Re: [FFmpeg-devel] avfilter/showvolume : add new options and minor clean

> > Please use short name for options, max 2 characters. > > The log scale shows nothing even thought there is no silence. Perhaps > use ceil()? > Doesn't understand this part. Do you mean, you think the log scale display option is not a good idea ? Or doesn't work in some case ? In my test, the d

Re: [FFmpeg-devel] avfilter/showvolume : add new options and minor clean

and when there is no silence i get completly empty display + > current volume in dB. > > You're right, my previous calc was wrong (doesn't display for level below -42 db) New patchs in attach use another calc for the log display : av_clipf(0.21 * log10(max) + 1, 0, 1); for the persistent max d

Re: [FFmpeg-devel] avfilter/showvolume : add new options and minor clean

> > Default white color is good decision ? Better use orange? > > Put { in own line like for every function: > > void its_me_function() > {<- have own line > > } > > Generally follow code style of file. > ___ > > New patchs in attach fix

Re: [FFmpeg-devel] avfilter/showvolume : add new options and minor clean

> > You forgot to update color in documentation. > > Patchset OK otherwise. > > Fix doc, and pushed. Thanks Martin ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] avcodec/prores_ks : use official quant matrix

2018-04-14 11:31 GMT+02:00 Martin Vignali : > > > >> > For the proxy chroma version is also the same matrix use in prores_aw >> >> Patch is OK then, assumming FATE still passes. >> >> >> Ok thanks for comments. > It passes fate test for me (but

Re: [FFmpeg-devel] swscale/swscale_unscaled : add X86_64 (SSE2, AVX) for uyvyto422

2018-04-08 22:21 GMT+02:00 Martin Vignali : > Hello, > > New patch in attach, > Remove the C part of the sse/avx func, and integrate all the process in > the ASM func > > A little bit faster, and avoid the macro in x86/rgb2rgb.c. > > > Remove the debug message i

Re: [FFmpeg-devel] fate/hap : add test for hap encoding

> > > > These tests are also broken, please fix or > > remove them: > > https://buildd.debian.org/status/fetch.php?pkg=ffmpeg&; > arch=i386&ver=7%3A4.0-1&stamp=152218&raw=0 > > ("Error 1") > > I've had a brief look at this error (and a similar error on s390x) and > it looks like a float roundin

Re: [FFmpeg-devel] fate/hap : add test for hap encoding

> > > Sorry, i missread the code before writing the test, and > > didn't see it use float for one step. > > I will send a patch to remove these tests. > > Or just commit. > > Tests removed. Martin ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org htt

Re: [FFmpeg-devel] avdevice/sdl output : fix window_size and add new option (WIP)

2018-04-20 22:24 GMT+02:00 Moritz Barsnick : > On Sun, Apr 08, 2018 at 17:28:24 +0200, Martin Vignali wrote: > > > - 001 : Fix -window_size option > > Before this patch, window_size is always set to the source size > > In other word, -window_size option have no effect. &g

Re: [FFmpeg-devel] avfilter/vf_interlace : add checkasm for lowpass_line and AVX2 version

> > >> > > >> This broke several interlace fate tests, including the new checkasm > one > > >> you added. > > >> > > New patch in attach for AVX2 version i add a process of only 1* mmsize, before the loop (who process 2 * mmsize at each loop) Pass

Re: [FFmpeg-devel] avfilter/vf_interlace : add checkasm for lowpass_line and AVX2 version

2017-12-30 19:57 GMT+01:00 Martin Vignali : > > > >> >> > >> This broke several interlace fate tests, including the new checkasm >> one >> > >> you added. >> > >> > > > New patch in attach for AVX2 version > i add a pro

[FFmpeg-devel] avcodec/utvideoenc : add SIMD (SSSE3) for sub_left_pred

Hello, in attach patch to add SIMD for sub_left_pred in utvideoenc 001 : add SIMD for utvideoenc 002 : add checkasm for llviddspenc (diff bytes and sub_left_pred) Encoding result : ./ffmpeg -i utvideo_file.avi -c:v utvideo -pred left res.avi Without frame= 3316 fps=194 q=-0.0 Lsize= 3613675kB

