ViewVC logotype

Contents of /linuxsampler/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 3988 - (show annotations) (download)
Wed Aug 4 18:23:05 2021 UTC (2 months, 2 weeks ago) by schoenebeck
File size: 103821 byte(s)
* AbstractEngineChannel: fix CCs not being forwarded to all

* Bumped version (2.2.0.svn6).

1 Version SVN trunk (?)
3 * general changes:
4 - AbstractEngineChannel: fixed CCs not being forwarded to all
5 VirtualMIDIDevices.
7 * Real-time instrument scripts:
8 - Fix: Do not share compiled instrument scripts among multiple engine
9 channels (sampler parts) as this can lead to crashes and would allow the
10 same global script variables to be accessible from multiple engine
11 channels which would not be the expected behaviour by instrument script
12 authors.
14 * GigaStudio/Gigasampler format engine:
15 - Use libgig's new IO-per-thread feature to avoid file I/O concurrency
16 issues with .gig file.
17 - Use libgig's new method gig::File::GetSample() with method
18 gig::File::CountSamples() instead of now deprecated
19 gig::File::GetFirstSample() / gig::File::GetNextSample() pair.
20 - Use libgig's new method gig::Instrument::GetRegionAt() with method
21 gig::Instrument:CountRegions() instead of now deprecated
22 gig::Instrument::GetFirstRegion() / gig::Instrument::GetNextRegion() pair.
23 - Use libgig's gig::File::GetInstrument() method with method
24 gig::File::CountInstruments() instead of now deprecated
25 gig::File::GetFirstInstrument() / gig::File::GetNextInstrument() pair.
27 * MIDI driver:
28 - VirtualMidiDevice: added method Reset() which is e.g. intended to be used
29 in "panic" / MIDI "all sound off" / MIDI "all note off" situations.
30 - VirtualMidiDevice: release all notes on MIDI CC #120 and CC #123.
32 Version 2.2.0 (9 May 2021)
34 * general changes:
35 - Fixed compiler error in Pool.h.
36 - Require C++14 compiler support.
37 - Autoconf: Added m4/ax_cxx_compile_stdcxx.m4 macro which is used
38 for checking in configure for C++14 support (as mandatory
39 requirement) and automatically adds compiler argument if required
40 (e.g. -std=C++14).
41 - RTMath: Implemented floating point comparison methods
42 fEqual32(float,float) and fEqual64(double,double) which take the
43 expected floating point tolerances into account.
44 - Added public API C++ class "LFO", which is a cluster class encapsulating
45 all the sampler's LFO implementations to be used by 3rd party applications
46 (e.g. by Gigedit).
47 - Added int math square LFO implementation.
48 - Added int math saw LFO implementation.
49 - Added numeric complex nr sine LFO implementation.
50 - Marked class LFOTriangleDiHarmonic as deprecated
51 (will be removed in future).
52 - Added LFOAll.h which includes all LFO implementation's header files.
53 - Refactored LFO class names and their file names (PulseLFO -> LFOPulse,
54 LFOSawIntMath -> LFOSawIntMathNew, SawLFO -> LFOSawIntMathOld,
55 SineLFO -> LFOSineBuiltinFn, LFOSine -> LFOSineNumericComplexNr,
56 SquareLFO -> LFOSquarePulse and separated the latter to its own header
57 file).
58 - Renamed type LFOSigned -> LFOTriangleSigned.
59 - Renamed type LFOUnsigned -> LFOTriangleUnsigned.
60 - Using now cubic interpolation as resampling algorithm by default; and
61 linear interpolation is deprecated now.
62 - Event:Init(): memset() entire Event::Param union to zero instead of trying
63 to do that for its individual members; the latter became error prone due
64 to the union's growing complexity meanwhile.
65 - EngineChannel: Overhaul of RPN/NRPN related methods: Added new
66 *Rpn*Parameter*() methods as replacement for previous *Rpn*Controller*()
67 methods and marked the old ones as deprecated; and likewise added new
68 *Nrpn*Parameter*() methods as replacement for previous *Nrpn*Controller*()
69 methods and marked those old ones deprecated as well.
70 - Added support for MIDI CC #38 (data entry LSB): this allows higher value
71 resolution for certain (e.g. hard coded) RPN and NRPN parameters.
72 - Implemented support for compressed RPN/NRPN message sequences (that is to
73 allow processing MIDI data entry messages without always having to be
74 preceded strictly by RPN/NRPN parameter selection messages on each value
75 change).
76 - Added support for MIDI CC #96 (data increment) and MIDI CC #97 (data
77 decrement).
78 - Added support for 'null function RPN/NRPN' (which is RPN/NRPN parameter
79 MSB=127 and LSB=127) and which is disabling subsequent data
80 entry/increment/decrement CC messages.
81 - Fixed crash when editing key groups (a.k.a. exclusive groups) with an
82 instrument editor.
83 - Fixed playing noise at the end of a disk stream under certain conditions.
84 - Fixed potential crash with high pitch at end of RAM playback.
85 - Added configure checks whether C++ compiler supports so called
86 'designated initializers'.
87 - Added global backtraceAsString() function for debugging purposes
88 (POSIX systems only, no Windows implementation yet).
89 - Thread class: Added new methods pushCancelable(), popCancelable(), name(),
90 nameOfCaller() and setNameOfCaller().
91 - configure: Always check for availability of pthread_testcancel() and use
92 it if so (previously it was only used if explicitly enabled by configure
93 option).
94 - Thread class: raise compiler warning if pthread_testcancel() is not
95 available, because it can lead to deadlocks.
96 - Mutex class: Implemented optional bug detection and deadlock debugging
97 features.
98 - Ref<> class is now thread safe (as lock-free and wait-free
99 implementation).
100 - Fixed compilation error with Bison >= 3.6.0.
101 - Fixed rare crash when switching between instruments that had instrument
102 scripts loaded.
103 - Allow EffectChain::InsertEffect() as alternative to
104 EffectChain::AppendEffect() (i.e. with index equal to current chain size).
105 - ConstCapacityArray class: Fixed bug in remove() method.
106 - Fixed crash on certain MIDI RPN and NRPN messages that occurred if no
107 real-time instrument script was loaded.
109 * Real-time instrument scripts:
110 - Added method ScriptVM::setExitResultEnabled() which allows to
111 explicitly enable the built-in exit() function to optionally accept
112 one function argument; the value of the passed exit() function
113 argument will then become available by calling
114 VMExecContext::exitResult() after script execution.
115 - 64 bit support for NKSP integer scripts variables (declare $foo).
116 - Variable names, function names and preprocessor condition names must start
117 with a regular character (a-z or A-Z); starting them with a digit or
118 underscore is not allowed.
119 - NKSP parser fix: equal comparison operator "=" and not equal comparison
120 operator "#" must only accept integer operands.
121 - NKSP language: Implemented support for standard units like Hertz, seconds,
122 Bel including support for metric unit prefixes; so one can now e.g.
123 conveniently use numbers in scripts like "5us" meaning "5 microseconds",
124 or e.g. "12kHz" meaning "12 kilo Hertz", or e.g. "-14mdB" meaning
125 "minus 14 Millidecibel", or e.g. "28c" meaning "28 cents" (for tuning).
126 - NKSP language: Introduced "final" operator "!" which is specifically
127 intended for synthesis parameter values to denote that the synthesis
128 parameter value is intended to be the "final" value for that synthesis
129 parameter that should explicitly be used by the engine and thus causing
130 the sampler engine to ignore all other modulation sources for the same
131 synthesis parameter (like e.g. LFO, EG); by simply prefixing a value,
132 variable or formula with this new "!" operator the expression is marked as
133 being "final".
134 - NKSP script editor API: Added support for detecting standard unit tokens
135 and their potential metric prefix token.
136 - NKSP language: Added support for NKSP real number literals and
137 arithmetic operations on them (e.g. "(3.9 + 2.9) / 12.3 - 42.0").
138 - NKSP language: Added support for NKSP real number (floating point) script
139 variables (declare ~foo).
140 - NKSP language: Added support for NKSP real number (floating point) array
141 script variables (declare ?foo[]).
142 - Built-in script function "message()" accepts now real number argument as
143 well.
144 - Added built-in script function "real_to_int()" and its short hand form
145 "int()" for casting from real number to integer in NKSP scripts.
146 - Added built-in script function "int_to_real()" and its short hand form
147 "real()" for casting from integer to real number in NKSP scripts.
148 - Allow built-in exit() function to potentially accept real number type
149 argument as well.
150 - Built-in script functions may have a different return type depending on
151 the arguments passed to the function.
152 - Built-in script function "abs()" optionally accepts and returns real
153 number.
154 - Built-in script functions "min()" and "max()" optionally accept real
155 number arguments and return real number as result in that case.
156 - NKSP VM API: Allow units and 'final'ness to be returned as result from
157 built-in functions (added methods VMFunction::returnUnitType() and
158 VMFunction::returnsFinal() for that purpose which must be implemented by
159 built-in function implementations).
160 - NKSP language: Allow metric unit prefixes of numeric scalar and array
161 variables to be changed freely at runtime (unlike unit types like Hz etc.
162 which are still sticky, parse-time features of variables which cannot be
163 changed at runtime for the intentional sake of determinism).
164 - NKSP language: 'final' values are prohibited for array variables for now
165 (attempt causes a parsers error).
166 - NKSP language: expressions with unit types (e.g. Hz) are prohibited for
167 conditions of runtime control structures like if(), while(), select()
168 (attempt causes a parser error).
169 - NKSP VM API: Allow built-in functions to perform their own, individual
170 parse time checks of arguments going to be passed to the function at
171 runtime (added method VMFunction::checkArgs() for that purpose).
172 - NKSP language: raise parser warning if only one operand of binary
173 operators (like logical 'or' comparison) contain a 'final' value (because
174 it would always yield in a 'final' result in such cases).
175 - NKSP language: Allow comparison (=, #, <, >, <=, >=) of values with
176 different metric unit prefixes, which will behave as expected (e.g.
177 result of expression '1000us < 2ms' is true).
178 - NKSP language: Allow adding values with different metric unit prefixes
179 (e.g. result of expression '100Hz + 5kHz' is '5100Hz').
180 - NKSP language: Allow subtracting values with different metric unit
181 prefixes (e.g. result of expression '1ms - 20us' is '980us').
182 - NKSP language: Allow multiplying with any metric unit prefixes
183 (e.g. result of expression '2k * 3ms' is '6s'), however multiplications
184 with unit types on both sides (e.g. '2s * 2s') is still prohibited since
185 we don't have any considerable practical use for a term like '4s^2'
186 (hence any attempt multiplying two unit types still causes parser error).
187 - NKSP language: Allow dividing by any metric unit prefixes and allow
188 division of same unit type on both sides (e.g. expression '8kHz / 1000Hz'
189 yields in unit free result '8'). So this is now a way to cast units away
190 e.g. for passing the result to other expressions, certain function calls
191 or variables which are not accepting any units (or that specific unit).
192 - NKSP language: integer arrays and real number arrays can now be converted
193 to strings (e.g. for dumping their content with message() calls for
194 script debugging purposes).
195 - NKSP language: expressions and variables with units are now correctly
196 casted to strings (e.g. with message() calls).
197 - NKSP language: comparing real numbers for equalness (e.g. '~foo = 3.1') or
198 unequalness (e.g. '~foo # 3.1') is now less strict and takes the expected
199 floating point tolerances into account.
200 - NKSP VM API: Added methods VMScalarNumberExpr::evalCastInt() and
201 VMScalarNumberExpr::evalCastReal().
202 - NKSP VM API: Added base class 'VMNumberArrayExpr' for classes
203 'VMIntArrayExpr' and 'VMRealArrayExpr'.
204 - NKSP VM API: replaced all unitPrefix() (parse time) methods by
205 unitFactor() (runtime) methods.
206 - Built-in function "exit()" supports now returning units and 'final'ness
207 for test cases.
208 - The following built-in functions support now units as well: "abs()",
209 "random()", "inc()", "dec()", "in_range()", "min()", "max()",
210 "real_to_int()", "int()", "int_to_real()" and "real()".
211 - Built-in functions "array_equal()", "search()" and "sort()" support now
212 real number arrays (correctly) as well.
213 - Added individual parse time checks of arguments to be passed to built-in
214 functions "random()", "inc()", "dec()", "in_range()", "min()", "max()",
215 "array_equal()" and "search()" specific for their individual purposes.
