pvq: use small LUT for integer sqrt((n+3)/2) sqrt((n+2)/2)
[daala.git] / src / pvq.c
2016-08-26 Tristan Matthewspvq: use small LUT for integer sqrt((n+3)/2) sqrt(...
2016-08-26 Tristan Matthewspvq: use od_beta_rcp in od_gain_compand
2016-08-25 Tristan Matthewspvq: use od_rcp_beta in od_compute_k
2016-08-24 Tristan Matthewspvq: use od_rcp_beta in od_compute_max_theta
2016-08-23 Tristan Matthewspvq: use int beta in od_pow
2016-08-19 Tristan Matthewspvq: keep beta int until used in computation
2016-08-19 Tristan Matthewspvq: fix documentation
2016-08-19 Tristan Matthewspvq: convert beta tables to int
2016-08-18 Tristan Matthewspvq: use od_log2 int approximation
2016-08-17 Tristan Matthewspvq: use approximation for od_rcp
2016-08-11 Tristan Matthewspvq: fix build for OD_FLOAT_PVQ
2016-08-11 Tristan Matthewspvq: fixed-point od_apply_householder
2016-08-11 Tristan Matthewspvq: convert compute_theta to fixed
2016-08-04 Tristan Matthewspvq: convert double constant to int
2016-08-01 Tristan Matthewspvq: convert od_log2 to int16_t input
2016-08-01 Tristan Matthewspvq: use integer approximation for exp2
2016-07-27 Tristan Matthewspvq: change shift to multiply for clarity
2016-07-27 Tristan Matthewspvq: convert compute_max_theta to fixed (aside from...
2016-07-21 Tristan Matthewspvq: use od_sqrt in od_pvq_compute_gain
2016-07-21 Tristan Matthewspvq: convert !noref branch of od_compute_k to integer
2016-07-20 Tristan Matthewspvq: convert 1st part of compute_k to integer (aside...
2016-07-20 Tristan Matthewspvq: fix return type for od_sqrt
2016-07-18 Tristan Matthewspvq: avoid double division in od_gain_compand
2016-07-15 Nathan E. EggeRemove compiler warnings with --enable-float-pvq.
2016-07-02 Tristan Matthewspvq: use od_pow() when beta != 1 and beta != 1.5
2016-07-01 Tristan Matthewspvq: use rounded exp2 and log2 when raising g to 1...
2016-06-01 Tristan Matthewspvq: use integer approximation of sqrt
2016-06-01 Tristan Matthewsod_rsqrt: s only depends on log2(x) and inshift
2016-06-01 Tristan Matthewsod_sqrt: s only depends on log2(x) and the input shift
2016-04-07 Tristan Matthewspvq: use int16_t for rsqrt calculations
2016-04-02 Jean-Marc ValinNew k-tokenizer based on recursively splitting vectors
2016-03-31 Tristan Matthewspvq: use float rounded version of sqrt(cg)
2016-03-31 Tristan Matthewspvq: update parameter doc for od_pvq_compute_gain
2016-03-25 Tristan Matthewspvq: add --enable-float-pvq option to configure
2016-03-14 Tristan Matthewspvq: convert non activity masking case to integer
2016-03-12 Tristan Matthewspvq: use integer approximation of 1/sqrt(yy)
2016-03-12 Tristan Matthewspvq: use float rounded version of 1/sqrt(yy)
2016-02-18 Tristan Matthewspvq: use integer approximations of cos and sin
2016-02-15 Tristan Matthewspvq: change OD_TRIG_SCALE to 32768, saturate to 32767
2016-02-11 Jean-Marc ValinPVQ: od_pvq_sin() and od_pvq_cos() now output an integer
2016-02-11 Jean-Marc ValinPVQ: Adds some missing casts
2016-02-11 Jean-Marc ValinPVQ: Change PVQ code to use int32_t as companded gain...
2016-02-10 Jean-Marc ValinPVQ: Minor type fix (no change in behaviour)
2016-02-10 Jean-Marc ValinPVQ: change the scaling of companded gain variables...
2016-02-10 Jean-Marc ValinPVQ: Quantize companded gain
2016-02-10 Jean-Marc ValinPVQ: Theta values converted to integer
2016-02-10 Jean-Marc ValinPVQ: Changing the scale of theta to that of a 16-bit...
