*2008-03-26* |
Jean-Marc Valin | optimisation: shaving a few cycles off prev_cwrs* by... |
tree | commitdiff |

*2008-03-26* |
Jean-Marc Valin | optimisations: faster handling of the zero for compute_... |
tree | commitdiff |

*2008-03-26* |
Jean-Marc Valin | changed 1*rcp(x) to just rcp(x) |
tree | commitdiff |

*2008-03-26* |
Jean-Marc Valin | optimisation: intra_prediction() no longer needs to... |
tree | commitdiff |

*2008-03-25* |
Jean-Marc Valin | optimisation: The "simple" Rxy/sqrt(Ryy) case in alg_qu... |
tree | commitdiff |

*2008-03-25* |
Jean-Marc Valin | Decision on whether to use pitch is now taken only... |
tree | commitdiff |

*2008-03-25* |
Jean-Marc Valin | properly defined EPSILON for the float case |
tree | commitdiff |

*2008-03-25* |
Jean-Marc Valin | A bunch of pointers marked as "restrict" to ease the... |
tree | commitdiff |

*2008-03-25* |
Jean-Marc Valin | optimisation: spreading_func now in-place with no branc... |
tree | commitdiff |

*2008-03-25* |
Jean-Marc Valin | oops. find_max32() now uses VERY_LARGE32 (instead of... |
tree | commitdiff |

*2008-03-25* |
Jean-Marc Valin | optimisations: Another bunch of simplifications to... |
tree | commitdiff |

*2008-03-25* |
Jean-Marc Valin | optimisation: Making use of restrict in find_spectral_p... |
tree | commitdiff |

*2008-03-25* |
Jean-Marc Valin | optimisations: caching sign of x in alg_quant(), change... |
tree | commitdiff |

*2008-03-25* |
Jean-Marc Valin | Optimised intra prediction a bit -- removed a condition... |
tree | commitdiff |

*2008-03-25* |
Jean-Marc Valin | Removed support for more than one MDCT blocks per frame... |
tree | commitdiff |

*2008-03-25* |
Jean-Marc Valin | Removed the "pitch compression" in the residual quantis... |
tree | commitdiff |

*2008-03-24* |
Jean-Marc Valin | Unrolled the inner loop in vq_index() so that the codeb... |
tree | commitdiff |

*2008-03-24* |
Jean-Marc Valin | Making bits2pulses() use a fixed number of iterations... |
tree | commitdiff |

*2008-03-23* |
Jean-Marc Valin | include "dsplib.h" in fixed_c5x.h |
tree | commitdiff |

*2008-03-23* |
Jean-Marc Valin | replaced divisions by recoprocals in intra prediction... |
tree | commitdiff |

*2008-03-22* |
Jean-Marc Valin | defined find_max16 and overrode it for C55x |
tree | commitdiff |

*2008-03-22* |
Jean-Marc Valin | Made a second version of ec_{en|de}code optimised for... |
tree | commitdiff |

*2008-03-22* |
Jean-Marc Valin | No longer trying to save bits when encoding integers... |
tree | commitdiff |

*2008-03-22* |
Jean-Marc Valin | fixed-point: added cheap celt_div() division using... |
tree | commitdiff |

*2008-03-22* |
Jean-Marc Valin | more use of restrict |
tree | commitdiff |

*2008-03-22* |
Jean-Marc Valin | Using restrict to make it clear there's no aliasing... |
tree | commitdiff |

*2008-03-21* |
Jean-Marc Valin | Added a few "restrict" keywords and changed some divisi... |
tree | commitdiff |

*2008-03-21* |
Jean-Marc Valin | fixed TI fft code -- again |
tree | commitdiff |

*2008-03-21* |
Jean-Marc Valin | Removed potentially unused var in MDCT init |
tree | commitdiff |

*2008-03-21* |
Jean-Marc Valin | oops (TI FFT) |
tree | commitdiff |

*2008-03-21* |
Jean-Marc Valin | local var name maxval was shadowing the TI function... |
tree | commitdiff |

*2008-03-21* |
Jean-Marc Valin | make sure TRIG_UPSCALE is properly defined |
tree | commitdiff |

*2008-03-21* |
Jean-Marc Valin | fix for TI version of celt_maxabs16() |
tree | commitdiff |

