ViewVC logotype

Contents of /linuxsampler/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 1800 - (show annotations) (download)
Sun Dec 7 01:26:46 2008 UTC (11 years, 9 months ago) by schoenebeck
File size: 25066 byte(s)
* maximum voices and disk streams can now be altered at runtime (added new
  and accordingly new LSCP events "GLOBAL_INFO:VOICES" and
* bumped version to

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

  ViewVC Help
Powered by ViewVC