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

Contents of /linuxsampler/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1599 - (show annotations) (download)
Fri Dec 28 15:47:33 2007 UTC (16 years, 3 months ago) by schoenebeck
File size: 19377 byte(s)
* removed the nonsense audio channel constraint (which was hard coded to
  max. 100 audio channels) for most audio drivers

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

  ViewVC Help
Powered by ViewVC