*2008-03-21* |
Jean-Marc Valin | fixed-point: fix for 32-bit TI FFT |
tree | commitdiff |

*2008-03-21* |
Jean-Marc Valin | fixed-point: Wrapper for the 32-bit complex FFT used... |
tree | commitdiff |

*2008-03-21* |
Jean-Marc Valin | fixed-point: defined celt_maxabs16() as basic operator |
tree | commitdiff |

*2008-03-20* |
Jean-Marc Valin | fixed-point: MULT16_32_Q15 for TI DSP (not entirely... |
tree | commitdiff |

*2008-03-20* |
Jean-Marc Valin | s/ROUND/ROUND16/ |
tree | commitdiff |

*2008-03-20* |
Jean-Marc Valin | fixed-point: using TI intrinsic for celt_ilog2() if... |
tree | commitdiff |

*2008-03-20* |
Jean-Marc Valin | Wrapper for the TI dsplib FFT |
tree | commitdiff |

*2008-03-20* |
Jean-Marc Valin | Making the real/single FFT easier to replace |
tree | commitdiff |

*2008-03-16* |
Jean-Marc Valin | fixed-point: more TI macros. Comments on the existing... |
tree | commitdiff |

*2008-03-15* |
Jean-Marc Valin | Making sure not to use the C library calls directly |
tree | commitdiff |

*2008-03-15* |
Jean-Marc Valin | Some work on assertions. |
tree | commitdiff |

*2008-03-14* |
Jean-Marc Valin | New C55 macro |
tree | commitdiff |

*2008-03-14* |
Jean-Marc Valin | fixed-point: Fixed a "shift too large" warning by using... |
tree | commitdiff |

*2008-03-14* |
Jean-Marc Valin | fixed-point: playing it safe. SHL32() now automatically... |
tree | commitdiff |

*2008-03-14* |
Jean-Marc Valin | fixed-point: fixed a 16-bit issue in normalise_bands... |
tree | commitdiff |

*2008-03-14* |
Jean-Marc Valin | Removed unnecessary header inclusions |
tree | commitdiff |

*2008-03-13* |
Jean-Marc Valin | Adjusting/fixing warnings |
tree | commitdiff |

*2008-03-13* |
Jean-Marc Valin | fixed-point: Using reciproqual instead of DIV32_16... |
tree | commitdiff |

*2008-03-13* |
Jean-Marc Valin | Enabling warnings and restricting symbol visibility |
tree | commitdiff |

*2008-03-13* |
Jean-Marc Valin | fixed-point: simplification of the gain in mix_pitch_an... |
tree | commitdiff |

*2008-03-13* |
Jean-Marc Valin | fixed-point: simplifying the arithmetic in alg_quant() |
tree | commitdiff |

*2008-03-12* |
Jean-Marc Valin | More accurate sqrt approximation using MULT16_16_Q15... |
tree | commitdiff |

*2008-03-12* |
Jean-Marc Valin | Increased accuracy of the reciprocal approximation... |
tree | commitdiff |

*2008-03-12* |
Jean-Marc Valin | Added info for properly exporting symbols |
tree | commitdiff |

*2008-03-12* |
Jean-Marc Valin | Changed definition of VARDECL to make it more flexible... |
tree | commitdiff |

*2008-03-12* |
Jean-Marc Valin | Using reciprocal approximation instead of full 32-bit... |
tree | commitdiff |

*2008-03-12* |
Jean-Marc Valin | Defined the celt_rcp() reciprocal approximation |
tree | commitdiff |

*2008-03-11* |
Jean-Marc Valin | __GNUC_PREREQ should be causing problems anymore. |
tree | commitdiff |

*2008-03-11* |
Jean-Marc Valin | Added macro definitions for the TI C5x family (untested) |
tree | commitdiff |

*2008-03-11* |
Jean-Marc Valin | Better heuristic for deciding when to use the 32-bit... |
tree | commitdiff |

*2008-03-10* |
Jean-Marc Valin | const correctness. celt_mode_create() now makes a copy... |
tree | commitdiff |

*2008-03-10* |
Jean-Marc Valin | oops, forgot to disable the main() |
tree | commitdiff |

*2008-03-10* |
Jean-Marc Valin | Changed some of the mode data from int to celt_int16_t... |
tree | commitdiff |