Re: [FFmpeg-devel] avcodec/utvideoenc : add SIMD (SSSE3) for sub_left_pred

> > this changes the output: > make -j12 && ./ffmpeg -i ~/videos/matrixbench_mpeg2.mpg -an -vcodec > utvideo -t 1 -pix_fmt yuv420p -pred left -t 1 test2.avi > > -rw-r- 1 michael michael 3744402 Jan 12 04:20 test2.avi > -rw-r- 1 michael michael 3753358 Jan 12 04:19 test.avi > > > Hello, Th

Re: [FFmpeg-devel] avcodec/utvideoenc : add SIMD (SSSE3) for sub_left_pred

Hello, Following Henrik Gramner's comments, new patch in attach i try to change int width -> ptrdiff_t width to remove movsxdifnidn but i have a segfault if height > 1 pass fate test for me. Martin 0001-avcodec-utvideoenc-add-SIMD-avx-for.patch Description: Binary data 0002-checkasm-add-tes

Re: [FFmpeg-devel] avcodec/utvideoenc : add SIMD (SSSE3) for sub_left_pred

Hello, new patch in attach no more segfault if i remove movsxdifnidn, after change in the "declare_func_emms" part I also modify the checkasm test, following your comments Martin 0001-avcodec-utvideoenc-add-SIMD-avx-for-sub_left_predict.patch Description: Binary data 0002-checkasm-add-test-f

[FFmpeg-devel] avfilter/x86/vf_blend : add avx2 for 8b func (v2)

Hello, following Henrik Gramner comments (in discussion "avfilter/x86/vf_blend : add avx2 version for 8b func (WIP)") in attach new patch to add AVX2 version for each 8b func (except divide) 001 : avutil : add ABS2 for avx2 002 : avfilter : add AVX2 version for most of the func, the AVX2 is a si

Re: [FFmpeg-devel] avfilter/x86/vf_blend : add avx2 for 8b func (v2)

2018-01-16 23:00 GMT+01:00 James Darnley : > On 2018-01-16 22:26, Martin Vignali wrote: > > diff --git a/libavutil/x86/x86util.asm b/libavutil/x86/x86util.asm > > index d7cd996842..9db2d90e57 100644 > > --- a/libavutil/x86/x86util.asm > > +++ b/libavutil/x86/x86ut

Re: [FFmpeg-devel] avfilter/x86/vf_blend : add avx2 for 8b func (v2)

Hello, New patch in attach with modification in average, grain extract, multiply, screen, grain merge -- blend Average -- Prev patch : average_c: 15605.4 average_sse2: 1205.9 average_avx2: 772.4 New patch : average_c: 15604.4 average_sse2: 490.9 average_avx2: 265.2 With 3 operand : using %if

Re: [FFmpeg-devel] [PATCH 4/4] avfilter/vf_framerate: add SIMD functions for frame blending

> if (s->bitdepth == 8) { > s->blend_factor_max = 1 << BLEND_FACTOR_DEPTH8; > -s->blend = blend_frames_c; > +if (ARCH_X86 && EXTERNAL_AVX2(cpu_flags)) > +s->blend = ff_blend_frames_avx2; > I think it's : if (EXTERNAL_AVX2_FAST(cpu_flags) > +else

Re: [FFmpeg-devel] [PATCHv2 4/4] avfilter/vf_framerate: add SIMD functions for frame blending

> + > + > +%if HAVE_AVX2_EXTERNAL > + > +INIT_YMM avx2 > +BLEND_FRAMES > + > +INIT_YMM avx2 > Don't think it's necessary to repeat INIT_YMM avx2. > +BLEND_FRAMES16 > + > +%endif > -- Martin ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http:

Re: [FFmpeg-devel] avcodec/utvideoenc : add SIMD (SSSE3) for sub_left_pred

2018-01-14 14:31 GMT+01:00 Martin Vignali : > Hello, > > new patch in attach > no more segfault if i remove movsxdifnidn, after change in the > "declare_func_emms" > part > > I also modify the checkasm test, following your comm

Re: [FFmpeg-devel] avfilter/x86/vf_blend : add avx2 for 8b func (v2)

2018-01-17 21:13 GMT+01:00 Martin Vignali : > Hello, > > > New patch in attach > > with modification in average, grain extract, multiply, screen, grain merge > > > -- blend Average -- > Prev patch : > average_c: 15605.4 > average_sse2: 1205.9 > average_

