space), as well as a
168 global volume factor for each entry, so the user can adjust the volume
169 dynamic of his mapped instrument collection without having to modify
170 the instrument files, also one can manage arbitrary amount of such
171 MIDI instrument maps and assign each sampler channel individually
172 a certain map, so that i.e. separation of normal instruments and
173 drumkits is possible
174 - new notification events for tracking changes to audio/MIDI devices,
175 MIDI instrument maps, MIDI instruments, FX sends, global volume.
176 - sampler was limited to load max. 200 instruments in the background
177 due to a constant size RingBuffer FIFO which is now replaced by a
178 dynamic (unlimited) size std::list FIFO
179 - added FX sends, these allow to route audio signals to arbitrary audio
180 output channels for being processed by external effect processors
181 (i.e. jack-rack), the send levels are controllable via arbitrary MIDI
182 controllers
183 - global (sampler wide) volume can now be controlled at runtime
184 - Implemented new, improved notification system
185 - fixed compilation errors regarding OSX
186 (patch by Toshi Nagata)
187 - implemented instruments database
188 - added support for escape sequences to the instruments database
189 - added highly experimental support for on-the-fly instrument editing
190 within the sampler's process (by using instrument editor plugins),
191 you'll notice the new "Registered instrument editors:" message on
192 startup, added a new LSCP command:
193 "EDIT CHANNEL INSTRUMENT <sampler-channel>"
194 to spawn a matching instrument editor for the instrument on the
195 given sampler channel, the plugin path can be overridden at compile
196 time with ./configure --enable-plugin-dir=/some/dir
197 - added experimental code for synchronizing instrument editors hosted
198 in the sampler's process to safely edit instruments while playing
199 without a crash (hopefully) by either suspending single regions
200 wherever possible or if unavoidable whole engine(s)
201 - fixed several issues in fundamental "Thread" class: set scheduling
202 policy and priority on thread level, set a minimum stack size for
203 thread (TODO: a reasonable value yet to be tested), bugfix: non-RT
204 threads simply inherited properties of starting thread instead of
205 setting their own policy and priority
206 - minor fix in our automatic stack trace mechanism on crashes, the
207 main process did not wait for the stack trace process to finish
208 its output
209 - fixed some minor memory leaks
210 - reenabled assembly features support, at the moment only for
211 enabling a fast denormal FPU mode (x86 platforms supporting SSE2)
212 - minor assembly fix in x86 features detection (don't use the PIC
213 register, to avoid relocations in the text segment at runtime)
214 - POSIX callback functions of Thread.h are hidden
215
216 * MIDI driver:
217 - dispatch bank select (MSB and LSB) messages
218
219 * audio driver:
220 - the ALSA audio output driver parameters now reflect the correct
221 parameter value ranges for the respective selected sound card
222 (patch by Till Wimmer, a bit fixed and extended)
223
224 * Gigasampler format engine:
225 - if a filter is used and EG2 finishes before EG1, let the voice
226 die when EG2 has finished (this fixes a problem with clicks and
227 voice starvation for some gigs)
228 - playback is no longer disabled during instrument loading
229 - all notes playing on a channel that changes its instrument keep
230 playing with the old instrument until they get a note off
231 command
232 - EG fix: a release value of zero could cause noises or crash
233 - handle MIDI coarse tuning messages (MIDI RPN #0 MSB #2 LSB)
234 - EG fine tuning: when attack is zero the EG starts at a level
235 above max sustain level, which means that there is a decay phase
236 even if sustain is 100%
237 - more EG fixes: the level could sometimes go below zero and cause
238 noises or crashes
239 - minor fix of EGDecay (patch by Toshi Nagata)
240 - fixed compiler error when --enable-override-filter-type was
241 supplied to the configure script (fixes #46)
242 - disk thread: queue sizes are now proportional to CONFIG_MAX_STREAMS
243 instead of a fix value
244 - behavior fix: on MIDI CC# 65 (portamento on / off), 126 (mono mode),
245 127 (solo mode) only kill voices if the respective mode really
246 changed
247
248 * LSCP server:
249 - fixed compile time error for old Bison versions
250 (i.e. v1.28 found on MacOS 10.4, patch by Toshi Nagata)
251 - parser now supports extended ASCII character set
252 (up to ASCII code 255, i.e. includes now umlauts and accents)
253 - filename arguments in LSCP commands now allow to use escape
254 sequences, that is directly literal as one of: \', \", \\, \n, \r,
255 \f, \t, \v, or as octal ASCII code value like \132, or as hex ASCII
256 code value like \xf2) (fixes bug #24)
257 - the following LSCP commands now also support escape sequences for at
258 least one of their text-based parameters (i.e. name, description):
259 "ADD MIDI_INSTRUMENT_MAP", "MAP MIDI_INSTRUMENT",
260 "SET MIDI_INSTRUMENT_MAP NAME", "SET FX_SEND NAME", "CREATE FX_SEND",
261 "SET DB_INSTRUMENT_DIRECTORY NAME",
262 "SET DB_INSTRUMENT_DIRECTORY DESCRIPTION", "SET DB_INSTRUMENT NAME",
263 "SET DB_INSTRUMENT DESCRIPTION", "FIND DB_INSTRUMENTS",
264 "FIND DB_INSTRUMENT_DIRECTORIES"
265 - returns verbose syntax errors (line and column where syntax error
266 occured, the unexpected character and the actually expected, possible
267 character(s), the latter only if less than 5 possibilities)
268 - made sure that LSCP syntax is not affected by gigedit locale
269 settings
270 - bugfix regarding strings parameter lists: all comma separated lists
271 of strings were treated as being one string containing commas
272 (fixes #57)
273
274 * Bug fixes:
275 - fixed crash occurring on certain LSCP scripts (Bug 39)
276 - another thread safety fix for lscp "load engine" and "set
277 channel audio output device"
278 - fixed a crash which occurs when reassigning the same engine
279 on a sampler channel with connected MIDI device
280 - fixed a crash which occurs when changing the number of ports of a MIDI
281 device connected to a sampler channel to number less then or equal
282 to the index of the port to which the sampler channel is connected.
