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

Contents of /linuxsampler/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1601 - (show annotations) (download)
Sat Dec 29 20:27:53 2007 UTC (16 years, 3 months ago) by schoenebeck
File size: 19583 byte(s)
* bugfix: on some POSIX systems instrument editor plugins refused to
  load as we used a non-portable Linux specific struct field
  (fixes bug #70, patch by Ronald Baljeu)

1 Version CVS HEAD (?)
2
3 * general changes:
4 - bugfix: on some POSIX systems instrument editor plugins refused to
5 load as we used a non-portable Linux specific struct field
6 (fixes bug #70, patch by Ronald Baljeu)
7
8 * audio driver:
9 - removed the nonsense audio channel constraint (which was hard coded to
10 max. 100 audio channels) for most audio drivers
11
12 Version 0.5.1 (6 December 2007)
13
14 * packaging changes:
15 - added autoconf checks for pthread library
16 - added autoconf check for pthread bug found on certain NPTL-enabled
17 glibc versions (see Gentoo bug report #194076)
18 - added autoconf checks for MS Windows
19
20 * general changes:
21 - bugfix: the thread used by an editor plugin didn't die when the
22 editor closed
23 - bugfix: Ringbuffer.h: fill_write_space_with_null() did not zero
24 out all the space. operator--() did not apply size_mask after
25 decrementing the read_ptr. DEFAULT_WRAP_ELEMENTS set to 0 in
26 order to avoid problems with the _NonVolatileReader functions.
27 - bugfix: Stream.h: added a missing divide by BytesPerSample in
28 GetWriteSpace(). Since this function is currently only used in
29 the stream's qsort() compare function, it didn't trigger any
30 bugs.
31 - Resampler.h, Synthesizer.h: cubic interpolation now works in
32 24bit mode too. Faster method to read 24bit words on little
33 endian machines (x86): replaced 3 byte reads + shifts with a 1
34 unaligned 32bit read and shift
35 - experimental support for MS Windows (MIDI input via MME, AUDIO
36 output via ASIO)
37 - made handling of SIGINT signal (Ctrl-C) a bit more robust
38 - support for monitoring the total number of active disk streams
39 (new LSCP commands: GET TOTAL_STREAM_COUNT,
40 SUBSCRIBE TOTAL_STREAM_COUNT, UNSUBSCRIBE TOTAL_STREAM_COUNT)
41
42 * AUDIO driver:
43 - added Windows ASIO low latency audio driver
44
45 * MIDI driver:
46 - added MME Windows MIDI driver
47
48 * LSCP server:
49 - added support for Windows style path / filenames, however with
50 forward slash path separators instead of backslash
51 (i.e. "C:/foo/bar.gig")
52 - allow naughty liblscp to send non-string device parameters within
53 apostrophes as well
54 - added new LSCP commands: "GET FILE INSTRUMENTS <file>",
55 "LIST FILE INSTRUMENTS <file>" and
56 "GET FILE INSTRUMENT INFO <file> <index>" for retrieving informations
57 about an arbitrary instrument file on the system where the sampler is
58 running on
59
60 Version 0.5.0 (15 October 2007)
61
62 * packaging changes:
63 - config.h is not going to be installed along with liblinuxsampler's
64 API header files anymore
65 - only the API relevant header (and source) files will be exported to
66 the Doxygen API documentation (explicitly listed in Doxyfile.in)
67 - added completely new XCode project files for Mac OSX which is now
68 capable to execute our autoconf environment, thus no need anymore
69 to maintain the compile time configuration file (osx/version.h) for
70 OSX manually (patch by Toshi Nagata)
71 - fixed buggy boolean --enable-foo / --disable-foo configure script
72 parameters
73 - global.h now only covers global definitions that are needed for the
74 C++ API header files, all implementation internal global definitions
75 are now in global_private.h