[FFmpeg-devel] avfilter/vf_blend : add 16 bit simd version for basic blend mode

Hello, Patchs in attach add SIMD (SSE and AVX2), for some blending mode of vf_blend in 16 bit Checkasm result : addition_16_c: 7921.6 addition_16_sse2: 485.1 addition_16_avx2: 280.6 and_16_c: 6821.1 and_16_sse2: 443.4 and_16_avx2: 429.4 darken_16_c: 8454.6 darken_16_sse4: 492.9 darken_16_avx2: 28

Re: [FFmpeg-devel] [PATCH]lavc/exr: Ignore long names flag

2018-01-30 14:40 GMT+01:00 Carl Eugen Hoyos : > 2018-01-30 13:16 GMT+01:00 Paul B Mahol : > > On 1/30/18, Carl Eugen Hoyos wrote: > >> 2018-01-30 10:30 GMT+01:00 Paul B Mahol : > >>> On 1/30/18, Carl Eugen Hoyos wrote: > > Attached patch fixes ticket #6994, unknown tag names are >

Re: [FFmpeg-devel] [PATCH]lavc/exr: Ignore long names flag

> > Until now, two types of files are supported: > Files with flags==0 and files with flags&0x02. > The specification requires that (flags&0x02) > implies flags==2 or possibly flags==6: "if bit > 9 is 1, bits 11 and 12 must be 0" > > After my patch, the following values are > accepted for flags: 0,

Re: [FFmpeg-devel] avfilter/vf_blend : add 16 bit simd version for basic blend mode

> > breaks build on x86-32 linux > > X86ASM libavfilter/x86/vf_blend.o > src/libavfilter/x86/vf_blend.asm:389: error: invalid combination of > opcode and operands > src/libavfilter/x86/vf_blend.asm:390: error: invalid combination of > opcode and operands > src/libavfilter/x86/vf_blend.asm:391: err

[FFmpeg-devel] fate/exr : add test for ticket 6994 (long name flag)

Hello, Patch in attach add test for ticket 6994 (flag long name) Sample can be found here : https://we.tl/WBnt10VSA1 and need to be put inside ./fate-suite/exr Can be test with : make fate-exr SAMPLES=fate-suite/ Need to be apply after one of the patch in discussion : "lavc/exr: Ignore long n

Re: [FFmpeg-devel] fate/exr : add test for ticket 6994 (long name flag)

2018-02-09 9:22 GMT+01:00 Martin Vignali : > Hello, > > Patch in attach add test for ticket 6994 (flag long name) > > Sample can be found here : > https://we.tl/WBnt10VSA1 > > and need to be put inside ./fate-suite/exr > > Can be test with : > make fate-exr SAM

Re: [FFmpeg-devel] [PATCH]lavc/exr: Ignore long names flag

2018-02-08 21:45 GMT+01:00 Martin Vignali : > > >> Until now, two types of files are supported: >> Files with flags==0 and files with flags&0x02. >> The specification requires that (flags&0x02) >> implies flags==2 or possibly flags==6: "if bit >>

[FFmpeg-devel] avfilter/vf/blend : add 16 bits SIMD version for basic mode (v2)

Hello, New patchs in attach Limit the asm version to x86_64 tested on osx X86_64 and osx X86_32 Martin 0001-checkasm-vf_blend-add-depth-param-in-order-to-add-te.patch Description: Binary data 0002-avfilter-x86-vf_blend-reorganize-init-in-order-to-ad.patch Description: Binary data 0003-avf

[FFmpeg-devel] avcodec/hapqa_extract_bsf : add bsf filter for haqqa (to hapq or hapalpha only) conversion

Hello, Like no comment have been made for the WIP version (November 2017), ready to apply patchs in attach pass fate on osx (x86_64) Update : - Add doc for new bsf filter The goal is to convert HAPQA file to HAPQ (removing alpha) or HAPAlphaOnly (remove rgb) HAPQA data, is separate in two part

Re: [FFmpeg-devel] avcodec/hapqa_extract_bsf : add bsf filter for haqqa (to hapq or hapalpha only) conversion

Hi Tom, > > if (ctx->texture == 1) { > > ... > > The ordering of sections in a Hap frame is not specified, so you can’t > assume the first is the colour and the second alpha - you will have to > query the section types for both sections and choose the appropriate one. > Ok > > > bytestream2_se

Re: [FFmpeg-devel] [PATCH] Added support for version field flag 0x04 in OpenEXR files

