Welcome to mirror list, hosted at ThFree Co, Russian Federation.

gitlab.xiph.org/xiph/opus.git - Unnamed repository; edit this file 'description' to name the repository.
summaryrefslogtreecommitdiff
BranchCommit messageAuthorAge
1.0.1Bump version to 1.0.1Jean-Marc Valin12 years
1.0.xAdds some checks to the surround APIJean-Marc Valin11 years
1.1.1bump version for upcoming 1.1.1Jean-Marc Valin8 years
1.1.2update releases.sha2Jean-Marc Valin8 years
1.1.xBump version numbersJean-Marc Valin7 years
analysis_fix1Improve silence handlingJean-Marc Valin5 years
analysis_fix2Reset nb_no_activity_frames when analysis DTX isn't usedJean-Marc Valin5 years
burg_optimizationfaster Burg implementationKoen Vos8 years
cng_fixFixes the transient detector on silenceJean-Marc Valin8 years
cng_fix2Always use the max of the decaying PLC energy and the background noiseJean-Marc Valin8 years
draftMerge branch 'draft-willson' into draftRalph Giles4 years
draft_wipadd security considerationJean-Marc Valin7 years
dred_bitrateSome general SILK CBR tuningJean-Marc Valin10 months
dred_bitrate2Properly account for SILK bits in CELT CBR codeJean-Marc Valin10 months
dred_bitrate3Implement allocation of the DRED bitsJean-Marc Valin10 months
exp-fwgan-plcUse FWGAN instead of LPCNet in PLCJean-Marc Valin9 months
exp-fwgan1Skeleton for FWGAN codeJean-Marc Valin9 months
exp-fwgan10Add FWGAN to lpcnet_demoJean-Marc Valin9 months
exp-fwgan2Add pre-emphasis filter before LPC filterJean-Marc Valin9 months
exp-fwgan8Bring back support for non-multiples of 8Jean-Marc Valin9 months
exp-highpassRemoves code that became useless with the Opus-level highpassJean-Marc Valin13 years
exp-lace-variable-lookaheadbugfixesJan Buethe6 months
exp-neural-silk-enhancementadded 16kHz output from silk decoderJan Buethe12 months
exp_24k_analysis1cleanupJean-Marc Valin7 years
exp_24k_analysis2Fix analysis for 2.5 ms framesJean-Marc Valin7 years
exp_24k_analysis3Adds support for running the analysis on 16 kHz inputJean-Marc Valin7 years
exp_24k_analysis4We no longer need 200 frames to remember 2 secondsJean-Marc Valin7 years
exp_24k_analysis5Cleaning up mlp_train codeJean-Marc Valin7 years
exp_analysisMakes the speech/music probability estimation mode conservativeJean-Marc Valin11 years
exp_analysis6OPUS_{GET,SET}_LSB_DEPTH for multichannel.Gregory Maxwell12 years
exp_analysis7Tuning the speech/music probability modelJean-Marc Valin12 years
exp_analysis8Makes variable framesize less aggressive at lower ratesJean-Marc Valin11 years
exp_analysis9Making multistream variable duration work for both the float and int APIJean-Marc Valin11 years
exp_arm_dotprod1Add dotprod code for dnn rtcdJean-Marc Valin6 months
exp_arm_dotprod2oopsJean-Marc Valin5 months
exp_avx2_nsq1oopsJean-Marc Valin6 months
exp_avx2_nsq2Oops, fix commentJean-Marc Valin6 months
exp_bandwidthFixes an integer overflow for lsb_depth=24Jean-Marc Valin11 years
exp_bandwidth_fix3Fix decision thresholds for larger frame sizesJean-Marc Valin6 years
exp_bandwidth_fix4Switch from narrowband to wideband at 9 kb/s, don't use mediumbandJean-Marc Valin6 years