2016-02-10 Jean-Marc ValinPVQ: Quantizing input and output of sin and cos
2016-02-09 Jean-Marc ValinPVQ: Extra comments about the fixed-point code
2016-02-08 Jean-Marc ValinPVQ: Changing accumulations to 32-bit int in od_apply_h...
2016-02-08 Jean-Marc ValinPVQ: Removes all double-precision calculations from...
2016-02-08 Jean-Marc ValinPVQ: Removes double-precision Householder reflection
2016-02-08 Jean-Marc ValinPVQ: Using 16-bit Householder in od_pvq_synthesis_partial()
2016-02-08 Jean-Marc ValinPVQ: QM scaling in od_pvq_synthesis_partial() in fixed...
2016-02-08 Jean-Marc ValinPVQ: Converting ypulse[] scaling to integer in od_pvq_s...
2016-02-08 Jean-Marc ValinPVQ: Linear gains changed to integer
2016-02-08 Jean-Marc ValinPVQ: Quantized gain
2016-02-08 Jean-Marc ValinPVQ: Perform PVQ search on 16-bit input
2016-02-08 Jean-Marc ValinPVQ: Use x16 as Householder input
2016-02-08 Jean-Marc ValinPVQ: Quantizing the Householder reflection vector
2016-02-08 Jean-Marc ValinPVQ: Quantizing the input and reference to 16 bits...
2016-01-21 Tristan Matthewspvq: remove logging
2015-12-15 Tristan Matthewspvq: improve packing
2015-12-13 Tristan Matthewspvq: fix integer overflow
2015-12-11 Tristan Matthewspvq: remove od_apply_qm
2015-11-01 Jean-Marc Valinmove PVQ state to pvq.h
2015-10-02 Tristan Matthewspvq: drop unused define
2015-08-20 Jean-Marc ValinAdds the option of using the Thor deblocking filter...
2015-07-28 Nathan E. EggeFix od_qm_get_index function.
2015-07-27 Jean-Marc ValinUsing 4-point lapping for all block sizes
2015-07-15 Tristan Matthewsencode: move QMs to where they are used
2015-07-09 Ian KronquistRemove dependency on libogg for core daala
2015-07-07 Tristan Matthewspvq: improve some QM names
2015-07-05 Tristan Matthewspvq: add comment explaining 1 vs. 1.5 for beta
2015-07-04 Nathan E. EggeAdd magnitude of 64-point basis functions.
2015-07-01 Tristan Matthewspvq: update doc to match code
2015-06-02 Jean-Marc ValinChanging block size variables from ln to bs
2015-05-21 Tristan MatthewsSignal in bitstream which QM is in use
2015-05-12 Tristan MatthewsSignal in bitstream if activity masking is in use
2015-03-23 Jean-Marc ValinMore fixes to address Tim's comments
2015-03-23 Jean-Marc ValinAdds quantization matrices
2015-03-23 Jean-Marc ValinUpdating quantization matrices to take into account...
2015-03-12 Nathan E. EggeAdding a 32-point lapping filter.
2015-02-10 Tristan Matthewspvq: use od_ prefix for non-static functions
2015-02-02 Jean-Marc ValinInterpolating quantization matrices across rates
2014-12-23 Jean-Marc ValinUse the same quantizers with and without activity maski...
2014-12-17 Jean-Marc ValinSignaling the quantization matrix in the bitstream
2014-12-02 Jean-Marc Valinfixes some comments
2014-12-02 Jean-Marc ValinMoves pvq.c functions to encoder/decoder-specific files
2014-12-02 Jean-Marc ValinMore simplifications to pvq_synthesis_partial().
2014-12-02 Jean-Marc ValinMore PVQ cleanup.
2014-11-23 Jean-Marc ValinReducing the use of pow() by optmizing the beta=1 and...
2014-11-23 Jean-Marc ValinUse special cases for "copy zero" and "copy reference"
2014-11-18 Jean-Marc ValinRDO on last two pulses
2014-11-18 Jean-Marc ValinAdding some RDO cost to using the H/V predictor
2014-11-14 Jean-Marc ValinMerge branch 'remove_intra2'
2014-11-13 Jean-Marc Valinfixes for issue 512
2014-11-13 Jean-Marc ValinChange ordering of CfL flip flag so that it's encoder...
2014-11-11 randomguyAdd casts to avoid precision loss warnings under MSVC.
2014-11-10 Nathan E. EggeUse OD_LOG2 macro instead of log2().