summaryrefslogtreecommitdiff
path: root/silk/fixed
Commit message (Collapse)AuthorAgeFilesLines
* Change pitch scaling behavior wrt nFramesPerPacketJean-Marc Valin2022-08-041-1/+1
| | | | | | | | Not sure if it was the original intent, but we now reduce the loss percentage threshold for pitch scaling as 1/nFramesPerPacket since only the first frame will have pitch scaling anyway. As a side effect, this brings back the original behavior of disabling pitch scaling for 0% loss.
* Re-tuning the use of LTP scalingJean-Marc Valin2022-07-241-2/+7
| | | | | Making LTP scaling depend on the bitrate and whether FEC is on. The thresholds for scaling 1 and 2 are now independent.
* Fixes valgrind failure caused by silk_find_pred_coefs_*()Jean-Marc Valin2022-06-281-1/+2
| | | | | | | | The function copies NLSFs from the stack to the state which for order 10 means we were copying uninitialized values. That in turn breaks check-asm when comparing the state under valgrind. Reviewed by Timothy B. Terriberry.
* Fixes wrap-around in silk_inner_prod16_sse4_1()Jean-Marc Valin2022-06-271-1/+2
| | | | Thanks Tim
* Update and re-enable SILK SSE4.1 optimisationsFrancis Quiers2022-03-075-206/+73
|
* Fix NEON optimizations buffer read overrunFelicia Lim2019-09-031-2/+7
| | | | Thanks to Ray Essick
* Fix build errors using Makefile.mipsMark Harris2019-04-082-4/+5
| | | | Broken by earlier ARM optimizations.
* Silk makes use of Opus VADGustaf Ullberg2018-05-142-3/+11
| | | | Signed-off-by: Jean-Marc Valin <jmvalin@jmvalin.ca>
* Harden SILK x86 codeJean-Marc Valin2018-03-272-3/+3
|
* Hardening silk/fixedJean-Marc Valin2018-03-2712-38/+38
|
* Fixed off-by-one issue in Silk DTXGustaf Ullberg2018-02-221-1/+1
| | | | Signed-off-by: Jean-Marc Valin <jmvalin@jmvalin.ca>
* Rename SSE 4.1 files to match their targets.Ralph Giles2017-09-112-0/+0
| | | | | | | | | Distinguish source files for the SSE 4.1 instruction set extension consistently by their filename. This makes it easier to check the correct flags are being set at build time. Signed-off-by: Jonathan Lennox <jonathan@vidyo.com>
* Optimize silk_warped_autocorrelation_FIX() for ARM NEONLinfeng Zhang2017-04-137-10/+347
| | | | | | | | | | | The optimization is bit exact with C function. This optimization speeds up fixed-point SILK encoder on NEON about 5% to 8%. (Tested on Acer Chromebook, ARMv7 Processor rev 3 (v7l).) Change-Id: I582f6f3585b7946149e16a2ad3084ebc0ae79a4f Signed-off-by: Jean-Marc Valin <jmvalin@jmvalin.ca>
* Fix compiler warningsMark Harris2017-02-263-12/+14
| | | | | | | | | | | | | | | | | | | - celt/modes.c:430:14: warning: cast from 'const unsigned char *' to 'opus_int16 *' increases required alignment from 1 to 2 [-Wcast-align] - 'C[0][1]' may be used uninitialized [-Wmaybe-uninitialized] - Unused variable/parameter - Value stored is never read - MSVC warnings about "possible loss of data" due to type conversions - MSVC warning C4146: unary minus operator applied to unsigned type - silk/NLSF_del_dec_quant.c:137:20: warning: array subscript is above array bounds [-Warray-bounds] (gcc -O3 false positive) - src/mlp_train.h:39:20: warning: function declaration isn't a prototype [-Wstrict-prototypes] - Remove SMALL_FOOTPRINT code from SSE 4.1 FIR implementation, matching the C implementation. The clang -Wcast-align warnings with SSE intrinsics are a known clang issue: https://llvm.org/bugs/show_bug.cgi?id=20670
* Optimize silk_LPC_inverse_pred_gain() for ARM NEONLinfeng Zhang2017-02-142-3/+3
| | | | | | | | The optimization is bit exact with C function. Change-Id: Ib3bdc26a5a4ebe02e7f24be85104e8e9a2a9a738 Signed-off-by: Jean-Marc Valin <jmvalin@jmvalin.ca>
* Eliminate trailing spacesMark Harris2016-11-021-1/+1
|
* Fixing some opus_int vs opus_int32 mismatchesJean-Marc Valin2016-10-051-2/+2
| | | | Reported by Mark Warner.
* Fixes an overflow in limit_warped_coefs()Jean-Marc Valin2016-07-251-4/+7
| | | | | For large values of maxabs_Q20, silk_MUL( maxabs_Q20, ind + 1 ) could overflow.