exp_bitstreamMake it possible to ignore inverted phase stereo for downmix purposesJean-Marc Valin8 years
exp_bitstream2Only enable new hybrid folding behaviour with ENABLE_UPDATE_DRAFTJean-Marc Valin8 years
exp_bitstream3Make it possible to ignore inverted phase stereo for downmix purposesJean-Marc Valin8 years
exp_bitstream5Make it possible to ignore inverted phase stereo for downmix purposesJean-Marc Valin8 years
exp_bitstream6Make it possible to ignore inverted phase stereo for downmix purposesJean-Marc Valin8 years
exp_bitstream7Make it possible to ignore inverted phase stereo for downmix purposesJean-Marc Valin8 years
exp_cbr_silkAdds padding code to fill up SILK frames to CBRJean-Marc Valin13 years
exp_celt_plc_constraint1Prevent overshoots from CELT PLC with predictionJean-Marc Valin5 months
exp_celt_plc_fix1Fix silly bug in CELT Deep PLCJean-Marc Valin6 months
exp_crc32wipJean-Marc Valin6 years
exp_delayed_decisionUpdates opus_demoJean-Marc Valin11 years
exp_densenet3Model updateJean-Marc Valin8 months
exp_densenet3xIncrease size, init convolutions with zerosJean-Marc Valin7 months
exp_detectionTraining now stops when stuck in a minimumJean-Marc Valin13 years
exp_dnn_rtcd2Adding RTCD for compute_activation()Jean-Marc Valin6 months
exp_dnn_rtcd3Adding RTCD for compute_activation()Jean-Marc Valin6 months
exp_dnn_rtcd4Adding RTCD for compute_activation()Jean-Marc Valin6 months
exp_dnn_rtcd5Addressing Tim's commentsJean-Marc Valin6 months
exp_dnn_rtcd6Adding RTCD for compute_activation()Jean-Marc Valin6 months
exp_dred_icasspLPCNet updateJean-Marc Valin15 months
exp_dred_refactor1Add blob loading for DRED encoder and decoderJean-Marc Valin11 months
exp_dred_tuning1Initial DRED tuningJean-Marc Valin4 months
exp_fargan33added FARGAN adversarial training scriptJan Buethe8 months
exp_fargan38fix gammaJean-Marc Valin8 months
exp_fargan43SaSignal domain, with pitch gatingJean-Marc Valin7 months
exp_fargan43gmore simplificationsJean-Marc Valin7 months
exp_fargan43g2more simplificationsJean-Marc Valin7 months
exp_fargan44constraining the pitchJean-Marc Valin7 months
exp_fargan_plcdebug codeJean-Marc Valin7 months
exp_fargan_plc3update weight blob codeJean-Marc Valin7 months
exp_fargan_shrink1Smaller version of farganJean-Marc Valin6 months
exp_fargan_shrink2Smaller version of farganJean-Marc Valin6 months
exp_fargancode1oops, add dumping scriptJean-Marc Valin7 months
exp_fec_tuneTune SILK rate for hybrid with FEC enabledJean-Marc Valin8 years
exp_gain_lock2CBR: reset gains to their original values where we're busting the budgetJean-Marc Valin8 years
exp_gain_lock4CBR: set gains to their previons values where we're busting the budgetJean-Marc Valin8 years
exp_gain_lock5CBR: set gains to their previons values where we're busting the budgetJean-Marc Valin8 years
exp_google_opt1bOptimize silk_NSQ_del_dec() for ARM NEONLinfeng Zhang7 years
exp_google_opt1cOptimize silk_NSQ_del_dec() for ARM NEONLinfeng Zhang7 years
exp_google_opt1eOptimize silk_NSQ_del_dec() for ARM NEONLinfeng Zhang7 years
exp_hardening2MUST_SUCCEED() fixesJean-Marc Valin6 years
exp_hybrid_lbr1Fixes integer overflow in SILK VAD for 10-ms framesJean-Marc Valin6 years