2018-02-21 1:25 GMT+01:00 Daniel Flehner Heen : > Hi! > > First time poster so please excuse any mistakes made. I've tried submitting > a patch through git send-email but got issues authentication through gmail. > Attached is a patch file created with git format-patch. > > Commit message: > > Adde

Re: [FFmpeg-devel] fate/exr : add test for ticket 6994 (long name flag)

2018-02-13 12:30 GMT+01:00 Martin Vignali : > > > 2018-02-09 9:22 GMT+01:00 Martin Vignali : > >> Hello, >> >> Patch in attach add test for ticket 6994 (flag long name) >> >> Sample can be found here : >> https://we.tl/WBnt10VSA1 >> >>

Re: [FFmpeg-devel] avfilter/vf/blend : add 16 bits SIMD version for basic mode (v2)

2018-02-17 21:51 GMT+01:00 Martin Vignali : > Hello, > > New patchs in attach > Limit the asm version to x86_64 > > tested on osx X86_64 and osx X86_32 > > > Martin > Pushed ___ ffmpeg-devel mailing list ffmpeg-deve

Re: [FFmpeg-devel] [PATCH]lavc/exr: Ignore long names flag

> > Hello, > > If no one is against, i will apply the patch in attach in few days (change > since previous patch (mention ticket in commit msg)) > > Martin > > Pushed ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo

Re: [FFmpeg-devel] [PATCH] checkasm/hevc_sao : add hevc_sao for checkasm

2018-03-02 12:48 GMT+01:00 Yingming Fan : > If you want to test checkasm of hevc_sao, please use `checkasm > --test=hevc_sao`, or `checkasm --test=hevc_sao --bench`. > > > > > On 2 Mar 2018, at 7:03 PM, Yingming Fan wrote: > > > > --- > > tests/checkasm/Makefile | 2 +- > > tests/checkasm/chec

Re: [FFmpeg-devel] [PATCH] checkasm/hevc_sao : add hevc_sao for checkasm

Works for me on MacOs 10.12 (x86_64 and x86_32). Martin ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

[FFmpeg-devel] swscale/rgb2rgb : add X86_64 SIMD (SSSE3 and AVX2) for shuffly_bytes func

Hello, Patch in attach add SIMD for the 5 shuffle_bytes func for rgb2rgb The new SIMD are write using external ASM. Also add checkasm test for theses func Restricted to x86_64, because the scalar part doesn't compile on x86_32 I consider for the scalar part that the src_size value is a multiple

[FFmpeg-devel] avformat/mov : add support for read/write Adobe Alpha Udta

Hello, Patch in attach add suport for reading ALFA udta set (at least), by Adobe Software and for writing it. Theses two patch, can keep the original metadata when remuxing it And the user can force the value using -metadata alpha= In order to have the right alpha mode set when importing the mov

[FFmpeg-devel] libavformat/segment : Patch proposal : Preserving timecode with segment

Hello, In attach, a first draft, for a patch in order to have a correct timecode when segment files (before this patch, the timecode is reset for each segment). This is not a ready to apply patch. One problem, is to get the real rate of the file, in order to correctly increment timecode (i set 2

Re: [FFmpeg-devel] libavformat/segment : add option to increment timecode

ping 2015-10-03 15:49 GMT+02:00 Martin Vignali : > Hello, > > in attach a patch, who add an option to increment timecode using segment. > > To test : > > ffmpeg -i src.mov -timecode 10:00:00:00 -vcodec copy -f segment > -segment_time 2 -reset_timestamps 1 -increment_tc 1

[FFmpeg-devel] Fwd: libavformat/segment : add option to increment timecode

Ping -- Forwarded message -- From: Martin Vignali Date: 2015-10-03 15:49 GMT+02:00 Subject: libavformat/segment : add option to increment timecode To: FFmpeg development discussions and patches Hello, in attach a patch, who add an option to increment timecode using segment

Re: [FFmpeg-devel] Fwd: libavformat/segment : add option to increment timecode

2016-01-29 12:35 GMT+01:00 Stefano Sabatini : > On date Friday 2016-01-29 12:18:02 +0100, Martin Vignali encoded: > > Ping > > > > -- Forwarded message ------ > > From: Martin Vignali > > Date: 2015-10-03 15:49 GMT+02:00 > > Subject: libavf