*2008-03-10* |
Jean-Marc Valin | No longer include redundent entries in the bit allocati... |
tree | commitdiff |

*2008-03-10* |
Jean-Marc Valin | Allocation cache can now be pre-computed as well. |
tree | commitdiff |

*2008-03-10* |
Jean-Marc Valin | Psychoacoustic decay coefficients can now be included... |
tree | commitdiff |

*2008-03-10* |
Jean-Marc Valin | Improvements to static modes |
tree | commitdiff |

*2008-03-10* |
Jean-Marc Valin | Moved the psycoacoustics data to the mode struct |
tree | commitdiff |

*2008-03-10* |
Jean-Marc Valin | Simplified spreading function so that only one set... |
tree | commitdiff |

*2008-03-10* |
Jean-Marc Valin | Initial support for static modes (defined STATIC_MODES) |
tree | commitdiff |

*2008-03-09* |
Jean-Marc Valin | fixed-point: done converting find_spectral_pitch() |
tree | commitdiff |

*2008-03-09* |
Jean-Marc Valin | fixed some build problems |
tree | commitdiff |

*2008-03-09* |
Jean-Marc Valin | testcelt no longer attempts to use ALLOC() |
tree | commitdiff |

*2008-03-09* |
Jean-Marc Valin | manual stack allocator should now work for sizeof(char)!=1 |
tree | commitdiff |

*2008-03-08* |
Jean-Marc Valin | fixed-point: stereo_mix() converted. |
tree | commitdiff |

*2008-03-08* |
Jean-Marc Valin | cleaning up some of the mode stuff |
tree | commitdiff |

*2008-03-07* |
Jean-Marc Valin | oops, forgot an int sampling rate somewhere |
tree | commitdiff |

*2008-03-07* |
Jean-Marc Valin | sampling rate has to be int32 |
tree | commitdiff |

*2008-03-07* |
Jean-Marc Valin | Adding code to dump the contents of a mode struct to... |
tree | commitdiff |

*2008-03-07* |
Jean-Marc Valin | fixed-point: defined HALF32() and used it for the forwa... |
tree | commitdiff |

*2008-03-07* |
Jean-Marc Valin | fixed-point: real FFT no longer needs float ops when... |
tree | commitdiff |

*2008-03-07* |
Jean-Marc Valin | Bit of cleaning up. No real code change (well, I hope... |
tree | commitdiff |

*2008-03-07* |
Jean-Marc Valin | fixed-point: removed the last float op in celt.c |
tree | commitdiff |

*2008-03-07* |
Jean-Marc Valin | fixed-point: finished intra_prediction(). No float... |
tree | commitdiff |

*2008-03-07* |
Jean-Marc Valin | fixed-point: converting intra-frame predictor (checkpoi... |
tree | commitdiff |

*2008-03-07* |
Jean-Marc Valin | fixed-point: intra_fold() converted |
tree | commitdiff |

*2008-03-07* |
Jean-Marc Valin | fixed-point: compute_band_energies() converted. Had... |
tree | commitdiff |

*2008-03-07* |
Jean-Marc Valin | fixed-point: converted normalise_bands (had to split... |
tree | commitdiff |

*2008-03-07* |
Jean-Marc Valin | Revert "compute_band_energies() merged with normalised_... |
tree | commitdiff |

*2008-03-07* |
Jean-Marc Valin | Changed the pulse spreading rotations so that the numbe... |
tree | commitdiff |

*2008-03-06* |
Jean-Marc Valin | fixed-point: done converting quant_bands() and unquant_... |
tree | commitdiff |

*2008-03-05* |
Jean-Marc Valin | Add option to #include "custom_support.h" |
tree | commitdiff |

*2008-03-05* |
Jean-Marc Valin | doing spreading function and gain quantisation in-place |
tree | commitdiff |

*2008-03-05* |
Jean-Marc Valin | fixed-point: part of stereo_mix() converted. |
tree | commitdiff |

*2008-03-05* |
Jean-Marc Valin | oops, unb0rked testcelt output |
tree | commitdiff |

*2008-03-05* |
Jean-Marc Valin | Defining RADIX_TWO_ONLY removes all butterflies, except... |
tree | commitdiff |

*2008-03-05* |
Jean-Marc Valin | fixed-point: changed find_spectral_pitch() to use singl... |
tree | commitdiff |

next |