216 - NKSP VM refactoring: Renamed all methods, functions and classes matching
217 pattern *ScalarNumber* to simply *Number* (that is i.e. classes
218 VMScalarNumberExpr -> VMNumberExpr, ScalarNumberExpr -> NumberExpr,
219 ScalarNumberVariable -> NumberVariable, ScalarNumberBinaryOp ->
220 NumberBinaryOp, VMScalarNumberResultFunction -> VMNumberResultFunction,
221 method VMExpr::asScalarNumberExpr() -> VMExpr::asNumber(), function
222 isScalarNumber() -> isNumber()).
223 - NKSP VM API: Added 4 overridden methods to class VMNumberExpr:
224 evalCastInt(MetricPrefix_t), evalCastInt(MetricPrefix_t,MetricPrefix_t),
225 evalCastReal(MetricPrefix_t), evalCastReal(MetricPrefix_t,MetricPrefix_t)
226 as convenient methods for automatically converting values to expected
227 metric value basis.
228 - Built-in function "wait()" accepts now both integers and real numbers as
229 argument.
230 - NKSP VM API cleanup: Get rid of legacy method
231 VMFunction::argType(vmint iArg) which was already superseded by its new
232 replacement VMFunction::acceptsArgType(vmint iArg, ExprType_t type).
233 - NKSP parser: if wrong argument type was passed to a built-in function and
234 that built-in function accepts more than one data type for the argument,
235 then show all supported data types as parser error message.
236 - Built-in function "play_note()" accepts now real numbers and seconds as
237 unit type as well for its 3rd and 4th function arguments.
238 - The following built-in functions accept now real numbers as well for their
239 2nd function argument: "change_vol()", "change_tune()", "change_cutoff()",
240 "change_attack()", "change_decay()", "change_release()",
241 "change_sustain()", "change_cutoff_attack()", "change_cutoff_decay()",
242 "change_cutoff_sustain()", "change_cutoff_release()",
243 "change_amp_lfo_freq()", "change_cutoff_lfo_freq()",
244 "change_pitch_lfo_freq()", "change_vol_time()", "change_tune_time()",
245 "change_pan_time()", "fade_in()", "fade_out()", "change_play_pos()".
246 - Fixed built-in function "change_play_pos()" not having accepted metric
247 prefixes at all.
248 - Fixed the following built-in functions having misinterpreted values given
249 with unit type (for their 2nd argument) as if they were relative values
250 (that is as if they were passed without a unit type): "change_attack()",
251 "change_decay()", "change_release()", "change_cutoff_attack()",
252 "change_cutoff_decay()", "change_cutoff_release()".
253 - Fixed the following built-in functions having applied completely wrong
254 'final' values: "change_sustain()", "change_cutoff_sustain()" (since the
255 respective EGs being their modulation sink assume uint data type with
256 value range 0..1000 instead of 0.0..1.0.
257 - Added individual parse-time checks of function arguments for the following
258 built-in functions: "play_note()", "note_off()", "set_event_mark()",
259 "delete_event_mark()", "by_marks()", "change_cutoff()", "change_attack()",
260 "change_decay()", "change_release()", "change_cutoff_attack()",
261 "change_cutoff_decay()", "change_cutoff_release()",
262 "change_amp_lfo_freq()", "change_cutoff_lfo_freq()",
263 "change_pitch_lfo_freq()", "change_vol_time()", "change_tune_time()" and
264 "change_pan_time()".
265 - Don't abort function call if unit type was used and at the same time
266 'final' operator was omitted for the primary value argument of the
267 following built-in functions: "change_cutoff()", "change_attack()",
268 "change_decay()", "change_release()", "change_cutoff_attack()",
269 "change_cutoff_decay()", "change_cutoff_release()",
270 "change_amp_lfo_freq()", "change_cutoff_lfo_freq()",
271 "change_pitch_lfo_freq()", "change_vol_time()", "change_tune_time()",
272 "change_pan_time()", instead imply 'final'ness at runtime and raise an
273 appropriate parser warning at parse time.
274 - Added built-in real number functions "round()", "ceil()", "floor()",
275 "sqrt()", "log()", "log2()", "log10()", "exp()", "pow()", "sin()",
276 "cos()", "tan()", "asin()", "acos()", "atan()".
277 - Added built-in script real number constant "~NI_MATH_PI".
278 - Added built-in script real number constant "~NI_MATH_E".
279 - NKSP language: Allow unary '+' operator.
280 - Added built-in script functions "msb()" and "lsb()".
281 - NKSP language: Added support for RPN event handler
282 ("on rpn ... end on" in instrument scripts).
283 - NKSP language: Added support for NRPN event handler
284 ("on nrpn ... end on" in instrument scripts).
285 - Added built-in read-only variables "$RPN_ADDRESS" and "$RPN_VALUE" which
286 may be read from the new RPN/NRPN script handlers to get the (N)RPN
287 parameter that had been changed and its new value.
288 - Added built-in const variables "$NI_CB_TYPE_RPN" and "$NI_CB_TYPE_NRPN"
289 which are identifying the new (N)RPN handlers as such at script runtime.
290 - Implemented built-in script function "set_rpn()".
291 - Implemented built-in script function "set_nrpn()".
292 - Fix: built-in functions returning an array type were broken and caused a
293 crash when the built-in function's (supposed to be array) result value was
294 used by the script VM.
295 - Removed upper value constraint for 2nd argument of built-in functions
296 "change_amp_lfo_freq()", "change_cutoff_lfo_freq()",
297 "change_pitch_lfo_freq()","change_reso()", "change_cutoff()",
298 "change_amp_lfo_depth()", "change_cutoff_lfo_depth()" and
299 "change_pitch_lfo_depth()", to allow e.g. passing value 2000000 to double
300 the LFO frequency / resonance / cutoff frequency.
301 - Reduced code involved for parsing NKSP variable declarations by using 2
302 generalized grammar rules with optional components (i.e. optional value
303 assignment and optional variable qualifiers) instead of previous linear
304 combinations of grammar rules, which also fixes some edge case.
305 - NKSP language parser: track code block locations also by a) raw byte
306 position in NKSP source code string and b) length of that code block in
307 bytes (along to the already existing code block tracking by line/column);
308 and added appropriate new members to public C++ API classes like
309 'firstByte' and 'lengthBytes' to class VMSourceToken and struct CodeBlock.
310 - NKSP language: Added support for 'patch' variable qualifier (capture their
311 locations and allow replacement of their assignment expression via VM C++
312 API).
313 - Fixed incorrect EG times if standard measuring unit was passed to built-in
314 functions "change_attack()", "change_decay()" or "change_release()".
315 - Fixed intermediate function result values never having reflected any
316 standard measuring unit type.
317 - Fixed re-entrant issue with function calls which caused wrong result
318 values if the same function was called multiple times in a term
319 (specifically if metric prefixes were used).
320 - NKSP language: Just throw a warning, not an error if an array variable of
321 size zero was declared.
322 - NKSP language: Allow omitting explicit array size on array variable
323 declaration if combined with immediate value assignment
324 (e.g. declare %foo[] := ( 1, 2, 3 ) ).
325 - NKSP language: emit warning if an array variable was declared with bigger
326 array size than amount of initial values been assigned, and initialize the
327 missing array elements with zero in this case.
328 - Fixed built-in NKSP exit() function to behave as return statement.
329 - NKSP parser: Fixed crash if unary '-' operator was used on a non-number
330 data type.
331 - NKSP VM: Fixed the "release" handler being executed too often under
332 certain situations if polyphonic data was passed from "note" handler to
333 "release" handler.
334 - Fixed memory leak in NKSP parser: string tokens were allocated as C
335 strings and never freed.
336 - Fixed incorrect polyphonic data transfer from wrong note handler to
337 release handler; which also fixes no release handler being executed
338 sometimes, and due to the latter it also fixes potential crashes as some
339 polyphonic script events were never released and the engine hence was
340 running out of free script events.
341 - Fixed crash when handling array results of built-in functions
342 (e.g. "num_elements(by_marks($MARK_1)").
343 - NKSP VM: Fixed dead note events not been removed from user's event groups
344 (i.e. causing incorrect behaviour of built-in NKSP functions by_marks()
345 and set_event_mark() after max. size of event group list was exhausted).
347 * test cases:
348 - Fixed compiler errors in test cases.
349 - Updated README for how to compile & run test cases.
350 - Updated test case
351 MutexTest::testDoubleLockStillBlocksConcurrentThread() to latest
352 expected behaviour of the Mutex class implementation (recursive
353 mutex type).
354 - Added test cases for NKSP core language aspects and core built-in
355 functions.
356 - Fixed thread tests segfaulting on Linux.
357 - NKSP: Added real number test cases for built-in functions exit(),
358 int_to_real(), real(), real_to_int() and int(), as well as for the
359 plus, minus and negate language operators.
360 - Added massive amount of NKSP test cases for standard measuring units and
361 'final' operator usage cases.
362 - Added NKSP test cases for (floating point tolerance aware) real number
363 equalness / unequalness comparison.
364 - Added NKSP int array and real array tests for value assignment and
365 initialization of arrays.
366 - Added NKSP test cases for built-in functions "round()", "ceil()",
367 "floor()", "sqrt()", "log()", "log2()", "log10()", "exp()", "pow()",
368 "sin()", "cos()", "tan()", "asin()", "acos()", "atan()".
369 - Added NKSP test cases for unary '+' operator.
370 - Added NKSP test cases for core built-in variables.
371 - Added thorough NKSP test cases for variable declarations.
373 * GigaStudio/Gigasampler format engine:
374 - LFOTriangleIntMath and LFOTriangleIntAbsMath: Fixed FlipPhase=true
375 behaviour for start_level_mid.
376 - Changed LFO start levels: LFO1 and LFO2 both to mid, LFO3 to max. start
377 level (see discussion "GigaStudio LFO compatibility" on mailing list
378 from 2019-09-26 for details).
379 - Changed default wave form for all 3 LFOs to sine (instead of triangle;
380 see discussion "GigaStudio LFO compatibility" on mailing list from
381 2019-09-26 for details).
382 - Format extension: Added support for different LFO wave forms (currently
383 either sine [default], triangle, saw or square).
384 - Format extension: Added support for LFO phase displacement (0��..360��).
385 - Format extension: Added support for flipping LFO polarity on LFO 3
386 (in the original gig format this was only available for LFO 1 and LFO 2).
387 - Format extension: Support for additional filter types:
388 lowpass 1/2/4/6-pole, highpass 1/2/4/6-pole, bandpass 2-pole,
389 bandreject 2-pole.
390 - Fixd potential crash with polynomial interpolator and max. pitch.
391 - If minimum cutoff is defined, remap cutoff controller CC values to really
392 span the entire possible cutoff range (and avoiding a dead controller
393 zone).
394 - Implemented support for loading instrument scripts with 'patch' variables
395 (by patching these script variables with potentially overridden ones by
396 gig Instrument).
398 * SFZ format engine:
399 - Fixed support for regions with loccN/hiccN conditions on more than one
400 MIDI controller.
401 - Fixed internal opcode 'xfin_hivel' falsely being set by user opcode
402 'xfin_lovel' (patch by Christian Czezatke).
404 * Benchmarks:
405 - Fixed benchmarks/triang.cpp falsely having favoured "int math abs"
406 algorithm (since result of 2nd run was not accumulated).
407 - Added benchmark for saw wave (benchmarks/saw.cpp).
408 - Added benchmark for sine wave (benchmarks/sine.cpp).
409 - Added benchmark for square wave (benchmarks/square.cpp).
410 - Increased amount of benchmarks runs by factor 6 to achieve benchmark times
411 which are large enough on modern systems.
413 Version 2.1.1 (27 Jul 2019)
415 * Real-time instrument scripts:
416 - Fixed behavior of built-in NKSP functions change_sustain(),
417 change_cutoff_attack(), change_cutoff_decay(), change_cutoff_sustain()
418 and change_cutoff_release().
420 * general changes:
421 - Only play release trigger samples on sustain pedal up if this behaviour
422 was explicitly requested by the instrument (otherwise only on note-off).
423 - Fixed compiler warnings.
424 - Fixed compilation error when cross-compiling to Mac.
425 - FX Sends: Provide more useful error messages on routing problems
426 (see bug #169).
427 - LSCP doc: Be more clear describing the two distinct approaches
428 of using external vs. internal effects (see bug #169).
429 - "optional" class: Fixed comparison operators.
431 * Gigasampler/GigaStudio format engine:
432 - Format extension: If requested by instrument then don't play release
433 trigger sample on note-off events.