283 - The previous bindings were not been disconnected when altering
284 the ALSA_SEQ_BINDINGS parameter. Introduced a NONE keyword for
285 unsubscribing from all bindings (e.g. ALSA_SEQ_BINDINGS=NONE).
286 - The active stream/voice count statistic was incorrect.
287 - notification events were not been sent for some sampler
288 channel changes
289 - added default min and max values to restrict the number of allowed
290 audio output channels and MIDI input ports
291 - the connection to the PCM interface is now closed when destroying
292 an audio output device
293 - files with slash in their path or filename could not be loaded
294
295 * test cases:
296 - updated and fixed (haven't been touched in a while)
297
298 Version 0.4.0 (24 November 2006)
299
300 * packaging changes:
301 - changed deprecated copyright attribute to license; added ldconfig
302 to post-(un)install steps; added devel package for liblinuxsampler;
303 to linuxsampler.spec (RPM)
304 - install necessary development header files for allowing 3rd party
305 applications to link against liblinuxsampler
306 - liblinuxsampler's API documentation can be generated with 'make docs'
307 (Doxygen required)
308 - added benchmark to automatically detect the best triangle LFO
309 implementation (currently either an integer math solution or a
310 di-harmonic approximation), automatic detection can be overriden
311 with --enable-signed-triang-algo=x and --enable-unsigned-triang-algo=x
312 configure script argument though (mandatory for cross-compilation)
313 - do not automatically pick optimized gcc flags if the user already
314 provided some on his own (as CXXFLAGS)
315 - added compile time option to disable processing of All-Notes-Off MIDI
316 messages
317 - added compile time options to allow disabling the various audio and
318 MIDI drivers
319 - fixed automatic GCC CPU switch detection on PPC
320 (patch by Ebrahim Mayat)
321
322 * Gigasampler format engine:
323 - extensive synthesis optimization
324 (reimplementation of EGs and LFO(s), removed synthesis parameter
325 prerendering and the synthesis parameter matrix in general, splitting
326 each audio fragment into subfragments now where each subfragment uses
327 constant pitch and filter coefficients. The volume coefficient is
328 linearly interpolated inside a subfragment, unless
329 --disable-interpolate-volume is set.)
330 - fine tuning of the EG modulation parameters
331 - improved filter cutoff calculation by adding support for the
332 following gig parameters: Cutoff freq (used when no cutoff
333 controller is defined), Control invert, Minimum cutoff, Velocity
334 curve and Velocity range. The keyboard tracking now scales
335 cutoff frequency instead of resonance.
336 - added support for gig parameter Resonance.
337 - fixed bug in sysex handling (patch by Juan Linietsky)
338 - added global volume attenuation of -9 dB (0.35f) to prevent clipping
339 which can be overridden with --enable-global-attenuation
340 - EG fixes: made the length of "attack hold" stage more
341 accurate. Release stage can now start before attack stage
342 ends. Cancel release didn't work when sustain was zero. Attack
343 time now has a minimal value to prevent clicks.
344 - fixed pitch changes larger than one octave
345 - fixed EG3 (pitch envelope) synthesis which was neutral all the time
346 - implemented portamento mode and solo mode (a.k.a 'mono mode'):
347 all modes can be altered via standard GM messages, that is CC5 for
348 altering portamento time, CC65 for enabling / disabling portamento
349 mode, CC126 for enabling solo mode and CC127 for disabling solo mode
350 - fine tuning of the curves for volume (CC7), pan (CC10 and gig
351 parameter) and crossfade
352 - added support for the "attenuation controller threshold" gig
353 parameter
354 - added smoothing of volume changes caused by control change
355 messages
356 - sample loop parameters are now taken from the DimensionRegion
357 instead of the wave chunk
358 - fixed keyswitching for v3 gigs with a number of keyswitch splits
359 not equal to a power of two
360 - reimplementation of the filter algorithm. The new filter is more
361 accurate and supports all gig filter types, including bandreject
362 and lowpass turbo.