Re: [FFmpeg-devel] Fwd: libavformat/segment : add option to increment timecode

2016-01-30 10:44 GMT+01:00 Stefano Sabatini : > On date Friday 2016-01-29 17:38:13 +0100, Martin Vignali encoded: > > 2016-01-29 12:35 GMT+01:00 Stefano Sabatini : > [...] > > > +AVDictionaryEntry * tcr = av_dict_get(s->metadata, "timecode", > &

[FFmpeg-devel] Fwd: Fwd: libavformat/segment : add option to increment timecode

2016-01-30 10:44 GMT+01:00 Stefano Sabatini : > On date Friday 2016-01-29 17:38:13 +0100, Martin Vignali encoded: > > 2016-01-29 12:35 GMT+01:00 Stefano Sabatini : > [...] > > > +AVDictionaryEntry * tcr = av_dict_get(s->metadata, "timecode", > &

[FFmpeg-devel] libavformat/segment : add option to increment timecode

Hello, in attach a patch, who add an option to increment timecode using segment. To test : ffmpeg -i src.mov -timecode 10:00:00:00 -vcodec copy -f segment -segment_time 2 -reset_timestamps 1 -increment_tc 1 target_%03d.mov the second file have timecode 10:00:02:00 (after the patch) instead of 1

[FFmpeg-devel] Patch proposal : Fix crash with astats filter when no audio input

Hello, Exemple to create the crash : ffmpeg -i fileWithoutSound -filter astats -f null - A patch in attach avoid the crash by avoiding Zero division in print_stats. I test the patch on several files, and the crash doesn't appear anymore. Best regards Martin 0001-Fix-crash-in-astats-filter-wh

Re: [FFmpeg-devel] Patch proposal : Fix crash with astats filter when no audio input

annels == 0) || (nb_samples !=0) return; Best regards Martin 2015-03-08 0:07 GMT+01:00 Nicolas George : > Le septidi 17 ventôse, an CCXXIII, Martin Vignali a écrit : > > Exemple to create the crash : > > ffmpeg -i fileWithoutSound -filter astats -f null - > > I experienc

Re: [FFmpeg-devel] avfilter/af_astats: Avoid Zero division in print part.

The divide-by-zero in channel part is not executed, when there is no sound (like with file i put in attach in m previous email) because, nb_channels = 0 in this case, so the for loop is not executed. 2015-03-08 10:36 GMT+01:00 Nicolas George : > L'octidi 18 ventôse, an CCXXIII, Paul B Mahol a écr

Re: [FFmpeg-devel] avfilter/af_astats: Avoid Zero division in print part.

In attach a sample file I use this line to create the crash : ffmpag -i path_to_file_in_attach -filter astats -f null - 2015-03-08 11:05 GMT+01:00 Nicolas George : > L'octidi 18 ventôse, an CCXXIII, Paul B Mahol a écrit : > > Well, nb_channels is for certain 0 here. > > Well, I am not "here" and

[FFmpeg-devel] Patch proposal : filter ebur128 : add filter name before each line of summary print

In attach a patch proposal, in order to make easier parse result of ebur128 filter by external scripts command line for testing : ffmpeg -i aFile.wav -filter ebur128=peak=true+sample -f null - Before patch, summary look like this : [Parsed_ebur128_0 @ 0x7fe8b0d0] Summary: Integrated loudne

[FFmpeg-devel] Ticket 3530 : Fix proposal

In attach a patch for ticket 3530 : https://trac.ffmpeg.org/ticket/3530 the file Apache_cut.aif of ticket 3530, works after the patch But i'm not sure, this is the best way to correct it, because with this sample file, avio_feof return 1 before the end of the file. (tested with Mac os 10.9, Clang

[FFmpeg-devel] Patch proposal : Remove some unused variable

I'm not sure about ffmpeg choice for compiler warning. An old post on the mailing list, say : it depends !! So i propose remove some unused variables, to remove some 'unused variable' warning. (Clang compiler Mac os 10.9) Patch in attach. Martin 0001-remove-unused-var-in-vfpp7.patch Descriptio

[FFmpeg-devel] Remove some compiler warning

in attach patchs in order to avoid some compiler warning (CLang Mac os 10.9) 001, and 002 : remove unused variable 003, and 004 : Convert const AVFrame * to AVFrame * Martin 0001-remove-unused-variable-lsp16s.patch Description: Binary data 0002-remove-unused-variable-bark_tab_s16_64.patch D