435 * SFZ format engine:
436 - Fixed memory leak when releasing samples
437 (fixes bug #307, patch by Jacek Roszkowski)
438 - Fixed potential crash when a sample is shared by more than one region
439 (fixes bug #308, patch by Jacek Roszkowski).
440 - Opcode 'sample': Added support for built-in sample '*silence'
441 (fixes bug #310, patch by Jacek Roszkowski).
443 Version 2.1.0 (25 Nov 2017)
445 * SFZ format engine:
446 - added support for <global>, <master> and #define (patch by Alby M)
447 - Removed code duplication in SFZ file loading code.
448 - Added support for sfz extension opcode 'script' which may be used to
449 load real-time instrument script file (NKSP script language).
450 - Implemented opcode set_ccN (initial patch by Giovanni Senatore).
451 - Fixed unintended volume fade-in of voices under certain conditions.
452 - sfz parser: allow missing space between header and opcode
454 * Gigasampler/GigaStudio format engine:
455 - Fixed clicks and pumping noise with Lowpass Turbo filter on very low
456 cutoff settings.
457 - Got rid of resembling an ancient GSt misbehavior which did not pitch at
458 all if an up-pitch of more than 40 semi tones was requested (I don't
459 think there is any stock gig sound that requires this behavior to
460 resemble its original sound).
461 - Added support for controlling whether the individual EGADSR stages may
462 be aborted (as LinuxSampler extension to the original GigaStudio 4
463 format).
465 * general changes:
466 - fixed printf type errors (mostly in debug messages)
467 - use unique_ptr instead of auto_ptr when building with C++11
468 - Added RTAVLTree class which is a real-time safe ordered multi-map, thus
469 allowing to sort data efficiently in real-time safe manner.
470 - RTList class: added methods for moving/inserting elements to arbitrary
471 position within a list.
472 - RTList class: added method fromPtr() for creating an Iterator object from
473 a raw element pointer.
474 - Attempt to partly fix resetting engine channels vs. resetting engine, an
475 overall cleanup of the Reset*(), ConnectAudioDevice(),
476 DisconnectAudioDevice() API methods would still be desirable though,
477 because the current situation is still inconsistent and error prone.
478 - Active voices are now internally grouped to "Note" objects, instead of
479 being directly assigned to a keyboard key. This allows more fine graded
480 processing of voices, which is i.e. required for certain instrument
481 script features.
482 - Fix: Release trigger voices were not spawned on sustain pedal up
483 (CC #64) events.
484 - Fix: Release trigger voices can now also distinguish correctly between
485 note-off and sustain pedal up events.
486 - All engines: Increased ramp speed of volume smoother and pan smoother
487 (while slow rate of crossfade smoother is preserved) to allow quick
488 volume and pan changes by instrument scripts for instance.
489 - gig/sf2/sfz: Fixed aftertouch channel pressure events being processed
490 properly and thus were ignored for certain uses.
491 - Added new C++ API method
492 VirtualMidiDevice::SendChannelPressureToSampler().
493 - windows, 32-bit: fixed potential crashes by making sure the stack in
494 sub threads is 16-byte aligned
495 - fixed numerous compiler warnings
496 - Fixed invalid (note-on) event ID being assigned to new Note objects.
497 - Revised fundamental C++ classes "Thread", "Mutex" and "Condition" which
498 fixes potential undefined behavior.
499 - Fixed Note object leak when triggering notes on keys which did not
500 have a valid sample mapped (fixes bug #252).
501 - Fixed compilation errors when compiling with CONFIG_DEVMODE enabled.
502 - linuxsampler binary fix: option --create-instruments-db ignored
503 subsequent optional argument due to glibc's implementation oddity
504 which expects a "=" sign, but no space between them.
506 * packaging changes:
507 - removed unnecessary dependency to libuuid
508 (originated by libgig's usage of it)
509 - Automake: set environment variable GCC_COLORS=auto to allow GCC to
510 auto detect whether it (sh/c)ould output its messages in color.
511 - Debian: Fixed packaging error about invalid "Source-Version"
512 substitution variable.
513 - Debian: Raised Debian compatibility level to Debian 9 "Stretch".
514 - Debian: Added build dependency to libsqlite3-dev for building
515 linuxsampler with instruments DB support.
517 * Real-time instrument scripts:
518 - Implemented scheduler for delayed MIDI events and for suspended scripts.
519 - Built-in script function "wait()": implemented support for function's
520 "duration-us" argument, thus scripts using this function are now
521 correctly resumed after the requested amount of microseconds.
522 - Built-in script function "play_note()": implemented support for
523 function's "duration-us" argument, thus notes triggered with this
524 argument are now correctly released after the requested amount of
525 microseconds.
526 - Fix: script events were not cleared when engine channel was reset,
527 potentially causing undefined behavior.
528 - Fixed crash which happened when trying to reference an undeclared
529 variable.
530 - Built-in script function "play_note()": Added support for passing
531 special value -1 for "duration-us" argument, which will cause the
532 triggered note to be released once the original note was released.
533 - Instrument script classes now exported with the liblinuxsampler C++ API.
534 - Added new API method ScriptVM::syntaxHighlighting() which provides
535 a convenient syntax highlighting backend for external instrument
536 script editor applications.
537 - Added new C++ API class "ScriptVMFactory".
538 - Extended parser issues to provide not only first line and first
539 column, but also last line and last column of issue (thus marking
540 the precise span of the issue within the source code).
541 - Reload script automatically after being modified by an instrument
542 editor.
543 - NKSP language grammar correction: allow empty event handler bodies
544 like "on note end on".
545 - Implemented built-in script function "change_vol()".
546 - Implemented built-in script function "change_tune()".
547 - Implemented built-in script function "change_pan()".
548 - Implemented built-in script function "change_cutoff()".
549 - Implemented built-in script function "change_reso()".
550 - Implemented built-in script function "event_status()".
551 - Added built-in script constants "$EVENT_STATUS_INACTIVE" and
552 "$EVENT_STATUS_NOTE_QUEUE" both for being used as flags for
553 "event_status()" function.
554 - NKSP language: Added support for bitwise operators ".or.", ".and."
555 and ".not.".
556 - NKSP language scanner: Fixed IDs matching to require at least one
557 character (i.e. when matching function names or variable names).
558 - NKSP language scanner: disabled unusued rules.
559 - Fixed behavior of play_note() and note_off() functions which must
560 be distinguished engine internally from "real" MIDI note on/off
561 events in order to avoid misbehaviors like hanging notes.
562 - Implemented built-in script variable "$KSP_TIMER".
563 - Implemented built-in script variable "$NKSP_REAL_TIMER".
564 - Implemented built-in script variable "$NKSP_PERF_TIMER".
565 - Implemented built-in script variable "$ENGINE_UPTIME".
566 - Implemented built-in script function "inc()".
567 - Implemented built-in script function "dec()".
568 - NKSP language fix: division expressions were evaluated too often.
569 - NKSP language fix: string concatenation operator was right
570 associative instead of left (to right).
571 - NKSP language correction: allow empty statements in entire language
572 (i.e. "if end if").
573 - Implemented built-in script function "stop_wait()".
574 - Implemented built-in script variable "$NI_CALLBACK_ID".
575 - Implemented built-in script variable "$NI_CALLBACK_TYPE".
576 - Implemented built-in script variable "$NKSP_IGNORE_WAIT".
577 - Added support for read-only built-in variables (respectively
578 handled by the script parser).
579 - Added built-in script constant "$NI_CB_TYPE_INIT".
580 - Added built-in script constant "$NI_CB_TYPE_NOTE".
581 - Added built-in script constant "$NI_CB_TYPE_RELEASE".
582 - Added built-in script constant "$NI_CB_TYPE_CONTROLLER".
583 - NKSP Language: Added support for user defined script functions.
584 - Implemented built-in script function "change_attack()".
585 - Implemented built-in script function "change_decay()".
586 - Implemented built-in script function "change_release()".
587 - Fixed all change_*() built-in script functions to apply their
588 synthesis parameter changes immediately in case the respective note
589 was triggered at the same time, instead of scheduling the parameter
590 change, especially because it would cause some parameter types's
591 changes either to be ramped (i.e. change_vol()) or other types even
592 to have not effect at all (i.e. change_attack()).
593 - Implemented built-in script function "sh_left()".
594 - Implemented built-in script function "sh_right()".
595 - Implemented built-in script function "min()".
596 - Implemented built-in script function "max()".
597 - NKSP Fix: "init" event handler was not always executed when a script
598 was loaded.
599 - NKSP built-in wait() script function: abort script execution if
600 a negative or zero wait time was passed as argument (since this is
601 a common indication of a bug either of the script or even of the
602 engine, which could lead to RT instability or even worse).
603 - ScriptVM: Implemented automatic suspension of RT safety
604 threatening scripts.
605 - Provide more user friendly error messages on syntax errors.
606 - Fixed NKSP parser warning "Not a statement" when assigning an
607 initializer list to an array variable.
608 - Implemented built-in script array variable "%ALL_EVENTS".
609 - Implemented built-in script function "in_range()".
610 - Implemented built-in script function "change_amp_lfo_depth()".
611 - Implemented built-in script function "change_amp_lfo_freq()".
612 - Implemented built-in script function "change_pitch_lfo_depth()".
613 - Implemented built-in script function "change_pitch_lfo_freq()".
614 - Implemented built-in script function "change_vol_time()".
615 - Implemented built-in script function "change_tune_time()".
616 - Implemented built-in script function "fade_in()".
617 - Implemented built-in script function "fade_out()".
618 - Fixed acceptance of wrong data type of parameters passed to built-in
619 script functions "change_vol()", "change_tune()", "change_pan()",
620 "change_cutoff()", "change_reso()", "change_attack()", "change_decay()",
621 "change_release()", "change_amp_lfo_depth()", "change_amp_lfo_freq()",
622 "change_pitch_lfo_depth()" and "change_pitch_lfo_freq()".
623 - Added built-in script function "get_event_par()" and implemented some
624 of its possible parameter selections.
625 - Added built-in script function "set_event_par()" and implemented some
626 of its possible parameter selections.
627 - Fixed a bunch of scheduler time related bugs.
628 - Fixed polyphonic variables not being reset to zero after usage.
629 - Built-in "ignore_event()" function: argument is now optional, like with
630 built-in function "ignore_controller()".
631 - Implemented built-in script function "change_velo()".
632 - Implemented built-in script function "change_note()".
633 - Adjusted behavior of "change_vol()" and "change_tune()" to a more
634 intuitive behavior if used in combination with "change_vol_time()" or
635 "change_tune_time()" respectively: now tuning/volume changes are only
636 assigned (without delay) immediately to a new note if the respective
637 timing function has not been called before, otherwise the volume/tuning
638 changes are automatically faded (before, only the event's time stamp was
639 relevant).
640 - Implemented built-in script function "array_equal()".
641 - Implemented built-in script function "search()".
642 - Implemented built-in script function "sort()".
643 - NKSP Fix: Never suspend "init" event handlers.
644 - Implemented built-in script function "same_region()" (currently only
645 available for gig format engine).
646 - Added built-in script constant "$NKSP_LINEAR".
647 - Added built-in script constant "$NKSP_EASE_IN_EASE_OUT".
648 - Implemented built-in script function "change_vol_curve()".
649 - Implemented built-in script function "change_tune_curve()".
650 - built-in "play_note()" function now supports a sample playback start
651 offset with argument 3, where special value -1 means to use the regular
652 sample offset as defined by the instrument file.
653 - Built-in array variable %KEY_DOWN[] is now a read-only variable.
654 - Built-in variable $EVENT_NOTE is now a read-only variable.
655 - Built-in variable $EVENT_VELOCITY is now a read-only variable.
656 - built-in "play_note()" function now accepts -2 for its fourth argument
657 (note duration) which means the life time of the note shall be sticked
658 to the requested note number of argument 1.
659 - Fix: built-in "play_note()" function now returns 0 as result value if
660 -1 was passed for its fourth argument (note duration) and the respective
661 parent note is already gone.
662 - Implemented built-in script function "change_play_pos()".
663 - NKSP language: Added support for user declared const array variables.
664 - NKSP language: Raise parser warning if array variable is accessed with
665 an index that exceeds the array's size.
666 - NKSP language fix: Unknown characters were not handled correctly.
667 - NKSP language: Added support for "synchronized .. end synchronized"
668 code blocks.
669 - Implemented built-in script function "abort()" which allows to abort
670 another script handler by passing its callback ID.