exp_hybrid_stereoBetter rate allocation for stereo SILK in hybrid modeJean-Marc Valin7 years
exp_hybrid_tuneFixes foldingJean-Marc Valin8 years
exp_hybrid_tune2Don't use too much temporal resolution on hybrid frames at low rateJean-Marc Valin8 years
exp_ignoreAdd mising cmake file from make distJean-Marc Valin5 years
exp_kelly1Printing classifier probsJean-Marc Valin7 years
exp_koen4Fixes the high-band hybrid gain from the previous commitJean-Marc Valin12 years
exp_lbr_tuneAdds some smoothing to the energy quantizationJean-Marc Valin8 years
exp_lbr_tune1Oops, fixes max_decay for LFE in fixed-pointJean-Marc Valin8 years
exp_lbr_tune10Adds some smoothing to the energy quantizationJean-Marc Valin8 years
exp_lbr_tune11Adds some smoothing to the energy quantizationJean-Marc Valin8 years
exp_lbr_tune12Include config.h in mlp.cJean-Marc Valin8 years
exp_lbr_tune13Include config.h in mlp.cJean-Marc Valin8 years
exp_lbr_tune14Include config.h in mlp.cJean-Marc Valin8 years
exp_lbr_tune2Forces trim to 5 on hybrid modeJean-Marc Valin8 years
exp_lbr_tune3Fixes range coder mismatch introduced in 0eb652a38c4Jean-Marc Valin8 years
exp_lbr_tune4faster Burg implementationKoen Vos8 years
exp_lbr_tune5faster Burg implementationKoen Vos8 years
exp_lbr_tune6comment about reduced codelength component in LTP quantizer error metricKoen Vos8 years
exp_lbr_tune7Oops, edited the wrong comment in the previous commitJean-Marc Valin8 years
exp_lbr_tune8Taking loss rate into account in compute_equiv_rate()Jean-Marc Valin8 years
exp_lbr_tune9Adds some smoothing to the energy quantizationJean-Marc Valin8 years
exp_leakage4First attempt at modelling leakageJean-Marc Valin7 years
exp_leakage7Adding leakage modelling to boost bandsJean-Marc Valin7 years
exp_lowbitrate1Fix whitespaceJean-Marc Valin5 months
exp_lowbitrate2CBR tuningJean-Marc Valin5 months
exp_mips_altmips C equivalent: Using simple rounding rather than the original biasJean-Marc Valin10 years
exp_mips_alt_jun03C equivalent of the MIPS codeJean-Marc Valin10 years
exp_mips_optWhitespace fixesJean-Marc Valin10 years
exp_msc_build1try to use vec_avx.h for MSVC buildsMichael Klingbeil5 months
exp_multiframe1Refactor multi-frame encoding to be non-recursiveJean-Marc Valin5 months
exp_multiframe_cleanup25 does not equal 8, even for large values of 5Jean-Marc Valin5 months
exp_multiframe_cleanup3divide max payload tooJean-Marc Valin5 months
exp_multiframe_cleanup6-tuneCleanup previous commitsJean-Marc Valin4 months
exp_neural_fec2More general Laplace encoderJean-Marc Valin20 months
exp_neural_fec2bUsing LPCNet version with FECJean-Marc Valin17 months
exp_neural_fec2cCode for extracting DRED from packetJean-Marc Valin17 months
exp_neural_fec3LPCNet updateJean-Marc Valin12 months
exp_neural_fec4Github actions for neural fecMarcus Asteborg11 months
exp_new_classifier1Retrain RNN classifier weights to include reverberated speechJean-Marc Valin6 years
exp_new_tf2btuningJean-Marc Valin6 years
exp_ng_lfsadd h5 models to git lfsMichael Klingbeil10 months
exp_overflow2More headroom in fixed-point silk_burg_modified() to avoid an overflowJean-Marc Valin8 years
exp_overflow4SILK overflow fixesJean-Marc Valin8 years
