Skip to content
Snippets Groups Projects
Select Git revision
  • 1.0.1
  • 1.0.x
  • 1.1.1
  • 1.1.2
  • 1.1.x
  • analysis_fix1
  • analysis_fix2
  • burg_optimization
  • cng_fix
  • cng_fix2
  • draft
  • draft_wip
  • exp-highpass
  • exp_24k_analysis1
  • exp_24k_analysis2
  • exp_24k_analysis3
  • exp_24k_analysis4
  • exp_24k_analysis5
  • exp_analysis
  • exp_analysis6
  • v1.3.1
  • v1.3
  • v1.3-rc2
  • v1.3-rc
  • v1.3-beta
  • v1.2.1
  • v1.2
  • v1.2-rc1
  • v1.2-beta
  • v1.1.5
  • v1.1.4
  • v1.2-alpha2
  • v1.2-alpha
  • v1.1.3
  • v1.1.2
  • v1.1.1
  • v1.1.1-rc
  • draft-ietf-codec-oggopus-06
  • draft-ietf-codec-oggopus-05
  • v1.1.1-beta
40 results
You can move around the graph by using the arrow keys.
Created with Raphaël 2.2.07Jan654331Dec29282723222120171614111098543230Nov29272826252422212021201918191817161513121198428Oct23211714131211129Sep2816141312111096531Aug302928262112430Jul2523151312117212129Jun27262517161513109765427May26252425242322212019201918171613121110119876523Apr1915518Mar1211875128Feb2220191817161076516Jan111093222Dec2120191817141312117543130Nov2927262322211915141211987653231Oct3029272423211918Preventing unnecessary stack use when using a large decode bufferFixes SMALL_FOOTPRINT for floatAnother aliasing hack in the SILK PLC with SMALL_FOOTPRINTDon't allocate pulses on the stack when calling the SILK PLC.Moves CELT PLC pitch search to a separate function to reduce peak stackDelaying allocation of the SILK temporary output buffer to reduce peak stackAdds SMALL_FOOTPRINT hack to the CELT PLC tooMoves deemphasis() call out of celt_decode_lost() to reduce peak stackStore decoded SILK pulses as 16-bit vectorCleaning up leftovers of "freq" in celt_decode_with_ec()Reduces the decoder stack use by removing the pcm_silk buffer in fixed-pointMoves the remains of compute_inv_mdcts() to celt_synthesis()Hack that makes the SMALL_FOOTPRINT CELT decoder use only 4.25 kB of stack.Reduces decoder stack usage by only storing one channel of denormalized MDCTSome cleaning up of the synthesis code.Making exp_rotation1() use MAC16_16(), which saves a few cycles on ARMSilences unused parameter warningMoving the radix-2 to expose trivial twiddle factorsImproving the accuracy of the fixed-point radix-3 and radix-5Minor cleanup -- nothing to see hereFixed-point: slight accuracy improvement in the comb filterRemove a SAVE_STACK that was pasted accidentally in the previous commitUnifying scaling of fixed-point and float FFTFixes C89 issueGetting rid of some negationsSlightly improving the accuracy of the fixed-point MDCT downscaleCommit 99968ab was causing us to allocate too much stack in the MDCTRemoves the separate 1/8N rotation in the (I)MDCT and unmerges the MDCT sizesInverse MDCT no longer requires any scratch spaceReverse the ordering of the FFT stage to optimize a degenerate radix-4 case.Merges the FFT scaling with the MDCT pre-rotateGetting rid of the inverse FFT entirelyMoving bitrev step to forward MDCT tooApplying the forward FFT gain up-front for fixed-point tooMoves the bitrev step to the IMDCT pre-rotationFixes a warning about "conversion from '__int64' to 'int'" on MSVCconfigure.ac: fix bashism in ARM optimization handlingfixed-point: slight (but free) accuracy improvement in compute_band_energies()fixed-point: adds rounding to some shifts to eliminate biasRemove useless code in alloc_trim_analysis()