671 - Fixed potential memory access bug and potential undefined behavior of
672 "init" event handlers.
673 - Print a time stamp along to each call of built-in function "message()".
674 - ScriptVM API: Added VMParserContext::preprocessorComments() which allows
675 to retrieve all code blocks filtered out by the preprocessor.
676 - Added built-in script function "fork()".
677 - Added built-in array variable %NKSP_CALLBACK_CHILD_ID[].
678 - Added built-in variable $NKSP_CALLBACK_PARENT_ID.
679 - Fixed potential crash when accessing dynamic built-in array variables.
680 - Added built-in script function "callback_status()".
681 - Added built-in constant $CALLBACK_STATUS_TERMINATED.
682 - Added built-in constant $CALLBACK_STATUS_QUEUE.
683 - Added built-in constant $CALLBACK_STATUS_RUNNING.
684 - Removed max. value limitation of built-in functions "change_attack()",
685 "change_decay()" and "change_release()" to i.e. allow passing 2000000
686 for doubling the respective time.
687 - NKSP script editor syntax highlighting API: Fixed app termination due
688 to a lexer start condition stack underrun.
689 - NKSP preprocessor: Fixed wrong behavior on nested USE_CODE_IF() and
690 USE_CODE_IF_NOT() preprocessor statements.
691 - NKSP: Added built-in preprocessor condition NKSP_NO_MESSAGE, which
692 can be set to disable all subsequent built-in "message()" function calls
693 on preprocessor level.
694 - Implemented built-in script function "change_sustain()".
695 - NKSP script editor syntax highlighting API: catch all fatal lexer errors,
696 to avoid the editor app to crash on ill-formed text input.
697 - Added built-in script function "change_pan_time()".
698 - Added built-in script function "change_pan_curve()".
699 - Added built-in script function "change_cutoff_attack()".
700 - Added built-in script function "change_cutoff_decay()".
701 - Added built-in script function "change_cutoff_sustain()".
702 - Added built-in script function "change_cutoff_release()".
703 - Added built-in script function "change_cutoff_lfo_depth()".
704 - Added built-in script function "change_cutoff_lfo_freq()".
706 * Instruments DB:
707 - Fixed memory access bug of general DB access code which lead to
708 undefined behavior.
709 - Cleanup of instruments DB file creation and opening code.
710 - The instrument DB path of linuxsampler's --create-instruments-db argument
711 is now optional, if it is missing, then a default location is used.
712 - Added support for scanning SFZ (.sfz) files.
713 - Added support for scanning Sound Font (.sf2) files.
714 - Fixed undefined DB transaction behavior.
716 Version 2.0.0 (15 July 2015)
718 * packaging changes:
719 - fixed building with newer MinGW-w64
720 - Mac OS X: support the new dir for Core Audio SDK
721 - Mac OS X: fixed building outside source directory
722 - made sure all source files for hostplugins are included when
723 doing "make dist"
724 - removed empty directories left from the cvs to svn migration
725 - fixed compilation with gcc 4.6.1
726 - another "make dist" fix, for the LV2 plugin
727 - made --enable-pthread-testcancel default on Mac OS X
728 - minor Makefile fix for building lscpparser when build and source
729 directories are separate
730 - Mac OS X: made it possible to specify plugin installation dir to
731 configure
732 - Mac OS X: Makefile fix for the install-strip target
733 - fixed compilation with gcc 4.7
734 - fixed configure script error with old autoconf versions
735 - lsatomic.h: use gcc provided atomic functions if building with
736 gcc 4.7 and C++11
737 - modernized configure script
738 - fixed linkage error when building with
739 LDFLAGS="-Wl,--no-undefined" (#190)
740 - fixed compilation with Clang 3.2
741 - removed usage of deprecated Automake variable INCLUDES
742 - fixed building with C++11
743 - build fix: ChangeFlagRelaxed.h was missing in makefile
744 - build fix: libsndfile compiler flags were missing in some
745 makefiles
746 - fix for building with bison 3.0 (#202)
747 - Mac OS X: added temporary hack allowing to spawn gigedit as callback
748 on the process's main thread
749 - fixed build error on newer MinGW
750 - support building with older jack versions
751 - support building with spaces in vst sdk path
752 - enabled automake 'subdir-objects' option and moved external
753 source references (vst, au, asio) from makefiles to cpp files,
754 in order to get rid of warnings from automake 1.14
756 * general changes:
757 - Refactoring: moved the independent code from
758 the Gigasampler format engine to base classes
759 - command line option '--profile' is currently disabled, since the
760 respective profiling code is currently broken
761 - Introduced support for optional environment variable
762 "LINUXSAMPLER_PLUGIN_DIR" which allows to override the directory
763 where the sampler shall look for instrument editor plugins
764 (patch by Luis Garrido, slightly modified).
765 - implemented Roland GS NRPN 1ArrH which allows to set volume per note
766 - implemented Roland GS NRPN 1CrrH which allows to set pan per note
767 - implemented Roland GS NRPN 1DrrH which allows to set reverb send per
768 note (in this implementation of the sampler its simply hard routed to
769 the 1st effect send of the sampler channel, no matter what the actual
770 effect type is)
771 - implemented Roland GS NRPN 1ErrH which allows to set chorus send per
772 note (in this implementation of the sampler its simply hard routed to
773 the 2nd effect send of the sampler channel, no matter what the actual
774 effect type is)
775 - implemented support for internal LADSPA effects
776 - raised limit of program change queue from 100 to 512
777 (as suggested by Alex Stone)
778 - implemented sine LFO, pulse LFO and saw LFO
779 - experimental support for per voice equalization
780 - added command line option --exec-after-init
781 - Introduced new C++ API method:
782 EngineChannel::InstrumentFileName(int index)
783 allowing to retrieve the whole list of files used for the loaded
784 instrument on an engine channel (a.k.a. part). Some GigaStudio
785 instruments for example are splitted over several files like
786 "Foo.gig", "Foo.gx01", "Foo.gx02", ...
787 - Added new C++ API method Sampler::GetGlobalMaxVoices().
788 - Added new C++ API method Sampler::GetGlobalMaxStreams().
789 - Added new C++ API method Sampler::SetGlobalMaxVoices().
790 - Added new C++ API method Sampler::SetGlobalMaxStreams().
791 - Various "const" and "restrict" optimizations.
792 - all engines: add pan CC value to instrument pan parameter before
793 applying panning, instead of using two separate pan functions in
794 series (#182)
795 - added a lock guard class for exception safe mutex handling and
796 used it everywhere appropriate
797 - Immediately apply scale tuning changes to active voices.
798 - Exposed scale tuning to C++ API (along to the already existing standard
799 SysEx way).
800 - lsatomic.h fixes: seq_cst load and store were suboptimal for x86
801 and broken for ppc64. (Seq_cst loads and stores are actually not
802 used in LS, so the bug wasn't noticable.)
803 - lsatomic.h: added ARMv7 support
804 - Added support for multiple MIDI input ports per sampler channel (and
805 added various new C++ methods for this new feature / design change, old
806 C++ API methods for managing SamplerChannel's MIDI inputs are now marked
807 as deprecated but are still there and should provide full behavior
808 backward compatibility).
809 - AbstractEngine::GSChecksum(): don't allocate memory on the stack (was
810 unsafe and caused compilation error with clang 2.x).
811 - Bugfix: only process the latest MIDI program change event.
812 - Introducing the LSCP shell, which provides convenient control of the
813 sampler from the command line by providing LSCP aware features.
814 - VirtualMidiDevice: Added support for program change.
815 - VirtualMidiDevice: Added support for bank select (MSB & LSB).
816 - VirtualMidiDevice: Added support for pitch bend.
817 - Aftertouch: extended API to explicitly handle channel pressure and
818 polyphonic key pressure events (so far polyphonic pressure was not
819 supported at all, and channel pressure was rerouted as CC128 but not
820 used so far).
821 - Added initial support for real-time instrument scripts. The script VM
822 code is shared by all sampler engine implemementations, however only the
823 gig file format currently provides support for storing instrument scripts
824 (as LinuxSampler extension to the original GigaStudio 4 format).
826 * Gigasampler format engine:
827 - implemented the "round robin keyboard" dimension
828 - fixed round robin and random dimensions for cases when number of
829 dimension zones is not a power of two
830 - made round robin use a counter for each region instead of each
831 key
832 - bugfix: pitch LFO controller "internal+aftertouch" was broken
833 - bugfix: filter keyboard tracking was broken
834 - filter performance fix (an unnecessary copy was made of the
835 filter parameters in each sub fragment)
836 - handle special case when pan parameter in gig file has max or
837 min value
838 - Exclusive Groups: don't ever stop voices of the same note,
839 doesn't sound naturally with a drumkit
840 - fixed EG1 modulation when attack or release is zero
841 - Fixed support for 'aftertouch' attenuation controller.
842 - Fixed crash that happened with velocity split sounds under certain
843 conditions (see also previous commit on libgig).
844 - fixed behaviour of filter LFO
846 * SFZ format engine:
847 - Initial implementation (not usable yet)
848 - added support for v2 multiple stage envelope generators
849 - added a fine-tuned v1 envelope generator instead of using the
850 one from the gig engine
851 - fixed memory leak and memory handling errors
852 - added support for trigger=first, trigger=legato and sw_previous
853 - allow non-numerical key values ("C#4" for example)
854 - "key" opcode now sets pitch_keycenter too
855 - fixed error when unloading instrument with same sample used by
856 multiple regions
857 - added some opcode aliases, like loopmode for loop_mode, to be
858 more compatible
859 - added support for trigger=release and rt_decay
860 - added support for off_mode=normal
861 - added support for random, seq_position, seq_length and volume
862 - added v1 LFO opcodes to parser (no support in engine yet)
863 - added support for amp_veltrack and amp_velcurve_N
864 - fine-tuned the default velocity curve
865 - added support for transpose
866 - fixed crash when using sw_down/up
867 - improved logic for sw_lokey/hikey/up/down/last
868 - added more v1 aliases to parser
869 - reduced memory usage for sfz data
870 - RT-safeness: avoid malloc in audio thread
871 - fixed a bug that could cause voice stealing to fail
872 - optimized sample lookup
873 - improved support for exclusive groups (group, off_by and
874 off_mode)
875 - added support for controller triggered regions (on_locc/on_hicc)
876 - added support for loop_mode=one_shot
877 - fixed playback of 16 bit wav files on big endian CPUs
878 - added support for Ogg Vorbis sample files
879 - use loop markers from sample file if loop_start and loop_end are
880 not set in sfz file
881 - implemeted filters. Filter types: lowpass, bandpass, bandreject
882 and highpass. 1, 2, 4 and 6 pole filters. Opcodes: fil_type,
883 cutoff, resonance, fil_veltrack, fil_keytrack, fil_keycenter,
884 cutoff_cc, cutoff_chanaft.