76 - atomic.h is not exposed to the C++ API anymore
77 - no need to include config.h anymore for using LS's API header files
78 - fixed warnings in API doc generation
79
80 * general changes:
81 - replaced the old, confusing MIDI program change mechanism by a
82 flexible MIDI instrument mapper which allows to map arbitrary
83 (MIDI bank, MIDI program) pairs with arbitrary
84 (engine type, instrument file, file index) triplets which will be
85 loaded on the respective channel when such MIDI program change
86 messages arrive, beside that, each entry allows to define a life-time
87 strategy for the instrument, so the user can define whether the
88 instrument should i.e. be permanently loaded (to be able to switch
89 quickly among sounds) or i.e. loaded just on demand when the
90 respective program change arrives (to safe RAM space), as well as a
91 global volume factor for each entry, so the user can adjust the volume
92 dynamic of his mapped instrument collection without having to modify
93 the instrument files, also one can manage arbitrary amount of such
94 MIDI instrument maps and assign each sampler channel individually
95 a certain map, so that i.e. separation of normal instruments and
96 drumkits is possible
97 - new notification events for tracking changes to audio/MIDI devices,
98 MIDI instrument maps, MIDI instruments, FX sends, global volume.
99 - sampler was limited to load max. 200 instruments in the background
100 due to a constant size RingBuffer FIFO which is now replaced by a
101 dynamic (unlimited) size std::list FIFO
102 - added FX sends, these allow to route audio signals to arbitrary audio
103 output channels for being processed by external effect processors
104 (i.e. jack-rack), the send levels are controllable via arbitrary MIDI
105 controllers
106 - global (sampler wide) volume can now be controlled at runtime
107 - Implemented new, improved notification system
108 - fixed compilation errors regarding OSX
109 (patch by Toshi Nagata)
110 - implemented instruments database
111 - added support for escape sequences to the instruments database
112 - added highly experimental support for on-the-fly instrument editing
113 within the sampler's process (by using instrument editor plugins),
114 you'll notice the new "Registered instrument editors:" message on
115 startup, added a new LSCP command:
116 "EDIT CHANNEL INSTRUMENT <sampler-channel>"
117 to spawn a matching instrument editor for the instrument on the
118 given sampler channel, the plugin path can be overridden at compile
119 time with ./configure --enable-plugin-dir=/some/dir
120 - added experimental code for synchronizing instrument editors hosted
121 in the sampler's process to safely edit instruments while playing
122 without a crash (hopefully) by either suspending single regions
123 wherever possible or if unavoidable whole engine(s)
124 - fixed several issues in fundamental "Thread" class: set scheduling
125 policy and priority on thread level, set a minimum stack size for
126 thread (TODO: a reasonable value yet to be tested), bugfix: non-RT
127 threads simply inherited properties of starting thread instead of
128 setting their own policy and priority
129 - minor fix in our automatic stack trace mechanism on crashes, the
130 main process did not wait for the stack trace process to finish
131 its output
132 - fixed some minor memory leaks
133 - reenabled assembly features support, at the moment only for
134 enabling a fast denormal FPU mode (x86 platforms supporting SSE2)
135 - minor assembly fix in x86 features detection (don't use the PIC
136 register, to avoid relocations in the text segment at runtime)
137 - POSIX callback functions of Thread.h are hidden
138
139 * MIDI driver:
140 - dispatch bank select (MSB and LSB) messages
141
142 * audio driver:
143 - the ALSA audio output driver parameters now reflect the correct
144 parameter value ranges for the respective selected sound card
145 (patch by Till Wimmer, a bit fixed and extended)
146
147 * Gigasampler format engine:
148 - if a filter is used and EG2 finishes before EG1, let the voice
149 die when EG2 has finished (this fixes a problem with clicks and
150 voice starvation for some gigs)
151 - playback is no longer disabled during instrument loading
152 - all notes playing on a channel that changes its instrument keep
153 playing with the old instrument until they get a note off
154 command
155 - EG fix: a release value of zero could cause noises or crash
156 - handle MIDI coarse tuning messages (MIDI RPN #0 MSB #2 LSB)
157 - EG fine tuning: when attack is zero the EG starts at a level
158 above max sustain level, which means that there is a decay phase
159 even if sustain is 100%
160 - more EG fixes: the level could sometimes go below zero and cause
161 noises or crashes
162 - minor fix of EGDecay (patch by Toshi Nagata)
163 - fixed compiler error when --enable-override-filter-type was
164 supplied to the configure script (fixes #46)
165 - disk thread: queue sizes are now proportional to CONFIG_MAX_STREAMS
166 instead of a fix value
167 - behavior fix: on MIDI CC# 65 (portamento on / off), 126 (mono mode),
168 127 (solo mode) only kill voices if the respective mode really
169 changed
170
171 * LSCP server:
172 - fixed compile time error for old Bison versions
173 (i.e. v1.28 found on MacOS 10.4, patch by Toshi Nagata)
174 - parser now supports extended ASCII character set
175 (up to ASCII code 255, i.e. includes now umlauts and accents)
176 - filename arguments in LSCP commands now allow to use escape
177 sequences, that is directly literal as one of: \', \", \\, \n, \r,
178 \f, \t, \v, or as octal ASCII code value like \132, or as hex ASCII
179 code value like \xf2) (fixes bug #24)
180 - the following LSCP commands now also support escape sequences for at
181 least one of their text-based parameters (i.e. name, description):
182 "ADD MIDI_INSTRUMENT_MAP", "MAP MIDI_INSTRUMENT",
183 "SET MIDI_INSTRUMENT_MAP NAME", "SET FX_SEND NAME", "CREATE FX_SEND",
184 "SET DB_INSTRUMENT_DIRECTORY NAME",
185 "SET DB_INSTRUMENT_DIRECTORY DESCRIPTION", "SET DB_INSTRUMENT NAME",
186 "SET DB_INSTRUMENT DESCRIPTION", "FIND DB_INSTRUMENTS",
187 "FIND DB_INSTRUMENT_DIRECTORIES"
188 - returns verbose syntax errors (line and column where syntax error
189 occured, the unexpected character and the actually expected, possible
190 character(s), the latter only if less than 5 possibilities)
191 - made sure that LSCP syntax is not affected by gigedit locale
192 settings
193 - bugfix regarding strings parameter lists: all comma separated lists
194 of strings were treated as being one string containing commas
195 (fixes #57)
196
197 * Bug fixes:
198 - fixed crash occurring on certain LSCP scripts (Bug 39)
199 - another thread safety fix for lscp "load engine" and "set
200 channel audio output device"
201 - fixed a crash which occurs when reassigning the same engine
202 on a sampler channel with connected MIDI device
203 - fixed a crash which occurs when changing the number of ports of a MIDI
204 device connected to a sampler channel to number less then or equal
205 to the index of the port to which the sampler channel is connected.
206 - The previous bindings were not been disconnected when altering
207 the ALSA_SEQ_BINDINGS parameter. Introduced a NONE keyword for
208 unsubscribing from all bindings (e.g. ALSA_SEQ_BINDINGS=NONE).
209 - The active stream/voice count statistic was incorrect.