exp_overflow6Fixes cap on gain in denormalise_bands()Jean-Marc Valin8 years
exp_pitchdnn1changed checkpoint formatJan Buethe7 months
exp_pitchdnn4model updateJean-Marc Valin7 months
exp_plc_fix1Improving PLCJean-Marc Valin4 months
exp_plc_fix2Remove run-time code for old TF2 modelsJean-Marc Valin4 months
exp_plc_fix5Remove run-time code for old TF2 modelsJean-Marc Valin4 months
exp_plc_hack4enable neural PLC by defaultJean-Marc Valin2 years
exp_plc_refactor1Support OPUS_SET_COMPLEXITY() on decoder sideJean-Marc Valin7 months
exp_pvq_opt1disable PVQ SSE code for fixed-pointJean-Marc Valin8 years
exp_ratecontrol1Improving rate control for low bitrateJean-Marc Valin6 years
exp_rdovae_state2updated modelJean-Marc Valin8 months
exp_rdovae_stats1Remove useless latent/state dimensionsJean-Marc Valin6 months
exp_rdovae_stats2Model updateJean-Marc Valin6 months
exp_refactoringmore bands.c cleaning upJean-Marc Valin11 years
exp_rnn1Using lookahead (when possible) to make the mode decision on the first frameJean-Marc Valin7 years
exp_rnn3Add RNN for VAD and speech/music classificationJean-Marc Valin7 years
exp_silk_cbrWe should now be able to be less conservative when using CBRJean-Marc Valin8 years
exp_silk_cbr2Using more fine-grained complexity settings for SILKJean-Marc Valin8 years
exp_silk_feature_dumpremove decoder delay for widebandJean-Marc Valin5 years
exp_silk_jm2Attenuate SILK PLC gain only for unvoiced speechFelicia Lim8 years
exp_sim_loss2Add simulated loss to opus_demoJean-Marc Valin5 months
exp_spreading_mask1Add a simple masking model to the spreading decisionJean-Marc Valin6 years
exp_stagingfixing unit testsJean-Marc Valin7 years
exp_stereo3No need to renormalizeJean-Marc Valin8 years
exp_stereo5Partially collapsing the stereo image when channels have different energyJean-Marc Valin8 years
exp_stereo6More LyX textJean-Marc Valin8 years
exp_stereo_vbr1experimental new stereo VBR codeJean-Marc Valin8 years
exp_stereo_vbr3wipJean-Marc Valin8 years
exp_surround1Avoids unnecessary copying in opus_encode for delayed decisionJean-Marc Valin11 years
exp_temporalSome tuning of temporal VBR.Jean-Marc Valin11 years
exp_temporal1Some tuning of temporal VBR.Jean-Marc Valin11 years
exp_theta10We need to copy the output folding, not the inputJean-Marc Valin8 years
exp_theta11Don't do theta RDO on intensity-stereo-coded bandsJean-Marc Valin8 years
exp_theta4distortion onlyJean-Marc Valin8 years
exp_theta6wipJean-Marc Valin8 years
exp_theta7mask is unsignedJean-Marc Valin8 years
exp_transient_fold1Using more folding on transients and less on non-transientJean-Marc Valin8 years
exp_ubsan1Ensuring we can see where crashes occurJean-Marc Valin22 months
exp_warnings1Fix C90-related warningsJean-Marc Valin22 months
exp_waveform_matchingUse new waveform matching option on ambisonicsJean-Marc Valin5 years
exp_waveform_matching2Use new waveform matching option on ambisonicsJean-Marc Valin5 years
exp_waveform_matching3Use new waveform matching option on ambisonicsJean-Marc Valin5 years
exp_wipoopsJean-Marc Valin11 years
exp_wip4Reverts the transient_analysis() changes from 4e8b9905Jean-Marc Valin12 years