* Disables MIPS silk_noise_shape_analysis_FIX() until it's updatedJean-Marc Valin2016-07-221-1/+2
|
* More headroom in fixed-point silk_burg_modified() to avoid an overflowJean-Marc Valin2016-07-221-1/+1
|
* Increase headroom in silk_warped_autocorrelation_FIX()Jean-Marc Valin2016-07-221-1/+1
|
* Increase headroom to avoid overflow in silk_pitch_analysis_core()Jean-Marc Valin2016-07-221-1/+1
|
* Avoids undefined behaviour from left-shifting negative valuesJean-Marc Valin2016-07-201-2/+2
|
* Fix include warningTristan Matthews2016-07-191-0/+1
|
* Moving FEC decision to the Opus encoder level (from SILK level)Jean-Marc Valin2016-07-191-1/+0
|
* CBR: set gains to their previons values where we're busting the budgetJean-Marc Valin2016-07-191-3/+21
|
* Set pulses to zero if we can't meet rate targetJean-Marc Valin2016-07-191-0/+13
|
* CBR: lock the gain on a subframe when the number of pulses stops going downJean-Marc Valin2016-07-191-1/+25
|
* merge back in the LTP gain safety limiterKoen Vos2016-07-191-1/+2
|
* CBR: Using the same gain increase for fixed-point as for floatJean-Marc Valin2016-07-191-5/+10
|
* increase order of noise shaping filterKoen Vos2016-07-192-2/+2
|
* more conservative scaling of LTP corrs; remove assert that checks for ↵Koen Vos2016-07-192-8/+8
| | | | negative residual energy; discard CB entries leading to such negative energies
* Clean up: alignment of commentsKoen Vos2016-07-171-4/+4
|
* Clean up: replace tabs by spacesKoen Vos2016-07-176-58/+58
|
* removed prefilterKoen Vos2016-07-177-389/+76
| | | | The NSQ SSE optimizations are disabled for now because they need to be updated
* undo some loop unrollingKoen Vos2016-07-171-9/+3
|
* pitch analysis (in FIX) now scales the input down at the start, instead of ↵Koen Vos2016-07-171-56/+37
| | | | at every stage
* minor clean up of pointer offsetsKoen Vos2016-07-173-13/+11
|
* simplified computation of LTP coefsKoen Vos2016-07-175-516/+89
|
* Apply 02f3ac225f99 fix for divide-by-zero to the SSE4.1 version of ↵Jean-Marc Valin2016-07-151-6/+8
| | | | silk_burg_modified()
* Fixes signed integer overflow in fixed-point BurgJean-Marc Valin2016-06-191-2/+5
| | | | | We just explicitly allow the overflow with silk_MLA_ovflw() since the result seems to be correct because the overflows cancel each other.
* fix for divide by zeroFelicia Lim2016-03-221-6/+8
|
* Declare silk_warped_LPC_analysis_filter_FIX_c in silk/fixed/main_FIX.h.Jonathan Lennox2015-11-021-0/+11
| | | | Fixes build failure on platforms with MAY_HAVE_SSE4_1 (but not PRESUME_SSE4_1) with --enable-intrinsics.
* Fixes compile problems for MIPSRhishikesh Agashe2015-10-072-2/+10
| | | | | | Brings MIPS in sync with the ARM/SSE optimizations that added "arch" parameters. Signed-off-by: Jean-Marc Valin <jmvalin@jmvalin.ca>
* Fix for flutter with FECKoen Vos2015-08-101-0/+1
| | | | | | | | | The bug was caused by an improper feedback of the per-frame bitrate, causing the bitrate to jump up and down from frame to frame, within a packet. The patch avoids this, and also gives a slight improvement in general for multi-frame packets, even without FEC. Signed-off-by: Jean-Marc Valin <jmvalin@jmvalin.ca>
* There are no tabs in source code.Timothy B. Terriberry2014-10-033-8/+8
| | | | There is also no trailing whitespace.
* Cisco optimization for x86 & fixed pointxiangmingzhu2014-10-0316-52/+692
| | | | | | | | | | 1. Only for fixed point on x86 platform (32bit and 64bit, uses SIMD intrinsics up to SSE4.2) 2. Use "configure --enable-fixed-point --enable-intrinsics" to enable optimization, default is disabled. 3. Official test cases are verified and passed. Signed-off-by: Timothy B. Terriberry <tterribe@xiph.org>
* Using 64-bit acculumation for C0 in silk_burg_modified()Marcello Caramma (mcaramma)2014-07-011-16/+12
| | | | Signed-off-by: Jean-Marc Valin <jmvalin@jmvalin.ca>
* Whitespace fixesexp_mips_optJean-Marc Valin2014-06-195-9/+2
|
* MIPS optimizationsRhishikesh Agashe2014-06-196-0/+711
| | | | Signed-off-by: Jean-Marc Valin <jmvalin@jmvalin.ca>
View Lorry Controller Status