885 - bugfix: zero ampeg_sustain didn't work
886 - added support for velocity effect on amplifier envelope time
887 (ampeg_vel2attack, ampeg_vel2decay, ampeg_vel2sustain and
888 ampeg_vel2release)
889 - added support for EG hold (ampeg_hold)
890 - added support for sample offset (offset)
891 - use the newly introduced signal units model
892 - implemented opcodes lfoN_delay,
893 lfoN_freq, lfoN_pan, lfoN_cutoff, lfoN_resonance
894 - implemented opcodes ampeg_delay, ampeg_vel2delay
895 pitcheg_delay, pitcheg_start, pitcheg_attack, pitcheg_hold,
896 pitcheg_decay, pitcheg_sustain, pitcheg_release, pitcheg_vel2delay,
897 pitcheg_vel2attack, pitcheg_vel2hold, pitcheg_vel2decay,
898 pitcheg_vel2sustain, pitcheg_vel2release, pitcheg_depth
899 - implemented opcodes pitchlfo_delay, pitchlfo_freq,
900 pitchlfo_depth, fillfo_delay, fillfo_freq, fillfo_depth,
901 amplfo_delay, amplfo_freq, amplfo_depth
902 - implemented opcodes fileg_delay, fileg_start, fileg_attack,
903 fileg_hold, fileg_decay, fileg_sustain, fileg_release, fileg_vel2delay,
904 fileg_vel2attack, fileg_vel2hold, fileg_vel2decay, fileg_vel2sustain,
905 fileg_vel2release
906 - implemented opcode lfoN_wave
907 - implemented opcode pitchlfo_depthccN
908 - implemented opcodes lfoN_phase, lfoN_phase_onccX,
909 lfoN_pitch, lfoN_pitch_onccX
910 - implemented opcodes pitchlfo_fade,
911 fillfo_fade, amplfo_fade, lfoN_fade, lfoN_fade_onccX
912 - implemented opcodes fillfo_freqccN,
913 pitchlfo_freqccN, amplfo_freqccN, lfoN_freq_onccX
914 - implemented opcodes ampeg_delayccN, ampeg_startccN,
915 ampeg_attackccN, ampeg_holdccN, ampeg_decayccN, ampeg_sustainccN,
916 ampeg_releaseccN, egN_timeX_onccY, egN_levelX_onccY
917 - lfoN_* and egN_* opcodes defined in group sections
918 are now taken into account
919 - implemented curves
920 - implemented opcodes volume_onccN, volume_curveccN
921 - implemented opcode volume_smoothccN
922 - implemented opcodes fillfo_depthccN, amplfo_depthccN,
923 lfoN_volume, lfoN_volume_onccX, lfoN_volume_smoothccX,
924 lfoN_freq_smoothccX, lfoN_pitch_smoothccX, lfoN_pan_onccX,
925 lfoN_pan_smoothccX, lfoN_cutoff_onccX, lfoN_cutoff_smoothccX,
926 lfoN_resonance_onccX, lfoN_resonance_smoothccX, lfoN_delay_onccX
927 - implemented opcode end
928 - implemented opcodes egN_amplitude_onccX,
929 egN_volume, egN_volume_onccX, egN_cutoff, egN_cutoff_onccX,
930 egN_pitch, egN_pitch_onccX, egN_resonance, egN_resonance_onccX
931 - implemented opcodes xfin_lokey, xfin_hikey,
932 xfout_lokey, xfout_hikey, xf_keycurve, xfin_lovel, xfin_hivel,
933 xfout_lovel, xfout_hivel, xf_velcurve, xfin_loccN, xfin_hiccN,
934 xfout_loccN, xfout_hiccN, xf_cccurve
935 - implemented opcodes pan_onccN, pan_smoothccN,
936 pan_curveccN, egN_pan, egN_pan_curve, egN_pan_onccX,
937 egN_pan_curveccX, pitch_veltrack
938 - when failed to parse a sfz file
939 print the line number on which the error occurs
940 - use common pool of CC objects to minimize RAM usage
941 - implemented opcodes amplfo_delay_onccN,
942 amplfo_fade_onccN, fillfo_delay_onccN, fillfo_fade_onccN,
943 pitchlfo_delay_onccN, pitchlfo_fade_onccN
944 - implemented opcodes fileg_delay_onccN,
945 fileg_start_onccN, fileg_attack_onccN, fileg_hold_onccN,
946 fileg_decay_onccN, fileg_sustain_onccN, fileg_release_onccN,
947 fileg_depth_onccN, pitcheg_delay_onccN, pitcheg_start_onccN,
948 pitcheg_attack_onccN, pitcheg_hold_onccN, pitcheg_decay_onccN,
949 pitcheg_sustain_onccN, pitcheg_release_onccN, pitcheg_depth_onccN
950 - implemented automatic aliasing *ccN <-> *_onccN
951 - *lfo_freqccN wasn't working when the respective *lfo_freq
952 was not set or was set to zero
953 - lfoN_freq_onccX wasn't working when lfoN_freq
954 was not set or was set to zero
955 - implemented opcodes resonance_onccN, resonance_smoothccN,
956 resonance_curveccN, cutoff_smoothccN, cutoff_curveccN
957 - implemented opcodes fillfo_depthchanaft,
958 fillfo_freqchanaft, amplfo_depthchanaft, amplfo_freqchanaft,
959 pitchlfo_depthchanaft, pitchlfo_freqchanaft
960 - implemented opcodes pitch_onccN,
961 pitch_curveccN, pitch_smoothccN, pitch_stepccN
962 - implemented opcodes volume_stepccN, pan_stepccN,
963 cutoff_stepccN, resonance_stepccN, lfoN_freq_stepccX,
964 lfoN_volume_stepccX, lfoN_pitch_stepccX, lfoN_pan_stepccX,
965 lfoN_cutoff_stepccX, lfoN_resonance_stepccX
966 - implemented opcodes eq1_freq, eq2_freq, eq3_freq,
967 eq1_freqccN, eq2_freqccN, eq3_freqccN, eq1_bw, eq2_bw, eq3_bw,
968 eq1_bwccN, eq2_bwccN, eq3_bwccN, eq1_gain, eq2_gain, eq3_gain,
969 eq1_gainccN, eq2_gainccN, eq3_gainccN
970 - implemented opcodes delay, delay_onccN, delay_random,
971 delay_samples, delay_samples_onccN
972 - implemented opcodes egN_eq1gain, egN_eq2gain, egN_eq3gain,
973 egN_eq1gain_onccX, egN_eq2gain_onccX, egN_eq3gain_onccX, egN_eq1freq,
974 egN_eq2freq, egN_eq3freq, egN_eq1freq_onccX, egN_eq2freq_onccX,
975 egN_eq3freq_onccX, egN_eq1bw, egN_eq2bw, egN_eq3bw, egN_eq1bw_onccX,
976 egN_eq2bw_onccX, egN_eq3bw_onccX, lfoN_eq1gain, lfoN_eq2gain,
977 lfoN_eq3gain, lfoN_eq1gain_onccX, lfoN_eq2gain_onccX, lfoN_eq3gain_onccX,
978 lfoN_eq1gain_smoothccX, lfoN_eq2gain_smoothccX, lfoN_eq3gain_smoothccX,
979 lfoN_eq1gain_stepccX, lfoN_eq2gain_stepccX, lfoN_eq3gain_stepccX,
980 lfoN_eq1freq, lfoN_eq2freq, lfoN_eq3freq, lfoN_eq1freq_onccX,
981 lfoN_eq2freq_onccX, lfoN_eq3freq_onccX, lfoN_eq1freq_smoothccX,
982 lfoN_eq2freq_smoothccX, lfoN_eq3freq_smoothccX, lfoN_eq1freq_stepccX,
983 lfoN_eq2freq_stepccX, lfoN_eq3freq_stepccX, lfoN_eq1bw, lfoN_eq2bw,
984 lfoN_eq3bw, lfoN_eq1bw_onccX, lfoN_eq2bw_onccX, lfoN_eq3bw_onccX,
985 lfoN_eq1bw_smoothccX, lfoN_eq2bw_smoothccX, lfoN_eq3bw_smoothccX,
986 lfoN_eq1bw_stepccX, lfoN_eq2bw_stepccX, lfoN_eq3bw_stepccX
987 - implemented opcodes eq1_vel2freq, eq2_vel2freq,
988 eq3_vel2freq, eq1_vel2gain, eq2_vel2gain, eq3_vel2gain
989 - sfz parser: allow double spaces in sample filenames
990 - sfz parser: allow absolute paths for sample filenames
991 - use linear decay and release for filter and pitch EG
992 - bugfix: only the first amp_veltrack definition in a file was
993 used
994 - bugfix: looping was disabled if loop_start was set to 0
995 - allow regions with end=-1 to turn off other regions using the
996 group and off_by opcodes (#168)
997 - made end=0 play the whole sample
998 - fixed support for lochan and hichan opcodes (#155)
999 - fixed crash when using lochan/hichan opcodes (#187)
1000 - sfz parser: allow -200 to 200 for pan_oncc opcode (#182)
1001 - added FLAC support (#191)
1002 - sfz parser bugfix: lines starting with whitespace were ignored
1003 - added amplitude opcode
1004 - added support for "#include" instruction
1005 (modified patch which was originally posted by Sergey on LS mailing list)
1006 - bugfix: generation of velocity curves etc should not be done
1007 after each "#include", only after the main file is parsed
1008 - bugfix: line numbers in error messages were wrong after "#include"
1009 - added support for float and 32 bit sample files
1011 * SoundFont format engine:
1012 - Initial implementation (not usable yet)
1013 - RT-safeness: avoid malloc in audio thread
1014 - fixed a bug that could cause voice stealing to fail
1015 - fine-tuned amplitude EG (by switching from gig to sfz EG)
1016 - initial implementation of Vibrato LFO and Modulation LFO
1017 - initial implementation of cutoff filter
1018 - use linear decay and release for filter and pitch EG
1020 * Host plugins (VST, AU, LV2, DSSI):
1021 - AU bugfix: failed to destroy its audio/MIDI devices
1022 - Listen to all interfaces on Mac OS X (INADDR_ANY)
1023 - VST bugfix: If the host called resume() before and after
1024 changing sample rate or block size, the number of channels was
1025 incorrectly set to two. This caused silence in Cubase 5.
1026 - save engine type (gig, sfz or sf2) in plugin state
1027 - VST: when opening Fantasia, look for both 32 and 64 bit Java on
1028 64 bit Windows
1029 - AU: changed number of output channels from one stereo to 16
1030 stereo
1031 - VST: made it possible to build the VST plugin for Mac
1032 - AU: link AU plugin dynamically if --disable-shared isn't
1033 specified
1034 - LV2 "state" extension support (patch by David Robillard)
1035 - VST bugfix: instrument loading hang and crashed the host when
1036 the plugin was loaded a second time (#174)
1037 - plugin bugfix: instrument loading hang when the plugin was
1038 loaded a second time (this time it's for Linux and Mac, previous
1039 similar fix was for Windows)
1040 - thread safety fixes for the instrument loading thread
1041 - LV2: use the new lv2 package if present
1042 - VST: try to open Fantasia automatically on Linux and Mac too (on
1043 Linux, the Fantasia jar should be placed in <prefix>/share/java)
1044 - VST: fixed crashes on Linux Ardour and EnergyXT
1045 - DSSI bugfix: it wasn't possible to change engine type. The MIDI
1046 port and audio channel routing for DSSI plugins are now visible.
1047 - LV2: use urid and atom extensions instead of deprecated uri-map
1048 and event
1049 - LV2: lv2 package 1.0 is now required to build the LV2 plugin
1050 - LV2: changed number of output channels to 16 stereo, just like
1051 the VST and AU plugins
1052 - LV2: fixed save/restore of SFZ state (patch by David Robillard)
1053 - LV2: made LV2 plugin buildable on Windows and Mac
1054 - VST: implemented retrieval and switching of programs using the
1055 sampler's internal MIDI instrument mapping system
1057 * Instrument editor interface:
1058 - Changed instrument editor plugin interface, providing additional
1059 informations like the EngineChannel for which the instrument editor was
1060 spawned for. This allows the instrument editors to interact more actively
1061 with the sampler.
1063 * MIDI driver:
1064 - ALSA MIDI driver supports now "NAME" device parameter, for overriding
1065 the ALSA sequencer client name
1066 - removed limit of maximum amount of MIDI ports per MIDI device, since
1067 there is no reason for this limit
1068 - MME: fixed memory handling bug found with cppcheck
1069 - MME: removed compiler warning
1070 - CoreMIDI: implemented driver specific port parameter "CORE_MIDI_BINDINGS",
1071 which allows to retrieve the list of CoreMIDI clients / ports and to connect
1072 to them a la JACK, via the usual sampler APIs
1073 - CoreMIDI: added driver specific port parameter "AUTO_BIND", if enabled
1074 the driver will automatically connect to other CoreMIDI clients' ports
1075 (e.g. external MIDI devices being attached to the Mac)
1076 - added support for MIDI note on velocity filter
1077 - CoreMIDI: fixed memory deallocation error
1078 - Fixed variable underflow in VirtualMidiDevice, which caused graphical
1079 virtual keyboards in frontends / instrument editors being stuck.
1080 - Bugfix in VirtualMidiDevice: process note on with velocity 0 as note off.
1081 - Implemented missing handling of MIDI "running status".
1082 - CoreMIDI fix: a MIDIPacket can contain more than one event per packet.
1083 - MME bugfix: driver wasn't closed properly
1084 - CoreMIDI: automatically connect to all input sources by default (driver
1085 parameter "AUTO_BIND").
1086 - CoreMIDI: fixed auto bind feature to CoreMIDI ports that go online
1087 - CoreMIDI: fixed minor memory leak
1089 * audio driver:
1090 - ASIO driver fixes for newer gcc versions (fix from PortAudio)
1091 - JACK audio: react on sample rate changes.
1092 - JACK audio: react on buffer size changes.
