Age | Commit message (Collapse) | Author | |
---|---|---|---|
2018-05-19 | Range adjustments for 8-bit noise functions, now preserving a (very) slight ↵ | Mark Kriegsman | |
bit more precision, too. | |||
2018-05-16 | Fix discontinuties in noise functions, both 8- and 16- bit versions. If you ↵ | Mark Kriegsman | |
need the old behavior, you can edit fastled_config.h and change FASTLED_NOISE_FIXED to 0. Fixes #416. Thanks to everyone who reported this and helped track it down. | |||
2017-11-02 | Makefile Compilation Fix (#490) | Nico | |
* renamed boolean to bool * Rename byte to uint8_t * Fix compilation error: 'NULL' was not declared in this scope * Always typedef RoReg and RwReg for the AVR platform * Include stdlib.h if abs() is not available This is required as Arduino.h #define's its own abs() makro. Outside of Arduino environments stdlib is required. Also see: https://github.com/arduino/Arduino/issues/6684 * Replace digitalWrite() with Pin Class by FastLed * Include string.h for memset function This include is normally done by Arduino.h but still required for makefile environments $ grep string.h . -r ./Arduino.h:#include <string.h> * Include math.h for pow() math.h is normally included by Arduino.h but required for a makefile build $ grep math.h . -r ./Arduino.h:#include <math.h> | |||
2016-01-07 | 12% speedup in noise on AVR (net with previous commit) by using inline avg15 ↵ | Mark Kriegsman | |
on AVRs with MUL. Code size goes up 8 bytes for 16-bit, 3-D noise. | |||
2016-01-07 | 3% speedup in noise on AVR by eliminating a shift-loop. Math is identical, ↵ | Mark Kriegsman | |
output results tested and found identical in every case tested. | |||
2015-12-29 | Noise behavior can be reverted for backward-stability of animation design ↵ | Mark Kriegsman | |
with #define FASTLED_NOISE_ALLOW_AVERAGE_TO_OVERFLOW 1 placed BEFORE the #include of FastLED.h | |||
2015-12-29 | Fixed noise integer overflow/underflow that was causing discontinuities in ↵ | Mark Kriegsman | |
the output of the noise functions. The output should now be (nearly) continuous to the precision that the relevant numeric types permit, as designed and intended -- but it's going to look slightly different now that the overflow/underflow jumps are gone. | |||
2015-12-26 | Arduion 1.6.6 and later, when you Include a library from the Sketch menu, ↵ | Daniel Garcia | |
will include _every_ header file - instead of just the library's main header file. Guard against that particular [REDACTED]. | |||
2015-07-23 | PROGMEM made optional. Control via FASTLED_USE_PROGMEM, which defaults to ↵ | Mark Kriegsman | |
the right thing. | |||
2015-02-09 | Fix compilation errors surrounding namespace - also rename the namespace, ↵ | Daniel Garcia | |
having FastLED::FastLED was proving too ambiguous for teh compiler. | |||
2015-02-09 | SparkCore needs FastLED to be in a namespace. | Daniel Garcia | |
2014-12-14 | #105 | Daniel Garcia | |
2014-11-22 | We only need to see the FastLED version warning once, thank you very much | Daniel Garcia | |
2014-11-06 | Another 1-2% faster 8-bit noise on AVR using a small asm widget to replace ↵ | Mark Kriegsman | |
badly-generated C code. | |||
2014-11-06 | Slight speedup and code size reduction by changing 16-bit-shift+AND to 8-bit ↵ | Mark Kriegsman | |
shift (ASL) with no AND needed afterwards. The compiler should have done this itself but didn't; film at 0x0B. | |||
2014-11-06 | Speedup for 8-bit noise (~5%) by using avg7(u,v) | Mark Kriegsman | |
2014-11-05 | Shut up compiler warnings on arm | Daniel Garcia | |
2014-10-25 | resolving conflicts | Daniel Garcia | |
2014-10-24 | noise tweak | Daniel Garcia | |
2014-10-02 | Clean up warnings | Daniel Garcia | |
2014-10-02 | Fix indexing error in one of the noise functions | Daniel Garcia | |
2014-09-15 | Merge branch 'FastLED2.1' into parallel | Daniel Garcia | |
2014-07-10 | Fix compile errors for arduino 1.5.7 - no testing done yet. Issue #67 | Daniel Garcia | |
2014-06-29 | full digix set of pins | Daniel Garcia | |
2014-06-15 | tweaking noise bits for locus | Daniel Garcia | |
2014-06-05 | For the fill_raw functions for all of 8bit, 16bit, and 16bit into 8 bit, ↵ | Daniel Garcia | |
shift the centerpoint of the noise function from 127/32767 to 0. For the fill_noise functions this will result in the colors centering on reds instead of blues | |||
2014-06-05 | Fix a handful of stupid math/loop errors that were causing a lot of ↵ | Daniel Garcia | |
smearing/distortion | |||
2014-06-04 | Fixing noise on avr with the new raw noise functions | Daniel Garcia | |
2014-06-04 | fixing on avr noise generation, something was unhappy there... | Daniel Garcia | |
2014-06-04 | Tweaked the scaling values of noise8/16 to better line up with the real ↵ | Daniel Garcia | |
range of values (found via doing 4 billion runs over each on my laptop). Also added better gradient functions for 1d and 2d noise (though i'm not happy with the 1d distribution, yet). Finally, added raw versions of the noise functions to give back the unadjusted, unscaled, roughly -0.6,0.6 range of values. | |||
2014-06-03 | tweaking q44 values | Daniel Garcia | |
2014-06-03 | Make the default frequency 2.0 | Daniel Garcia | |
2014-06-03 | fixing merge | Daniel Garcia | |
2014-06-03 | Adding fractional types. Expanding the raw fill functions to include real ↵ | Daniel Garcia | |
16 bit as well as take parameters for frequency and amplitude. | |||
2014-06-02 | Restoring 'serpentine' functionality. Might be faster to move it out of the ↵ | Mark Kriegsman | |
loop, but at least it's functional this way. | |||
2014-06-02 | fix typo | Daniel Garcia | |
2014-06-02 | Only want progmem on avr | Daniel Garcia | |
2014-06-02 | Tweaking a bit. | Daniel Garcia | |
2014-06-02 | Here, have some more memory. PROGMEM has some cost, but getting 257 bytes ↵ | Daniel Garcia | |
of ram back is more important. | |||
2014-06-02 | Adding leds/[] operators on controllers and LEDS. | Daniel Garcia | |
2014-06-02 | Adding some sample fill functions and the NoisePlayground example for folks ↵ | Daniel Garcia | |
to play in. | |||
2014-06-01 | Adding 8 and 16 bit simplex noise implementations to the library. The 8 bit ↵ | Daniel Garcia | |
version can do over 30,000 noise points per second, the 16 bit version over 15,000. Also, add a debugging FPS counter to FastLED. |