[FFmpeg-devel] Ticket 4386 : crash create by previous patch

Hello, After some research, the crash when reading exr Piz file (ticket 4386), appear with this patch : https://github.com/FFmpeg/FFmpeg/commit/586ba24ff29468d2a4ee843a9650feea5b2be6f6 if, i use the previous line : memset(lut + k, 0, (USHORT_RANGE - k) * 2); instead of the new one : memset(lut +

Re: [FFmpeg-devel] [PATCH] libavcodec/exr.c : exr lossy compression B44 based on OpenEXR code

I doesn't test yet the patch. But if you need to, two samples create with After Effects from ( http://fr.wikipedia.org/wiki/FFmpeg#/media/File:FFmpeg_Logo_new.svg) EXR B44 Half Float http://www.datafilehost.com/d/b3fdd446 EXR B44 Float http://www.datafilehost.com/d/5fea8262 Martin

Re: [FFmpeg-devel] avdevice/sdl output : fix window_size and add new option (WIP)

> > Thanks for comments and testing. > Dropped the window_pos patch. > > New patch in attach : > - 001 : Fix -window_size option --> Unchanged patch > - 002 : Add option to disable quit action : Add doc > > > Pushed Martin ___ ffmpeg-devel mailing list f

[FFmpeg-devel] avcodec/proresenc_aw improvements

Hello, Patch in attach improve some part of the prores_aw encoder. 001 : use scantable in prores_data instead of a duplicate one. 002 : use for the frame header, primaries, transfert, colorspace like in proresenc_ks avoid color shift on some software (update fate test) 003 : change qp start value

Re: [FFmpeg-devel] avcodec/proresenc_aw improvements

> > Am I correct that the output files also change? > Patch 002 : change the header of the output file only, not the compress data. Patch 003 : change the output file mainly for near uniform bloc, where the low scale quantif value where used. But IMHO is better by default to be close to official

[FFmpeg-devel] avcodec/proresenc_aw : add support for prores 444

Hello Patch in attach add support for prores 444 profile in AW prores encoder Need to be apply after patch in discussion "avcodec/proresenc_aw improvments" Speed (from prores 422 HQ file -> Prores 444) ./ffmpeg -i prores422hqfile.mov -pix_fmt yuv444p10 -c:v prores -an -profile:v 4 res_aw.mov -y

Re: [FFmpeg-devel] avcodec/proresenc_aw : add support for prores 444

> > > > Speed (from prores 422 HQ file -> Prores 444) > > > > ./ffmpeg -i prores422hqfile.mov -pix_fmt yuv444p10 -c:v prores -an > > -profile:v 4 res_aw.mov -y > > ==> 25 fps > > > > ./ffmpeg -i prores422hqfile.mov -pix_fmt yuv444p10 -c:v prores_ks -an > > -profile:v 4 res_ks.mov -y > > ==> 10 fps

Re: [FFmpeg-devel] avcodec/proresenc_aw improvements

> > > 001 : use scantable in prores_data instead of a duplicate one. > > This could negativly affect the performance of the changed inner loop > have you checked that the changed function does not become slower ? > > > No, i will make some tests. > > -*buf++ = 6; > > +*buf++ = pict->color_

Re: [FFmpeg-devel] avcodec/proresenc_aw improvements

> > > > > Do you think it's better to only authorize few colorspace ? > > depends on what happens if "others" are stored. > if the official decoders fail with a blank screen then its probably > not a good idea to use such a value. If OTOH they ignore values they > do not support then it may be ok.

Re: [FFmpeg-devel] [PATCH 4/7] Adds gray floating-point pixel formats.