1093 - JACK audio: jack_port_get_buffer() was cached and called outside
1094 RT context.
1095 - ASIO driver: removed compiler warnings
1096 - CoreAudio: fixed minor error handling bug
1097 - ASIO driver: be more verbose when no ASIO card could be found (fixes #203)
1098 - JACK audio: return the JACK server's current sample rate as default value
1099 for audio device parameter "SAMPLERATE" (fixes #166).
1101 * LSCP server:
1102 - added support for sending MIDI CC messages via LSCP command
1103 "SEND CHANNEL MIDI_DATA CC <sampler-chan> <ctrl> <val>"
1104 - added LSCP command "GET AVAILABLE_EFFECTS"
1105 - added LSCP command "LIST AVAILABLE_EFFECTS"
1106 - added LSCP command "GET EFFECT INFO <effect-index>"
1107 - added LSCP command "CREATE EFFECT_INSTANCE <effect-index>"
1108 - added LSCP command
1109 "CREATE EFFECT_INSTANCE <effect-system> <module> <effect-name>"
1110 - added LSCP command "DESTROY EFFECT_INSTANCE <effect-instance>"
1111 - added LSCP command "GET EFFECT_INSTANCES"
1112 - added LSCP command "LIST EFFECT_INSTANCES"
1113 - added LSCP command "GET EFFECT_INSTANCE INFO <effect-instance>"
1114 - added LSCP command
1115 "GET EFFECT_INSTANCE_INPUT_CONTROL INFO <effect-instance> <input-control>"
1117 <effect-instance> <input-control> <value>"
1118 - added LSCP command "GET SEND_EFFECT_CHAINS <audio-device>"
1119 - added LSCP command "LIST SEND_EFFECT_CHAINS <audio-device>"
1120 - added LSCP command "ADD SEND_EFFECT_CHAIN <audio-device>"
1121 - added LSCP command
1122 "REMOVE SEND_EFFECT_CHAIN <audio-device> <effect-chain>"
1123 - added LSCP command
1124 "GET SEND_EFFECT_CHAIN INFO <audio-device> <effect-chain>"
1125 - added LSCP command "APPEND SEND_EFFECT_CHAIN EFFECT <audio-device>
1126 <effect-chain> <effect-instance>"
1127 - added LSCP command "INSERT SEND_EFFECT_CHAIN EFFECT <audio-device>
1128 <effect-chain> <effect-chain-pos> <effect-instance>"
1129 - added LSCP command "REMOVE SEND_EFFECT_CHAIN EFFECT <audio-device>
1130 <effect-chain> <chain-pos>"
1131 - added LSCP command "SET FX_SEND EFFECT <sampler_channel>
1132 <fx_send_id> <effect_chain> <chain_pos>"
1133 - added LSCP command "REMOVE FX_SEND EFFECT <sampler_channel> <fx_send_id>"
1137 - provide comprehensive error messages on LSCP syntax errors
1138 (suggesting expected next non-terminal symbols)
1139 - Fixed client connection not being closed after network errors.
1141 * LSCP shell:
1142 - Added support for auto correction of obvious and trivial syntax mistakes.
1143 - Added support for auto completion by tab key.
1144 - Show currently available auto completion while typing.
1145 - Added support for browsing command history with up / down keys.
1146 - Show all possible next symbols immediately right to the current command
1147 line while typing (no double tab required for this feature, as it would
1148 be the case in other shells).
1149 - Added support for moving cursor left/right with arrow keys.
1150 - Added support for built-in LSCP reference documentation, which will
1151 automatically show the relevant LSCP reference section on screen as soon
1152 as one specific LSCP command was detected while typing on the command
1153 line.
1155 * Real-time instrument scripts:
1156 - Implemented built-in script array variable %CC.
1157 - Implemented built-in script int variable $CC_NUM.
1158 - Implemented built-in script int variable $EVENT_NOTE.
1159 - Implemented built-in script int variable $EVENT_VELOCITY.
1160 - Implemented built-in script constant variable $VCC_MONO_AT.
1161 - Implemented built-in script constant variable $VCC_PITCH_BEND.
1162 - Implemented execution of script event handler "init".
1163 - Implemented execution of script event handler "controller".
1164 - Implemented execution of script event handler "note".
1165 - Implemented execution of script event handler "release".
1166 - Implemented built-in script function "play_note()" (only two of the
1167 max. four function arguments are currently implemented yet though).
1168 - Implemented built-in script int variable $EVENT_ID.
1169 - Implemented built-in script function "ignore_event()"
1170 - Implemented built-in script function "ignore_controller()" (may have one
1171 or no argument).
1172 - Implemented built-in script function "set_controller()".
1173 - Added extended script VM for the Gigasampler/GigaStudio format sampler
1174 engine, which extends the general instrument script VM with Giga format
1175 specific variables and functions.
1176 - Giga format scripts: added built-in script int constant variables
1187 - Giga format scripts: Implemented built-in script function
1188 "gig_set_dim_zone(event_id, dimension, zone)".
1189 - Implemented built-in script int array variable %KEY_DOWN.
1190 - Implemented built-in script function "abs()".
1191 - Implemented built-in script function "random()".
1192 - Implemented built-in script function "num_elements()".
1193 - Implemented built-in script function "note_off()".
1194 - Implemented built-in script function "set_event_mark()".
1195 - Implemented built-in script function "delete_event_mark()".
1196 - Implemented built-in script function "by_marks()".
1197 - Added built-in script int const variables $MARK_1 to $MARK_28.
1198 - Built-in script functions "ignore_event()", "note_off()" and
1199 "gig_set_dim_zone()" now also accept an array of event IDs as argument
1200 (i.e. return value of new script function "by_marks()").
1201 - Pass/preserve polyphonic variable data from respective "note" event
1202 handler to "release" event handler.
1203 - Fixed crash when using built-in script function "by_marks()".
1205 * Bug fixes:
1206 - Fixed crash which may occur when MIDI key + transpose is out of range
1207 - minor valgrind fixes
1208 - fixed crash which occurred when changing an already deployed sampler
1209 channel to a different engine type
1210 - fixed crash when deleting a sampler channel or changing engine
1211 type while an instrument load was in progress
1212 - bugfix: playing a note while changing the instrument could cause
1213 a crash, or give "Handing back unknown region" error messages
1214 - bugfix: calling SET VOICES (which Fantasia does on start and
1215 refresh) could cause instruments to be unloaded
1216 - Mac OS X: fixed name collision of enum in EffectControl and
1217 wrong name of destructor in AudioOutputDeviceCoreAudio.cpp
1218 - Mac OS X: fixed hanging threads
1219 - Windows: disabled the previous thread fix on non-Mac systems, as
1220 it caused hanging threads on Windows
1221 - Fixed possible crashes due to corrupted MIDI/audio device list
1222 after MIDI/audio device creation failure
1223 - When creating MIDI instrument map entries with "PERSISTENT" type, the
1224 instruments were uselessly precached with zero samples, however it still
1225 took the full preloading time and on 1st program change the respective
1226 instrument was completely reloaded again.
1227 - fixed handling of rapid bank select and program change messages
1228 sent to the same sampler channel (patch from the Open Octave
1229 project, slightly adjusted)
1230 - fixed crash when trying to create an effect instance with controls
1231 which min and/or max values depend on the sample rate
1232 - fixed bug #162
1233 - bugfix: LADSPA_PATH was not evaluated correctly when containing
1234 multiple paths (#165)
1235 - thread safety fixes for the instrument loading thread
1236 - bugfix: instrument loading crashed for sfz and sf2 in Ardour
1237 (#176)
1238 - more thread safety fixes for the instrument loading thread
1239 - sfz/sf2 engine: fixed crash when using small audio fragment size
1240 - Mac OS X: fixed crash when unloading plugin on 10.7 and later
1241 - Mac OS X: fixed process hang when unloading 32-bit plugin (bug
1242 introduced in previous fix)
1243 - fixed crash when a channel received a program change while
1244 playing a note in a key group
1245 - fixed erroneous error message piping in VoiceBase.h
1246 (labelled "Disk stream not available in time")
1247 - Update effects on sample rate & period size changes (to avoid
1248 crashes and noise on such transitions).
1249 - ignore missing LADSPA paths without ignoring valid LADSPA paths
1250 (fixes #208)
1251 - Fixed bug in LSCP grammar definition which caused a statement like
1252 "GET SERVER INFOasdf\n" to be accepted as valid statement (was so far
1253 practically irrelevant, however it caused problems with the new LSCP
1254 shell's auto completion feature).
1255 - Fixed MIDI program change messages being ignored if quickly executed
1256 after each other (fixes #231).
1258 Version 1.0.0 (31 July 2009)
1260 * packaging changes:
1261 - autoconf bugfix: the PKG_CONFIG variable wasn't initialized properly,
1262 causing e.g. the libgig test to fail when
1263 "./configure --disable-jack-driver" was used
1264 (patch by Alexis Ballier)
1265 - fixed compilation with gcc 4.3
1266 - fixes for building on OS X (thanks to Ebrahim Mayat for testing)
1267 - fixed configure so it detects x86_64 (#107)
1268 - fixes for building with newer MinGW versions
1269 - fix for building with bison 2.4 (#111)
1270 - fixed building with libgig installed in a non-standard directory
1271 - minor fix in configure for mmsystem.h detection on MinGW
1272 - Windows: look for editor plugins and Fantasia using base
1273 directory of liblinuxsampler dll (look in the same directory and one
1274 directory above)
1275 - configure script fix: removed unconditional use of SSE
1276 - fixed building with sqlite installed in a non-standard directory
1277 - when cross-compiling, don't try to create instruments.db
1278 - fix for new mingw-w64 version, which has usleep
1280 * general changes:
1281 - bugfix: on some POSIX systems instrument editor plugins refused to
1282 load as we used a non-portable Linux specific struct field
1283 (fixes bug #70, patch by Ronald Baljeu)
1284 - fixed endless loop which occured when loading an instrument editor
1285 plugin DLL on Windows systems failed
1286 - fixed memory leaks that occurred when liblinuxsampler was unloaded
1287 - optimized the SynchronizedConfig class so it doesn't wait
1288 unnecessarily long after an update
1289 - added support for notifying instrument editors on note-on / note-off
1290 events (e.g. to highlight the pressed keys on the virtual keyboard
1291 of gigedit)
1292 - added support for triggering notes by instrument editors (see above)
1293 - be verbose on DLL load errors (on Linux)
1294 - fixed di-harmonic triangle LFO implementation (this LFO implementation
1295 is only used by very few systems, most pick the int math
1296 implementation, which is usually faster)
1297 - fixes for audio drivers with varying buffer sizes
1298 - experimental support for running LinuxSampler as a DSSI, LV2 and
1299 VST plugin
1300 - notification events for stream/voice count statistics are now sent
1301 only when there are actual changes
1302 - added memory ordering constraints to improve stability on
1303 multi-core and multi-cpu systems
1304 - maximum voices and disk streams can now be altered at runtime
1305 - fixed CPU feature detection on x86_64 (maybe fixes #108)
1306 - automatic stacktrace mechanism is now turned off by default and can
1307 be switched on by command line option "--stacktrace" (the automatic
1308 stacktrace mechanism seems to be broken on most systems at the moment)
1309 - C++ API method InstrumentManager::LaunchInstrumentEditor() now returns
1310 a pointer to the launched InstrumentEditor object
1311 - added optional 3rd party user data parameter for following
1312 liblinuxsampler C++ API methods: InstrumentEditor::Main(),
1313 InstrumentEditor::Launch(),
1314 InstrumentManager::LaunchInstrumentEditor()
1315 - theoretical fix: made SynchronizedConfig follow C++0x memory
1316 model more strictly
1317 - fixes for using large audio device buffers
1318 - Windows: add the installation directory to the DLL search path
1319 when loading an editor plugin (solves problems with VST and
1320 gigedit on systems with other GTK versions installed)
1321 - updated linuxsampler man page
1323 * audio driver:
1324 - removed the nonsense audio channel constraint (which was hard coded to
1325 max. 100 audio channels) for most audio drivers
1326 - JACK audio driver did not offer a device parameter "SAMPLERATE" as
1327 opposed to the LSCP specs
1328 - bugfix: the SAMPLERATE parameter of some drivers (e.g. JACK)
1329 reflected the wrong value
1330 - fixed a memory management bug in ASIO driver
1331 - Makefile fix: JACK_CFLAGS wasn't used
1332 - JACK: use jack_client_open instead of the deprecated
1333 jack_client_new
1334 - added (experimental) CoreAudio driver
1335 - applied old fixes to the ASIO driver that were included in the
1336 previous binary release but accidentally never committed to CVS
1337 (fixes #117)
1338 - fixes for ASIO on mingw-w64 (iasio wrapper is not needed on
1339 win64)
1340 - VST: added support for sample rate and buffer size changes
1341 - VST: close editor (Fantasia) when the VST is removed
1342 - VST: avoid opening Fantasia more than once for each VST instance
1343 - VST: export main function as "main" on Linux too (fix for
1344 energyXT)
1345 - VST: changed number of output channels from one stereo to 16
1346 stereo
1347 - added channel routing, fxsends and midi maps to the settings
1348 stored in the plugin state
1349 - performance optimization of AudioChannel::MixTo() and
1350 AudioChannel::CopyTo() methods using GCC vector exensions
1351 (if available)
1352 - ASIO fixes: avoid initializing the device twice, avoid throwing
1353 exception when getting parameters from a disconnected device
1355 * MIDI driver:
1356 - added JACK MIDI driver
1357 - dispatch bank select as ordinary CC as well, the user might seriously
1358 want to (mis)use it for some purpose ("fixed" in all current MIDI
1359 input drivers: ALSA, CoreMIDI, JACK, MidiShare, MME)
1360 - bugfix: pitch bend wasn't working with JackMidi, VST, LV2, MME,
1361 CoreMidi or AU
1362 - fixed mingw-w64 compilation error in MME driver
1363 - made program change handling in MIDI thread real-time safe by
1364 moving the logic to a non-RT thread
1365 - fixed minor memory leak in ALSA driver
1367 * instruments database:
1368 - avoid time consuming samples scanning when adding instruments
1369 to the instruments database
1370 - added support for handling lost files in the instruments database
1371 - Implemented option for adding instruments in separate directories
1372 in the instruments database
1373 (patch by Chris Cherrett & Andrew Williams, a bit adjusted)
1374 - work-around for missing fnmatch function on Windows to make
1375 instrument database compilable
1376 - added instrument database support on Windows
1377 works with both standalone mode and VST plugin.