exp_wip5Reverts the transient_analysis() changes from 4e8b9905Jean-Marc Valin12 years
exp_wip6Add exp_wip5 tag to version string.Gregory Maxwell12 years
gain_sumAdds back gain sum checkJean-Marc Valin8 years
gitlab-cifixup gitlab-ci: job extends typoRalph Giles4 years
hybrid_vbrusing "hybrid" flag instead of start != 0Jean-Marc Valin8 years
hybrid_vbr3using "hybrid" flag instead of start != 0Jean-Marc Valin8 years
icassp2024Add pitch_demoJean-Marc Valin4 months
ietf-draftFixes a minor CELT->SILK switching glitch in the decoderJean-Marc Valin11 years
jon_patchesMake it possible to pass in paths to opus_demo and opus_compare.Jonathan Lennox9 years
krishna-neural-pitchneural pitch updateKrishna Subramani7 months
masterMatch silenced overflow checks in the sse4.1 versionFelicia Lim6 months
merge_attemptMerge branch 'exp_analysis7' into masterJean-Marc Valin12 years
msvcSquashed commit of the following:Jean-Marc Valin11 years
ne10_patchesAdd entcode.c to dump_modes build.Timothy B. Terriberry9 years
neural_fec3LPCNet updateJean-Marc Valin15 months
neural_plcUpdate build instructionsJean-Marc Valin24 months
opt_floatUsing celt_inner_prod() for input sanity check.Jean-Marc Valin10 years
opus-ngfixes in osce python codeJan Buethe4 months
opus-ng-2bcupdated to new neural pitch datasetJan Buethe7 months
opus-ng-fix-allocuse opus_(re)alloc and opus_free for dnn and DRED related functionsMichael Klingbeil5 months
opus-ng-lace-integrationnolace debugged and activeJan Buethe5 months
opus-ng-lace-integration2added missing copyright headerJan Buethe5 months
opus-ng-lace-integration2-model-tuningMerge remote-tracking branch 'origin/opus-ng-lace-integration2' into opus-ng-...Jan Buethe5 months
opus-ng-lace-integration2-quantizationupdated osce modelsJan Buethe4 months
opus-ng-lace-integration4float constantsJean-Marc Valin5 months
opus-ng-lace-integration5Merge LACE/NoLACE under OSCE frameworkJan Buethe5 months
opus-ng-linearswitched to FIFO ordering in weight-exchange and switched to direct include i...Jan Buethe9 months
opus-ng-linear3oopsJean-Marc Valin9 months
opus-ng-linear4Make RDOVAE encoder use LinearLayer directlyJean-Marc Valin9 months
opus-ng-neonwarningfixneon related alignment requirement warning fixJan Buethe10 months
opus-ng-osce-model-tuningclean-upJan Buethe4 months
opus-ng-osce-models-v2Updated LACE and NoLACE models to version 2Jan Buethe4 months
silk_LTP_corr_smoothingforward-backward smoothing of LTP correlations before quantizingKoen Vos8 years
silk_lsf_quant_with_hysteresisbias in LSF quantization towards previously quantized LSFsKoen Vos8 years
silk_pitch_fixMakes pitch gain control less aggressiveJean-Marc Valin10 years
silk_tuningsfix for divide by zeroFelicia Lim8 years
silk_tunings2set back to floating point by defaultKoen Vos8 years
silk_tunings3Clean up: alignment of commentsKoen Vos8 years
silk_tunings4Clean up: alignment of commentsKoen Vos8 years
silk_tunings6fix bug in previous commitKoen Vos8 years
stagingWork around VS2015 internal compiler errorMark Harris6 years
stereo_masking1biasing quantizationJean-Marc Valin8 years
surroundRevisit surround rate allocationJean-Marc Valin11 years