210 - notification events were not been sent for some sampler
211 channel changes
212 - added default min and max values to restrict the number of allowed
213 audio output channels and MIDI input ports
214 - the connection to the PCM interface is now closed when destroying
215 an audio output device
216 - files with slash in their path or filename could not be loaded
217
218 * test cases:
219 - updated and fixed (haven't been touched in a while)
220
221 Version 0.4.0 (24 November 2006)
222
223 * packaging changes:
224 - changed deprecated copyright attribute to license; added ldconfig
225 to post-(un)install steps; added devel package for liblinuxsampler;
226 to linuxsampler.spec (RPM)
227 - install necessary development header files for allowing 3rd party
228 applications to link against liblinuxsampler
229 - liblinuxsampler's API documentation can be generated with 'make docs'
230 (Doxygen required)
231 - added benchmark to automatically detect the best triangle LFO
232 implementation (currently either an integer math solution or a
233 di-harmonic approximation), automatic detection can be overriden
234 with --enable-signed-triang-algo=x and --enable-unsigned-triang-algo=x
235 configure script argument though (mandatory for cross-compilation)
236 - do not automatically pick optimized gcc flags if the user already
237 provided some on his own (as CXXFLAGS)
238 - added compile time option to disable processing of All-Notes-Off MIDI
239 messages
240 - added compile time options to allow disabling the various audio and
241 MIDI drivers
242 - fixed automatic GCC CPU switch detection on PPC
243 (patch by Ebrahim Mayat)
244
245 * Gigasampler format engine:
246 - extensive synthesis optimization
247 (reimplementation of EGs and LFO(s), removed synthesis parameter
248 prerendering and the synthesis parameter matrix in general, splitting
249 each audio fragment into subfragments now where each subfragment uses
250 constant pitch and filter coefficients. The volume coefficient is
251 linearly interpolated inside a subfragment, unless
252 --disable-interpolate-volume is set.)
253 - fine tuning of the EG modulation parameters
254 - improved filter cutoff calculation by adding support for the
255 following gig parameters: Cutoff freq (used when no cutoff
256 controller is defined), Control invert, Minimum cutoff, Velocity
257 curve and Velocity range. The keyboard tracking now scales
258 cutoff frequency instead of resonance.
259 - added support for gig parameter Resonance.
260 - fixed bug in sysex handling (patch by Juan Linietsky)
261 - added global volume attenuation of -9 dB (0.35f) to prevent clipping
262 which can be overridden with --enable-global-attenuation
263 - EG fixes: made the length of "attack hold" stage more
264 accurate. Release stage can now start before attack stage
265 ends. Cancel release didn't work when sustain was zero. Attack
266 time now has a minimal value to prevent clicks.
267 - fixed pitch changes larger than one octave
268 - fixed EG3 (pitch envelope) synthesis which was neutral all the time
269 - implemented portamento mode and solo mode (a.k.a 'mono mode'):
270 all modes can be altered via standard GM messages, that is CC5 for
271 altering portamento time, CC65 for enabling / disabling portamento
272 mode, CC126 for enabling solo mode and CC127 for disabling solo mode
273 - fine tuning of the curves for volume (CC7), pan (CC10 and gig
274 parameter) and crossfade
275 - added support for the "attenuation controller threshold" gig
276 parameter
277 - added smoothing of volume changes caused by control change
278 messages
279 - sample loop parameters are now taken from the DimensionRegion
280 instead of the wave chunk
281 - fixed keyswitching for v3 gigs with a number of keyswitch splits
282 not equal to a power of two
283 - reimplementation of the filter algorithm. The new filter is more
284 accurate and supports all gig filter types, including bandreject
285 and lowpass turbo.