> > +static int uint_y_to_float_y_wrapper(SwsContext *c, const uint8_t *src[], > + int srcStride[], int srcSliceY, > + int srcSliceH, uint8_t *dst[], int > dstStride[]) > +{ > +int y, x; > +int dstStrideFloat = dstStrid

[FFmpeg-devel] avcodec/psd : add support for gray float (WIP)

Hello, Like gray float pix fmt is now added, add support for PSD gray float picture Sample can be found here : https://we.tl/t-QHDNyfMr1S Can be test with : ./ffmpeg -i lena-gray_float.psd -pix_fmt gray8 res8.png Doesn't work with : ./ffmpeg -i lena-gray_float.psd res16.png I hoped to have som

[FFmpeg-devel] swscale : hScale16To19 : limit shift for float(32bits) input

Hello, Patch in attach fix for me, Gray Float to UINT16 conversion (see discussion : avcodec/psd : add support for gray float (WIP)) After this patch and psd patch, this command line works : ./ffmpeg -i lena-gray_float.psd res16.png Maybe not the best way to fix that. Comments welcome Martin

Re: [FFmpeg-devel] swscale : hScale16To19 : limit shift for float(32bits) input

Hello, New patch in attach Add an if case for float input, in order to define the shift value. I try to add some details in the commit msg. Pass fate test for me (Mac os 10.12, x86_64). Martin 0002-swscale-treat-float-input-data-as-uint-16bpc.patch Description: Binary data 0003-swscale-sws

Re: [FFmpeg-devel] swscale : hScale16To19 : limit shift for float(32bits) input

Updated patch in attach also fix the conversion when using float gray as input, and 8bpc uint as output Martin 0002-swscale-treat-float-input-data-as-uint-16bpc.patch Description: Binary data 0003-swscale-swscale-small-cosmetic.patch Description: Binary data ___

Re: [FFmpeg-devel] avcodec/psd : add support for gray float (WIP)

Hello, Better patch in attach (reuse the same "copy part" than uint 8bpc and 16 bpc pix fmt) Works only after applying swscale patch in discussion : swscale : hScale16To19 : limit shift for float(32bits) input Can be test with : ./ffmpeg -i lena-gray_float.psd -pix_fmt gray8 res8.png ./ffmpeg -i

[FFmpeg-devel] swscale/input : avoid float calc for GrayFloat to Gray16 conv

Hello, Patch in attach modify GrayFloat to Gray16 conversion using the same method currently use inside exr decoder (no float calc) duplicate the float_to_uint func inside swscale_internal Martin 0004-swscale-input-avoid-float-calc-for-grayFloat-to-uint.patch Description: Binary data _

Re: [FFmpeg-devel] [PATCH 4/7] Adds gray floating-point pixel formats.

Hello, I am having trouble reproducing this error. These tests are fine for 32-bit > VMs on > my computers. So the only thing I can do is to disable these tests for > these formats. > Otherwise, I need to test other changes somehow. Here is the patch, that > skips > pixfmts tests for these formats

Re: [FFmpeg-devel] [PATCH 4/7] Adds gray floating-point pixel formats.

> > > So the only thing I can do is to disable these tests for > > these formats. > > Otherwise, I need to test other changes somehow. Here is the patch, that > > skips > > pixfmts tests for these formats. > > in absence of another solution this should be ok > > I'm not against, removing these test

Re: [FFmpeg-devel] avcodec/psd : add support for gray float (WIP)

Le lun. 20 août 2018 à 15:39, Martin Vignali a écrit : > Hello, > > Better patch in attach (reuse the same "copy part" than uint 8bpc and 16 > bpc pix fmt) > > Works only after applying swscale patch in discussion : > swscale : hScale16To19 : limit shift for floa

Re: [FFmpeg-devel] swscale : hScale16To19 : limit shift for float(32bits) input

> > patches should be ok > > Pushed, thanks. Martin ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] swscale/input : avoid float calc for GrayFloat to Gray16 conv

> > --- > > libswscale/input.c| 10 +- > > libswscale/swscale_internal.h | 20 > > 2 files changed, 25 insertions(+), 5 deletions(-) > > please provide benchmark, what is the impact on speed from this ? > > In my tests, the patch increase speed by around 20

Re: [FFmpeg-devel] swscale/input : avoid float calc for GrayFloat to Gray16 conv

> > well then it should be ok > not sure if the ff_ prefix is ideal for a static inline function > > Do you think flt_2_uint16 is a better name ? (can probably be interesting to also add same kind of func for float to uint8 in unscaled part) Martin ___ f

Re: [FFmpeg-devel] [PATCH 4/7] Adds gray floating-point pixel formats.

> > But maybe to make tests simpler, we can use/add bit exact conversion > > for uint8 to float, we can generate a LUT without float calc > > and for uint16 to float, we can add a uint16 to float conversion without > > float calc, or maybe, generate a LUT in bit exact mode (probably faster, > if >

<    1   2   3   4   5   6