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
AgeCommit message (Collapse)Author
2016-06-03CBR: reset gains to their original values where we're busting the budgetexp_gain_lock2Jean-Marc Valin
2016-06-03Set pulses to zero if we can't meet rate targetJean-Marc Valin
2016-06-03Now that we're preventing huge gains, we no longer need to pretend there'sJean-Marc Valin
no dither
2016-06-03CBR: lock the gain on a subframe when the number of pulses stops going downJean-Marc Valin
2016-06-03CBR: Using the same gain increase for fixed-point as for floatJean-Marc Valin
2016-06-03Oops, edited the wrong comment in the previous commitexp_lbr_tune7Jean-Marc Valin
2016-06-03Re-tuning compute_equiv_rate()Jean-Marc Valin
2016-06-03Getting rid of the SILK penalty for hybrid CBR.Jean-Marc Valin
That was likely a mistake from a misunderstanding on the SILK bit allocation
2016-06-03Improving CBR for hybrid modeJean-Marc Valin
Hybrid CBR now simply forces the SILK CBR rate to the "target" value. Also, we're getting rid of the -2 kb/s offset for CBR, which appeared to be harmful. In the case of hybrid constrained VBR, the cap is computed in the same way as the target.
2016-06-02Making the encoder more aggressive at using wider bandwidthsJean-Marc Valin
2016-06-02comment about reduced codelength component in LTP quantizer error metricKoen Vos
2016-06-02avoid double if statement for finding best quantized LTP coefsKoen Vos
2016-06-02increase NSQ decision delay from 32 to 40Koen Vos
2016-06-02reduced noise floor in noise shaping analysis by 2 dBKoen Vos
2016-06-02increase order of noise shaping filterKoen Vos
2016-06-02more conservative scaling of LTP corrs; remove assert that checks for ↵Koen Vos
negative residual energy; discard CB entries leading to such negative energies
2016-06-02Clean up: alignment of commentsKoen Vos
2016-06-02Clean up: replace tabs by spacesKoen Vos
2016-06-02removed prefilterKoen Vos
The NSQ SSE optimizations are disabled for now because they need to be updated
2016-06-02slight clean upKoen Vos
2016-06-02undo some loop unrollingKoen Vos
2016-06-02pitch analysis (in FIX) now scales the input down at the start, instead of ↵Koen Vos
at every stage
2016-06-02minor clean up of pointer offsetsKoen Vos
2016-06-02simplified computation of LTP coefsKoen Vos
2016-06-02Attenuate SILK PLC gain only for unvoiced speechFelicia Lim
2016-06-02We should now be able to be less conservative when using CBRJean-Marc Valin
2016-06-02Makes the encoder more aggressive about meeting the rate targetJean-Marc Valin
2016-06-02Using more fine-grained complexity settings for SILKJean-Marc Valin
Only complexity 0-1 now uses the plain non-delayed-decision NSQ since its quality is pretty bad. Previously 0-3 were using that plain NSQ.
2016-06-02Fixing bandwidth selection hysteresisJean-Marc Valin
Previously, the bandwidth detection could (e.g.) change the bandwidth from fullband to superwideband, and the hysteresis would then cause bandwidth to be stuck in superwideband.
2016-06-02Boosting the high-band bitrate on transients for hybrid modeJean-Marc Valin
2016-06-02fix potential overflowKoen Vos
2016-06-02Add experimental support for ambisonic encodingMichael Graczyk
The implementation currently only codes each channel independently with no special allocation rules. Signed-off-by: Jean-Marc Valin <jmvalin@jmvalin.ca>
2016-06-02Fixes signed integer overlof in silk_ADD_POS_SAT32()Jean-Marc Valin
Removes unused 64-bit version
2016-06-02Turn on DTX only when the noise is sufficiently quietFelicia Lim
Computes a pseudo SNR as the ratio between the running peak signal energy to instantaneous noise energy. DTX is turned on only if this is higher than a pre-determined threshold. Signed-off-by: Jean-Marc Valin <jmvalin@jmvalin.ca>
2016-06-02New DTX that works in all modes (SILK/CELT/HYBRID)Felicia Lim
A frame is marked as valid for DTX if it contains noise or only digital silence. As before, there is an overhang period of 200 ms and a maximum consecutive DTX period of 400 ms. If the new DTX cannot be used because of the complexity setting and sampling frequency chosen, the SILK DTX will be used instead. Signed-off-by: Jean-Marc Valin <jmvalin@jmvalin.ca>
2016-06-02Skip analysis if the frame contains only digital silenceFelicia Lim
Signed-off-by: Jean-Marc Valin <jmvalin@jmvalin.ca>
2016-06-02NLSF_VQ_MAX_SURVIVORS is no longer usefulJean-Marc Valin
2016-06-02take advantage of more efficient NLSF quantization by reducing number of ↵Koen Vos
survivors in search JMV: edited to be a little more conservative
2016-06-02NLSF_del_dec_quant minor clean upKoen Vos
2016-06-02NLSF_encode now also uses NLSF weights from tableKoen Vos
2016-06-02NLSF VQ now uses absolute error, and predictionKoen Vos
2016-06-02NLSF decoding now uses tables for NLSF weightsKoen Vos
2016-06-02sum_sqr_shift: reduced headroom (ie more preserved resolution); shift ↵Koen Vos
increments by one instead of two
2016-06-02Simplifying fast_atan2f()Jean-Marc Valin
2016-06-02Quality: Increase CELT rate for voiced frames in hybrid modeJean-Marc Valin
2016-06-02Better modeling of the effect of CBR and complexityJean-Marc Valin
Should be able to make better mode/bandwidth decisions when CBR/complexity are involved.
2016-06-02Re-enabling high-band attenuation for hybrid modeJean-Marc Valin
The new attenuation is much smaller than previously.
2016-06-02Quality: Retrained classifier MLP with better dataJean-Marc Valin
New MLP doesn't attempt to classify silence as speech/music
2016-06-02Using "hybrid" flag instead of "start!=0"Jean-Marc Valin
2016-06-02Quality: Making the encoder more aggressive at using wider bandwidthsJean-Marc Valin
generated by cgit v1.2.3 (git 2.25.1) at 2024-06-13 14:26:06 +0300