1378 the instruments DB file is located in
1379 %USERPROFILE%\.linuxsampler\instruments.db which allows different
1380 databases for each windows user
1381 if no DB is present it automatically creates the .linuxsampler subdir
1382 and then creates an empty DB
1383 - fixed recursive import, which was broken on Windows
1385 * Gigasampler format engine:
1386 - fixed a memory leak that could happen when a channel was deleted
1387 while notes were playing
1388 - made it possible to load an instrument even if the audio thread
1389 isn't running
1390 - added partial support for the "Controller Triggered" MIDI rule,
1391 enough for piano gigs with pedal noise samples
1392 - minor fix: only mark FX sends as being modified if really the
1393 respective FX send MIDI controller was used
1394 - added support for GM global device volume SysEx message
1395 (can be selected at compile time to either apply globally to the
1396 whole sampler [default] or only to the sampler channels that are
1397 connected to the respective MIDI input port the SysEx message arrived
1398 on)
1399 - bugfix: notes triggered at position 0 in the audio buffer were
1400 sometimes wrongly killed in the same buffer, causing no sound to
1401 be played
1402 - added support for chromatic / drumkit mode Roland GS Sysex message
1403 which is usally used to switch a part between chromatic sounds and
1404 drumkit sounds (as opposed to their common default setting of MIDI
1405 part 10 being a drumkit part and all other ones chromatic parts), in
1406 LS however one can switch between the first 16 MIDI instrument maps
1407 defined for the sampler
1408 - made it possible to create multiple sample channels even if the
1409 audio thread isn't running
1410 - fixed crash when removing channel with active voices (#116)
1411 - bugfix: on sample reference changes (instrument editor), only
1412 un-cache the respective sample if it's really not used by any
1413 sampler engine anymore
1414 - re-cache samples in case they were changed by an instrument editor,
1415 e.g. when a sample was added while playing (#82)
1416 - fixed hanging notes which occured when note-off event had the exact
1417 same time stamp as the voice's note-on event and occured both in the
1418 same audio fragment cycle (fixes bug #112)
1419 - added support for the "fine tune" and "pitch bend range"
1420 instrument-level gig parameters
1421 - fixed minor artifacts in pitch bend handling
1422 - added support for GS Reset SysEx message
1423 - allow gig files to use unlimited downward pitch shifting
1424 - added a limit check for upward pitch shifting
1425 - bugfix: sometimes, when playing a note twice fast, the second
1426 note was silent
1427 - fixed crash happening when a pitch bend event arrived at the
1428 same time a new instrument was loading
1430 * LSCP server:
1431 - added new LSCP event "CHANNEL_MIDI" which can be used by frontends to
1432 react on MIDI data arriving on certain sampler channels (so far only
1433 Note-On and Note-Off events are sent via this LSCP event)
1434 - added new LSCP event "DEVICE_MIDI" which can be used by frontends to
1435 react on MIDI data arriving on certain MIDI input devices (so far only
1436 Note-On and Note-Off events are sent via this LSCP event)
1437 - added new LSCP commands: FIND LOST DB_INSTRUMENT_FILES and
1439 - added new LSCP command: SEND CHANNEL MIDI_DATA which can be used by
1440 frontends to send MIDI messages to specific sampler channel
1441 - added two additional fields to GET FILE INSTRUMENT INFO command -
1443 - bugfix: the bank number provided by MIDI_INSTRUMENT_INFO notifications
1444 was incorrect
1445 - Optimized the retrieval of the MIDI instrument mappings
1446 - added new LSCP command "SET VOICES" to globally alter the maximum
1447 amount of voices, added new LSCP event "GLOBAL_INFO:VOICES" which will
1448 be triggered respectively
1449 - added new LSCP command "SET STREAMS" to globally alter the maximum
1450 amount of disk streams, added new LSCP event "GLOBAL_INFO:STREAMS"
1451 which will be triggered respectively
1452 - bugfix: retry if "select" returns EINTR (this fixes a crash when
1453 a gigedit file dialog is opened)
1454 - close all connections when LSCPServer is deleted
1455 - hard close of all sockets on exit
1456 - bugfix: SET CHANNEL MIDI_INPUT_TYPE didn't work with the MME
1457 driver
1459 * Bug fixes:
1460 - fixed a crash which occurs when removing a sampler channel waiting
1461 to start instrument loading after another channel
1462 - fixed a crash which occurs when removing a sampler channel with
1463 instrument loading in progress (bug #113)
1464 - fixed termination caused by uncaught exception when adding MIDI
1465 instrument with PERSISTENT load mode
1466 - fixed possible iterator invalidations when resetting the sampler
1467 - fixed memory leaks when issuing the following LSCP commands:
1472 - fixed possible compilation error when sqlite is not present
1473 - fixed orphaned pointers when setting maximum voices limit (bug #118)
1474 - fixed crash when changing the audio output device of a sampler
1475 channel with loaded instrument and start playing notes
1476 - fixed endless loop in Engine::SuspendAll() (bug #120)
1477 - fixed a low-level atomic load function that was broken on 64-bit
1478 PowerPC, which probably could cause crashes on that platform
1479 - fixed a memory management error which could cause a crash when a
1480 plugin was unloaded
1481 - bugfix: two private structs had the same name, which could cause
1482 problems if the linker chose the wrong constructor
1483 - fixed low-level ConditionServer usage bug that caused lockups on
1484 Windows
1487 Version 0.5.1 (6 December 2007)
1489 * packaging changes:
1490 - added autoconf checks for pthread library
1491 - added autoconf check for pthread bug found on certain NPTL-enabled
1492 glibc versions (see Gentoo bug report #194076)
1493 - added autoconf checks for MS Windows
1495 * general changes:
1496 - bugfix: the thread used by an editor plugin didn't die when the
1497 editor closed
1498 - bugfix: Ringbuffer.h: fill_write_space_with_null() did not zero
1499 out all the space. operator--() did not apply size_mask after
1500 decrementing the read_ptr. DEFAULT_WRAP_ELEMENTS set to 0 in
1501 order to avoid problems with the _NonVolatileReader functions.
1502 - bugfix: Stream.h: added a missing divide by BytesPerSample in
1503 GetWriteSpace(). Since this function is currently only used in
1504 the stream's qsort() compare function, it didn't trigger any
1505 bugs.
1506 - Resampler.h, Synthesizer.h: cubic interpolation now works in
1507 24bit mode too. Faster method to read 24bit words on little
1508 endian machines (x86): replaced 3 byte reads + shifts with a 1
1509 unaligned 32bit read and shift
1510 - experimental support for MS Windows (MIDI input via MME, AUDIO
1511 output via ASIO)
1512 - made handling of SIGINT signal (Ctrl-C) a bit more robust
1513 - support for monitoring the total number of active disk streams
1514 (new LSCP commands: GET TOTAL_STREAM_COUNT,
1517 * AUDIO driver:
1518 - added Windows ASIO low latency audio driver
1520 * MIDI driver:
1521 - added MME Windows MIDI driver
1523 * LSCP server:
1524 - added support for Windows style path / filenames, however with
1525 forward slash path separators instead of backslash
1526 (i.e. "C:/foo/bar.gig")
1527 - allow naughty liblscp to send non-string device parameters within
1528 apostrophes as well
1529 - added new LSCP commands: "GET FILE INSTRUMENTS <file>",
1530 "LIST FILE INSTRUMENTS <file>" and
1531 "GET FILE INSTRUMENT INFO <file> <index>" for retrieving informations
1532 about an arbitrary instrument file on the system where the sampler is
1533 running on
1535 Version 0.5.0 (15 October 2007)
1537 * packaging changes:
1538 - config.h is not going to be installed along with liblinuxsampler's
1539 API header files anymore
1540 - only the API relevant header (and source) files will be exported to
1541 the Doxygen API documentation (explicitly listed in Doxyfile.in)
1542 - added completely new XCode project files for Mac OSX which is now
1543 capable to execute our autoconf environment, thus no need anymore
1544 to maintain the compile time configuration file (osx/version.h) for
1545 OSX manually (patch by Toshi Nagata)
1546 - fixed buggy boolean --enable-foo / --disable-foo configure script
1547 parameters
1548 - global.h now only covers global definitions that are needed for the
1549 C++ API header files, all implementation internal global definitions
1550 are now in global_private.h
1551 - atomic.h is not exposed to the C++ API anymore
1552 - no need to include config.h anymore for using LS's API header files
1553 - fixed warnings in API doc generation
1555 * general changes:
1556 - replaced the old, confusing MIDI program change mechanism by a
1557 flexible MIDI instrument mapper which allows to map arbitrary
1558 (MIDI bank, MIDI program) pairs with arbitrary
1559 (engine type, instrument file, file index) triplets which will be
1560 loaded on the respective channel when such MIDI program change
1561 messages arrive, beside that, each entry allows to define a life-time
1562 strategy for the instrument, so the user can define whether the
1563 instrument should i.e. be permanently loaded (to be able to switch
1564 quickly among sounds) or i.e. loaded just on demand when the
1565 respective program change arrives (to safe RAM space), as well as a
1566 global volume factor for each entry, so the user can adjust the volume
1567 dynamic of his mapped instrument collection without having to modify
1568 the instrument files, also one can manage arbitrary amount of such
1569 MIDI instrument maps and assign each sampler channel individually
1570 a certain map, so that i.e. separation of normal instruments and
1571 drumkits is possible
1572 - new notification events for tracking changes to audio/MIDI devices,
1573 MIDI instrument maps, MIDI instruments, FX sends, global volume.
1574 - sampler was limited to load max. 200 instruments in the background
1575 due to a constant size RingBuffer FIFO which is now replaced by a
1576 dynamic (unlimited) size std::list FIFO
1577 - added FX sends, these allow to route audio signals to arbitrary audio
1578 output channels for being processed by external effect processors
1579 (i.e. jack-rack), the send levels are controllable via arbitrary MIDI
1580 controllers
1581 - global (sampler wide) volume can now be controlled at runtime
1582 - Implemented new, improved notification system
1583 - fixed compilation errors regarding OSX
1584 (patch by Toshi Nagata)
1585 - implemented instruments database
1586 - added support for escape sequences to the instruments database
1587 - added highly experimental support for on-the-fly instrument editing
1588 within the sampler's process (by using instrument editor plugins),
1589 you'll notice the new "Registered instrument editors:" message on
1590 startup, added a new LSCP command:
1591 "EDIT CHANNEL INSTRUMENT <sampler-channel>"
1592 to spawn a matching instrument editor for the instrument on the
1593 given sampler channel, the plugin path can be overridden at compile
1594 time with ./configure --enable-plugin-dir=/some/dir
1595 - added experimental code for synchronizing instrument editors hosted
1596 in the sampler's process to safely edit instruments while playing
1597 without a crash (hopefully) by either suspending single regions
1598 wherever possible or if unavoidable whole engine(s)
1599 - fixed several issues in fundamental "Thread" class: set scheduling
1600 policy and priority on thread level, set a minimum stack size for
1601 thread (TODO: a reasonable value yet to be tested), bugfix: non-RT
1602 threads simply inherited properties of starting thread instead of
1603 setting their own policy and priority
1604 - minor fix in our automatic stack trace mechanism on crashes, the
1605 main process did not wait for the stack trace process to finish
1606 its output
1607 - fixed some minor memory leaks
1608 - reenabled assembly features support, at the moment only for
1609 enabling a fast denormal FPU mode (x86 platforms supporting SSE2)
1610 - minor assembly fix in x86 features detection (don't use the PIC
1611 register, to avoid relocations in the text segment at runtime)
1612 - POSIX callback functions of Thread.h are hidden
1614 * MIDI driver:
1615 - dispatch bank select (MSB and LSB) messages
1617 * audio driver:
1618 - the ALSA audio output driver parameters now reflect the correct
1619 parameter value ranges for the respective selected sound card
1620 (patch by Till Wimmer, a bit fixed and extended)
1622 * Gigasampler format engine:
1623 - if a filter is used and EG2 finishes before EG1, let the voice
1624 die when EG2 has finished (this fixes a problem with clicks and
1625 voice starvation for some gigs)
1626 - playback is no longer disabled during instrument loading
1627 - all notes playing on a channel that changes its instrument keep
1628 playing with the old instrument until they get a note off
1629 command
1630 - EG fix: a release value of zero could cause noises or crash
1631 - handle MIDI coarse tuning messages (MIDI RPN #0 MSB #2 LSB)
1632 - EG fine tuning: when attack is zero the EG starts at a level
1633 above max sustain level, which means that there is a decay phase
1634 even if sustain is 100%
1635 - more EG fixes: the level could sometimes go below zero and cause
1636 noises or crashes
1637 - minor fix of EGDecay (patch by Toshi Nagata)
1638 - fixed compiler error when --enable-override-filter-type was
1639 supplied to the configure script (fixes #46)
1640 - disk thread: queue sizes are now proportional to CONFIG_MAX_STREAMS
1641 instead of a fix value
1642 - behavior fix: on MIDI CC# 65 (portamento on / off), 126 (mono mode),
1643 127 (solo mode) only kill voices if the respective mode really
1644 changed
1646 * LSCP server:
1647 - fixed compile time error for old Bison versions
1648 (i.e. v1.28 found on MacOS 10.4, patch by Toshi Nagata)
1649 - parser now supports extended ASCII character set
1650 (up to ASCII code 255, i.e. includes now umlauts and accents)
1651 - filename arguments in LSCP commands now allow to use escape
1652 sequences, that is directly literal as one of: \', \", \\, \n, \r,
1653 \f, \t, \v, or as octal ASCII code value like \132, or as hex ASCII
1654 code value like \xf2) (fixes bug #24)
1655 - the following LSCP commands now also support escape sequences for at
1656 least one of their text-based parameters (i.e. name, description):
1663 - returns verbose syntax errors (line and column where syntax error
1664 occured, the unexpected character and the actually expected, possible
1665 character(s), the latter only if less than 5 possibilities)
1666 - made sure that LSCP syntax is not affected by gigedit locale
1667 settings
1668 - bugfix regarding strings parameter lists: all comma separated lists
1669 of strings were treated as being one string containing commas
1670 (fixes #57)
1672 * Bug fixes:
1673 - fixed crash occurring on certain LSCP scripts (Bug 39)
1674 - another thread safety fix for lscp "load engine" and "set
1675 channel audio output device"
1676 - fixed a crash which occurs when reassigning the same engine
1677 on a sampler channel with connected MIDI device
1678 - fixed a crash which occurs when changing the number of ports of a MIDI
1679 device connected to a sampler channel to number less then or equal
1680 to the index of the port to which the sampler channel is connected.
1681 - The previous bindings were not been disconnected when altering
1682 the ALSA_SEQ_BINDINGS parameter. Introduced a NONE keyword for
1683 unsubscribing from all bindings (e.g. ALSA_SEQ_BINDINGS=NONE).
1684 - The active stream/voice count statistic was incorrect.
1685 - notification events were not been sent for some sampler
1686 channel changes
1687 - added default min and max values to restrict the number of allowed
1688 audio output channels and MIDI input ports
1689 - the connection to the PCM interface is now closed when destroying
1690 an audio output device
1691 - files with slash in their path or filename could not be loaded
1693 * test cases:
1694 - updated and fixed (haven't been touched in a while)
1696 Version 0.4.0 (24 November 2006)
1698 * packaging changes:
1699 - changed deprecated copyright attribute to license; added ldconfig
1700 to post-(un)install steps; added devel package for liblinuxsampler;
1701 to linuxsampler.spec (RPM)
1702 - install necessary development header files for allowing 3rd party
1703 applications to link against liblinuxsampler
1704 - liblinuxsampler's API documentation can be generated with 'make docs'
1705 (Doxygen required)
1706 - added benchmark to automatically detect the best triangle LFO
1707 implementation (currently either an integer math solution or a
1708 di-harmonic approximation), automatic detection can be overriden
1709 with --enable-signed-triang-algo=x and --enable-unsigned-triang-algo=x
1710 configure script argument though (mandatory for cross-compilation)
1711 - do not automatically pick optimized gcc flags if the user already
1712 provided some on his own (as CXXFLAGS)
1713 - added compile time option to disable processing of All-Notes-Off MIDI
1714 messages
1715 - added compile time options to allow disabling the various audio and
1716 MIDI drivers
1717 - fixed automatic GCC CPU switch detection on PPC
1718 (patch by Ebrahim Mayat)
1720 * Gigasampler format engine:
1721 - extensive synthesis optimization
1722 (reimplementation of EGs and LFO(s), removed synthesis parameter
1723 prerendering and the synthesis parameter matrix in general, splitting
1724 each audio fragment into subfragments now where each subfragment uses
1725 constant pitch and filter coefficients. The volume coefficient is
1726 linearly interpolated inside a subfragment, unless
1727 --disable-interpolate-volume is set.)
1728 - fine tuning of the EG modulation parameters
1729 - improved filter cutoff calculation by adding support for the
1730 following gig parameters: Cutoff freq (used when no cutoff
1731 controller is defined), Control invert, Minimum cutoff, Velocity
1732 curve and Velocity range. The keyboard tracking now scales
1733 cutoff frequency instead of resonance.
1734 - added support for gig parameter Resonance.
1735 - fixed bug in sysex handling (patch by Juan Linietsky)
1736 - added global volume attenuation of -9 dB (0.35f) to prevent clipping
1737 which can be overridden with --enable-global-attenuation
1738 - EG fixes: made the length of "attack hold" stage more
1739 accurate. Release stage can now start before attack stage
1740 ends. Cancel release didn't work when sustain was zero. Attack
1741 time now has a minimal value to prevent clicks.
1742 - fixed pitch changes larger than one octave
1743 - fixed EG3 (pitch envelope) synthesis which was neutral all the time
1744 - implemented portamento mode and solo mode (a.k.a 'mono mode'):
1745 all modes can be altered via standard GM messages, that is CC5 for
1746 altering portamento time, CC65 for enabling / disabling portamento
1747 mode, CC126 for enabling solo mode and CC127 for disabling solo mode
1748 - fine tuning of the curves for volume (CC7), pan (CC10 and gig
1749 parameter) and crossfade
1750 - added support for the "attenuation controller threshold" gig
1751 parameter
1752 - added smoothing of volume changes caused by control change
1753 messages
1754 - sample loop parameters are now taken from the DimensionRegion
1755 instead of the wave chunk
1756 - fixed keyswitching for v3 gigs with a number of keyswitch splits
1757 not equal to a power of two
1758 - reimplementation of the filter algorithm. The new filter is more
1759 accurate and supports all gig filter types, including bandreject
1760 and lowpass turbo.
1761 - real support for 24 bit samples - samples are not truncated to
1762 16 bits anymore
1763 - support for aftertouch (channel pressure, not polyphonic
1764 aftertouch)
1765 - LFO1 behaviour fixed (dampening from max volume instead of
1766 amplifying from 0)
1768 * LSCP server:
1769 - fixed application exit on broken pipe error (fixes bug #20)
1770 - fixed the notification messages delay due to lack of
1771 network activity (fixes bug #26)
1772 - fixed parser bug which occured on space(s) within device parameters
1775 * audio driver:
1776 - added aRts audio output driver (by no means RT safe)
1778 * MIDI driver:
1779 - fixed legacy sysex code which caused dispatching of MIDI SysEx
1780 messages several times instead of once
1781 - API extension for MIDI drivers which already supply exact time stamps
1782 for events (i.e. for offline rendering based MIDI drivers)
1783 - added checks for bad MIDI data
1785 * linuxsampler application:
1786 - show available sampler engine types on startup
1788 * general changes:
1789 - support for muting sampler channels and solo mode of the same, two new
1790 LSCP commands ("SET CHANNEL MUTE" and "SET CHANNEL SOLO") and two new
1791 fields ("MUTE" and "SOLO") for command "GET CHANNEL INFO" were
1792 introduced for this, the behavior is the same like on a mixer console
1793 (patch by Grigor Iliev, a bit adjusted). Also added configure option
1794 --enable-process-muted-channels which can be used to enable the
1795 processing of muted channels.
1796 - support for sostenuto pedal
1797 - support for monitoring the total number of active voices
1800 - fixed some memory management errors
1801 - fixed some concurrency problems that could lead to crashes when
1802 LSCP commands were executed
1803 - fixed crash when instrument loading failed previously on the same
1804 sampler channel (fixes bug #36)
1806 Version 0.3.3 (15 July 2005)
1808 * packaging changes:
1809 - fixed compilation with gcc 4.0
1811 * Gigasampler format engine:
1812 - LFO preprocessing was performed even though the respective LFO was
1813 disabled by the instrument patch (minor efficiency fix)
1814 - if period time of chosen audio device is too small (< MIN_RELEASE_TIME)
1815 for volume ramp downs in the same fragment (needed for current voice
1816 stealing implementation) then simply inform the user with a warning
1817 about possible click sounds and reduce the volume ramp down
1818 appropriately instead of cancelling the audio device connection
1820 Version 0.3.2 (24 June 2005)
1822 * packaging changes:
1823 - updated autotools build files to compile on OS X
1824 (conditional compilation of CoreMIDI and MidiShare drivers)
1825 - hand-crafted assembly optimization code can be disabled with
1826 './configure --disable-asm' (definitely not recommended)
1827 - fixed 'make dist' rule to include all necessary files
1828 - require automake (>= 1.5) for 'make -f Makefile.cvs'
1829 (due to 'dist-bzip2' automake option)
1831 * Gigasampler format engine:
1832 - support for the gig parameters for "release velocity response" curves
1833 - fine tuning of the EGADSR envelope
1834 - volume of release triggered samples now depends on note-on velocity,
1835 note length and gig parameter "release trigger decay" instead of
1836 note-off velocity.
1837 - revised voice stealing
1838 (fixes crash and endless loop caused by voice stealing)
1839 - don't reset scale tuning on instrument or audio output device change
1840 - handle key group conflicts right at the beginning of each fragment
1841 instead of when the respective voice is actually launched
1842 (fixes undefined behavior if stolen voices belonged to a key group -
1843 this case was followed by a "killed voice survived" error message)
1844 - fixed minor issue with null/silence samples
1845 (those stole voices even though they don't need a voice at all which
1846 resulted in "voice stealing didn't work out" messages)
1847 - don't reset volume, pan, pitch and MIDI controller values on
1848 instrument or audio output device change
1850 * LSCP server:
1851 - fixed some crashes (patch by Grigor Iliev, fixes #19)
1852 - fixed LSCP event "CHANNEL_INFO" notification
1853 (e.g. did not notify on volume changes or MIDI program change events)
1855 * linuxsampler application:
1856 - added command line parameters --lscp-addr and --lscp-port to override
1857 default IP address and port of LSCP server
1859 Version 0.3.1 (24 May 2005)
1861 * initial release

  ViewVC Help
Powered by ViewVC