update_draftPrevents ridiculously large gains from causing inf/NaNs in float decoderJean-Marc Valin8 years
 
TagDownloadAuthorAge
v1.4commit 82ac57d9f1...Jean-Marc Valin13 months
v1.3.1commit e85ed7726d...Jean-Marc Valin5 years
v1.3commit 83d5155f15...Jean-Marc Valin6 years
v1.3-rc2commit 4a643d98c3...Jean-Marc Valin6 years
v1.3-rccommit 693098f6ee...Jean-Marc Valin6 years
v1.3-betacommit ff6532c3be...Jean-Marc Valin6 years
v1.2.1commit defbc370ec...Jean-Marc Valin7 years
v1.2commit 5bb3cbc34c...Jean-Marc Valin7 years
v1.2-rc1commit 44d653003b...Jean-Marc Valin7 years
v1.2-betacommit 3609a2218e...Jean-Marc Valin7 years
v1.1.5commit d3977edab4...Jean-Marc Valin7 years
v1.1.4commit 901c24328d...Jean-Marc Valin7 years
v1.2-alpha2commit aef475641a...Jean-Marc Valin7 years
v1.2-alphacommit 92282c1ae4...Jean-Marc Valin8 years
v1.1.3commit aa32042a50...Jean-Marc Valin8 years
v1.1.2commit 655cc54c56...Jean-Marc Valin8 years
v1.1.1commit 1f22ae379b...Jean-Marc Valin8 years
v1.1.1-rccommit 699b643042...Jean-Marc Valin9 years
draft-ietf-codec-oggopus-06commit 8ab1968e13...Ralph Giles10 years
draft-ietf-codec-oggopus-05commit 3731a57a1b...Ralph Giles10 years
v1.1.1-betacommit fd562eefa2...Jean-Marc Valin10 years
draft-ietf-codec-oggopus-04commit fd2992aaab...Ralph Giles10 years
draft-ietf-codec-oggopus-03commit 4e8acd5ca3...Ralph Giles10 years
draft-ietf-codec-oggopus-01commit 25ffd5cd91...Ralph Giles10 years
draft-ietf-codec-oggopus-02commit cf33d32440...Ralph Giles10 years
v1.1commit 65471dd567...Jean-Marc Valin10 years
v1.1-rc3commit 17dcb27026...Jean-Marc Valin10 years
v1.1-rc2commit c16a96b458...Jean-Marc Valin10 years
v1.1-rccommit d6eb9c468f...Jean-Marc Valin10 years
v1.1-betacommit ea95c533c8...Jean-Marc Valin11 years
v1.0.3commit fc66793e1f...Jean-Marc Valin11 years
v1.1-alphacommit d683c76b88...Jean-Marc Valin11 years
v1.0.2commit 6f2ce267af...Jean-Marc Valin11 years
v1.0.1commit e7f668b3d4...Jean-Marc Valin12 years
v1.0.1-rc3commit 12190653b2...Jean-Marc Valin12 years
v1.0.0commit a69a6f8bc4...Jean-Marc Valin12 years
v1.0.1-rc2commit 9a35da4659...Jean-Marc Valin12 years
v1.0.1-rccommit 4ee682f52a...Jean-Marc Valin12 years
v1.0.0-rccommit 028dacd407...Jean-Marc Valin12 years
draft-15commit eec7590a58...Jean-Marc Valin12 years
v0.9.14commit ad49daa473...Jean-Marc Valin12 years
draft-14commit ad49daa473...Jean-Marc Valin12 years
v0.9.11commit 88ddeee581...Jean-Marc Valin12 years
draft-13commit 88ddeee581...Jean-Marc Valin12 years
draft-12commit b72de4afda...Jean-Marc Valin12 years
v0.9.10commit b72de4afda...Jean-Marc Valin12 years
v0.9.9commit 17c5966045...Jean-Marc Valin12 years
draft-10commit a9d342e61e...Jean-Marc Valin13 years
v0.9.8commit a9d342e61e...Jean-Marc Valin13 years
draft-09commit 38c6a15092...Jean-Marc Valin13 years
v0.9.7commit 38c6a15092...Jean-Marc Valin13 years
draft-08commit 8a23fe977f...Jean-Marc Valin13 years
v0.9.6commit 8a23fe977f...Jean-Marc Valin13 years
draft-07commit 3596fedb65...Jean-Marc Valin13 years
v0.9.5commit 3596fedb65...Jean-Marc Valin13 years
v0.9.4commit 0ed6577500...Jean-Marc Valin13 years
generated by cgit v1.2.3 (git 2.25.1) at 2024-05-09 13:35:11 +0300