363 - real support for 24 bit samples - samples are not truncated to
364 16 bits anymore
365 - support for aftertouch (channel pressure, not polyphonic
366 aftertouch)
367 - LFO1 behaviour fixed (dampening from max volume instead of
368 amplifying from 0)
369
370 * LSCP server:
371 - fixed application exit on broken pipe error (fixes bug #20)
372 - fixed the notification messages delay due to lack of
373 network activity (fixes bug #26)
374 - fixed parser bug which occured on space(s) within device parameters
375 (e.g. "SET AUDIO_OUTPUT_CHANNEL_PARAMETER 0 0 SOMEPARAM='foo bar'")
376
377 * audio driver:
378 - added aRts audio output driver (by no means RT safe)
379
380 * MIDI driver:
381 - fixed legacy sysex code which caused dispatching of MIDI SysEx
382 messages several times instead of once
383 - API extension for MIDI drivers which already supply exact time stamps
384 for events (i.e. for offline rendering based MIDI drivers)
385 - added checks for bad MIDI data
386
387 * linuxsampler application:
388 - show available sampler engine types on startup
389
390 * general changes:
391 - support for muting sampler channels and solo mode of the same, two new
392 LSCP commands ("SET CHANNEL MUTE" and "SET CHANNEL SOLO") and two new
393 fields ("MUTE" and "SOLO") for command "GET CHANNEL INFO" were
394 introduced for this, the behavior is the same like on a mixer console
395 (patch by Grigor Iliev, a bit adjusted). Also added configure option
396 --enable-process-muted-channels which can be used to enable the
397 processing of muted channels.
398 - support for sostenuto pedal
399 - support for monitoring the total number of active voices
400 (new LSCP commands: GET TOTAL_VOICE_COUNT, GET TOTAL_VOICE_COUNT_MAX,
401 SUBSCRIBE TOTAL_VOICE_COUNT, UNSUBSCRIBE TOTAL_VOICE_COUNT)
402 - fixed some memory management errors
403 - fixed some concurrency problems that could lead to crashes when
404 LSCP commands were executed
405 - fixed crash when instrument loading failed previously on the same
406 sampler channel (fixes bug #36)
407
408 Version 0.3.3 (15 July 2005)
409
410 * packaging changes:
411 - fixed compilation with gcc 4.0
412
413 * Gigasampler format engine:
414 - LFO preprocessing was performed even though the respective LFO was
415 disabled by the instrument patch (minor efficiency fix)
416 - if period time of chosen audio device is too small (< MIN_RELEASE_TIME)
417 for volume ramp downs in the same fragment (needed for current voice
418 stealing implementation) then simply inform the user with a warning
419 about possible click sounds and reduce the volume ramp down
420 appropriately instead of cancelling the audio device connection
421
422 Version 0.3.2 (24 June 2005)
423
424 * packaging changes:
425 - updated autotools build files to compile on OS X
426 (conditional compilation of CoreMIDI and MidiShare drivers)
427 - hand-crafted assembly optimization code can be disabled with
428 './configure --disable-asm' (definitely not recommended)
429 - fixed 'make dist' rule to include all necessary files
430 - require automake (>= 1.5) for 'make -f Makefile.cvs'
431 (due to 'dist-bzip2' automake option)
432
433 * Gigasampler format engine:
434 - support for the gig parameters for "release velocity response" curves
435 - fine tuning of the EGADSR envelope
436 - volume of release triggered samples now depends on note-on velocity,
437 note length and gig parameter "release trigger decay" instead of
438 note-off velocity.
439 - revised voice stealing
440 (fixes crash and endless loop caused by voice stealing)
441 - don't reset scale tuning on instrument or audio output device change
442 - handle key group conflicts right at the beginning of each fragment
443 instead of when the respective voice is actually launched
444 (fixes undefined behavior if stolen voices belonged to a key group -
445 this case was followed by a "killed voice survived" error message)
446 - fixed minor issue with null/silence samples
447 (those stole voices even though they don't need a voice at all which
448 resulted in "voice stealing didn't work out" messages)
449 - don't reset volume, pan, pitch and MIDI controller values on
450 instrument or audio output device change
451
452 * LSCP server:
453 - fixed some crashes (patch by Grigor Iliev, fixes #19)
454 - fixed LSCP event "CHANNEL_INFO" notification
455 (e.g. did not notify on volume changes or MIDI program change events)
456
457 * linuxsampler application:
458 - added command line parameters --lscp-addr and --lscp-port to override
459 default IP address and port of LSCP server
460
461 Version 0.3.1 (24 May 2005)
462
463 * initial release

  ViewVC Help
Powered by ViewVC