286 - real support for 24 bit samples - samples are not truncated to
287 16 bits anymore
288 - support for aftertouch (channel pressure, not polyphonic
289 aftertouch)
290 - LFO1 behaviour fixed (dampening from max volume instead of
291 amplifying from 0)
292
293 * LSCP server:
294 - fixed application exit on broken pipe error (fixes bug #20)
295 - fixed the notification messages delay due to lack of
296 network activity (fixes bug #26)
297 - fixed parser bug which occured on space(s) within device parameters
298 (e.g. "SET AUDIO_OUTPUT_CHANNEL_PARAMETER 0 0 SOMEPARAM='foo bar'")
299
300 * audio driver:
301 - added aRts audio output driver (by no means RT safe)
302
303 * MIDI driver:
304 - fixed legacy sysex code which caused dispatching of MIDI SysEx
305 messages several times instead of once
306 - API extension for MIDI drivers which already supply exact time stamps
307 for events (i.e. for offline rendering based MIDI drivers)
308 - added checks for bad MIDI data
309
310 * linuxsampler application:
311 - show available sampler engine types on startup
312
313 * general changes:
314 - support for muting sampler channels and solo mode of the same, two new
315 LSCP commands ("SET CHANNEL MUTE" and "SET CHANNEL SOLO") and two new
316 fields ("MUTE" and "SOLO") for command "GET CHANNEL INFO" were
317 introduced for this, the behavior is the same like on a mixer console
318 (patch by Grigor Iliev, a bit adjusted). Also added configure option
319 --enable-process-muted-channels which can be used to enable the
320 processing of muted channels.
321 - support for sostenuto pedal
322 - support for monitoring the total number of active voices
323 (new LSCP commands: GET TOTAL_VOICE_COUNT, GET TOTAL_VOICE_COUNT_MAX,
324 SUBSCRIBE TOTAL_VOICE_COUNT, UNSUBSCRIBE TOTAL_VOICE_COUNT)
325 - fixed some memory management errors
326 - fixed some concurrency problems that could lead to crashes when
327 LSCP commands were executed
328 - fixed crash when instrument loading failed previously on the same
329 sampler channel (fixes bug #36)
330
331 Version 0.3.3 (15 July 2005)
332
333 * packaging changes:
334 - fixed compilation with gcc 4.0
335
336 * Gigasampler format engine:
337 - LFO preprocessing was performed even though the respective LFO was
338 disabled by the instrument patch (minor efficiency fix)
339 - if period time of chosen audio device is too small (< MIN_RELEASE_TIME)
340 for volume ramp downs in the same fragment (needed for current voice
341 stealing implementation) then simply inform the user with a warning
342 about possible click sounds and reduce the volume ramp down
343 appropriately instead of cancelling the audio device connection
344
345 Version 0.3.2 (24 June 2005)
346
347 * packaging changes:
348 - updated autotools build files to compile on OS X
349 (conditional compilation of CoreMIDI and MidiShare drivers)
350 - hand-crafted assembly optimization code can be disabled with
351 './configure --disable-asm' (definitely not recommended)
352 - fixed 'make dist' rule to include all necessary files
353 - require automake (>= 1.5) for 'make -f Makefile.cvs'
354 (due to 'dist-bzip2' automake option)
355
356 * Gigasampler format engine:
357 - support for the gig parameters for "release velocity response" curves
358 - fine tuning of the EGADSR envelope
359 - volume of release triggered samples now depends on note-on velocity,
360 note length and gig parameter "release trigger decay" instead of
361 note-off velocity.
362 - revised voice stealing
363 (fixes crash and endless loop caused by voice stealing)
364 - don't reset scale tuning on instrument or audio output device change
365 - handle key group conflicts right at the beginning of each fragment
366 instead of when the respective voice is actually launched
367 (fixes undefined behavior if stolen voices belonged to a key group -
368 this case was followed by a "killed voice survived" error message)
369 - fixed minor issue with null/silence samples
370 (those stole voices even though they don't need a voice at all which
371 resulted in "voice stealing didn't work out" messages)
372 - don't reset volume, pan, pitch and MIDI controller values on
373 instrument or audio output device change
374
375 * LSCP server:
376 - fixed some crashes (patch by Grigor Iliev, fixes #19)
377 - fixed LSCP event "CHANNEL_INFO" notification
378 (e.g. did not notify on volume changes or MIDI program change events)
379
380 * linuxsampler application:
381 - added command line parameters --lscp-addr and --lscp-port to override
382 default IP address and port of LSCP server
383
384 Version 0.3.1 (24 May 2005)
385
386 * initial release

  ViewVC Help
Powered by ViewVC