Re: [FFmpeg-user] Testing transcode speeds on Raspberry Pi 4
Hey, > An attempt to do some hardware acceleration... >> ffmpeg -i path/to/source.mp4 -c:v h264_omx -b:v 9000k -vf >> scale=1280:720,format=yuv420p path/to/dest.mp4 > Gave the exact same same results That sounds to me like the encoder is waiting for filtered frames in both cases. Did you try without the -vf ? Regards, Ted Park ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
[FFmpeg-user] Testing transcode speeds on Raspberry Pi 4
Running a few tests to see if there's any practical value in using a RPI4 to generate h.264s from camera originals using FFMpeg. So far it's not exactly overwhelming: > ffmpeg -i path/to/source.mp4 -b:v 9000k -vf scale=1280:720,format=yuv420p > path/to/dest.mp4 ran at about 25% of real time (compared to a Mac that ran the same file at >4x real time) An attempt to do some hardware acceleration... > ffmpeg -i path/to/source.mp4 -c:v h264_omx -b:v 9000k -vf > scale=1280:720,format=yuv420p path/to/dest.mp4 Gave the exact same same results I don't know I would expect more from a $50 computer -- the fact that it works at all is already impressive -- but I thought I'd ask if anyone has any ideas on how to get more speed out of the little thing. ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-user] Flutter FFmpeg
Hi, > ``` > int resultCode = await _ffmpeg.executeWithArguments([ >"-noautorotate", >"-i", >fileToCompress.path, >"-movflags", >"+faststart", >"-vcodec", >"libx264", >"-crf", >"", >'-preset:v', >'', >"-vf", >"format=yuv420p", >result.path > ]); > ``` > > **Current behaviour** > This works well on android (testing with a oneplus 6) as it always seems to > compress the file to <20MB in a reasonable time. However when i test this on > IOS (Iphone SE) the imagepicker seems to compress the video first e.g. 150MB > to 25MB. Which I then attempt to compress the 25MB again with above command > with params "veryfast" and 18. > > The file always gets larger that the original 25mb and it takes forever to > execute (>3min). Similar behaviour is experienced when I use different params > e.g. 24 and faster. Someone else recently uploaded an iPhone recording that qtfaststart completely fubar'ed, I think something to do with the export/share option adding a bunch of top-level metadata atoms after the moov? Could you possibly try importing the file doing your best not to let iOS alter it, and compare it to the file you transferred using the original method? If you're on a mac, a USB cable and Image Capture app should work, but drag the file somewhere, instead of importing it into Photos or other library. Regards, Ted Park ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-user] concat demuxer filter_complex (fade)
Hi, > I learned, that it's possible to define multiple inputs and reference then > later in the filter_complex to the different inputs like this: > > ffmpeg -i vid1.mkv -i vid2.mkv -i vid3.mkv -filter_complex > "[0:v]fade=t=in:st=0:d=1[v0]; [1:v]fade=t=in:st=0:d=1[v1]; > [2:v]fade=t=in:st=0:d=1[v2]; > [v0][0:a][v1][1:a][v2][2:a]concat=n=3:v=1:a=1[v][a]" -map "[a]" -map "[v]" > out.mkv > > But is this possible with using the concat demuxer (providing the files by > writing them into a file) too? And if yes how? > (Goal is to specify the videos through a file, using the concat demuxer, but > apply the fade filter to each of the video files before concatenating them. > Is it possible to make this in one step?) > Can you please elaborate what you mean by fragile and complicated? I mean, you've got to see how that is complicated if nothing else. Anyway, I remembered that using the ffconcat file, you can define streams within the virtual single input from the concat demuxer, but I've only ever seen it used to specify specific elementary streams within vob files, not sure if it can be used with other demuxers. Maybe someone else does? Regards, Ted Park ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-user] ffplay options for setting audio volume and locating player windows on the screen
Hi, > Have you tried the good ole tried and true X / XQuartz -geometry options? > I have not used OSX in a very long time (since the switch to Quartz), so > there may be some differences between the old Xorg and XFree nomenclature, > but I imagine those types of options would still be there for initial > window placements and sizes. Unfortunately X11 is entirely deprecated and not even supported as an optional install. The latest release of XQuartz is something like 5 years old and includes server version 1.17 or something. You can build yourself with the help of ports etc but a lot of things don't work and is a pain in the neck in general. I'm pretty sure the static build is appkit. Regards, Ted Park ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-user] Decombing via screening - 'tblend' bug (?)
Hello! On 04/15/2020 07:13 PM, pdr0 wrote: On 04/15/2020 00:25 AM, I wrote: This: ffmpeg -i IN -filter_complex "telecine=pattern=5,split[A][B],[A]select='not(eq(mod(n+1\,5)\,3))'[C],[B]select='eq(mod(n+1\,5)\,3)',split[E][F],[E][F]blend[D],[C][D]interleave" OUT outputs 598 frames. 'blend' outputs as expected. May I make it plainer? Filtergraph using 'blend': IN (240 fr) > telecine (600 fr) > (cont. below) split#1 [A] > select (480 progressive frames) > [C] [B] > select (120 combed frames) > [D] [C] > (480 progressive frames) > interleave (600 fr) [D] > split#2 > (120 Cframes) > blend (120 fr) > > (120 Cframes) > Note: interleave output is actually 598 frames. This: ffmpeg -i IN -filter_complex "telecine=pattern=5,split[A][B],[A]select='not(eq(mod(n+1\,5)\,3))'[C],[B]select='eq(mod(n+1\,5)\,3)',separatefields,scale=height=2*in_h:sws_flags=neighbor,setsar=1,tblend[D],[C][D]interleave" OUT outputs 716 frames. 'tblend' (documented in the same article) outputs extra frames. Filtergraph using 'tblend': IN (240 fr) > telecine (600 fr) > (cont. below) [C] > (480 progressive frames) > interleave (720 fr) [D] > separatefields (240 fr) > tblend (240 fr) > Note: interleave output is actually 716 frames. My point is that 'blend' outputs 120 frames while 'tblend' outputs 240 (actually, 219) frames. Due to slipshod documentation -- the documentation is combined -- the difference is not noted. Of course, my complaint began as a bug complaint, not a documentation complaint. I can live with behavior differences. You would expect 719 frame output if you started with a 240 frame, 23.976p clip Sure. tblend changes the frame count by -1 . Default mode doesn't appear to do anything except drop the original frame zero. When you use all_mode=average, it blends adjacent frames. The new frame zero becomes 50/50 mix of (old frame 0,1). New frame one becomes a 50/50 mix of (old frame 1,2). You can test the filter by itself to verify this When you separate fields, you have 2 times the number of original frames. If you resize them to full height and treat them as frames, you still have 2 times the number of frames on that selection set Of course. If you started with a 240 frame clip , you should end up with a 600 frame clip after that telecine filter only with those settings. Of course. If you take every 5th frame from the telecine output ; 600/5 =120 . Separating fields give you 120*2 =240. Applying tblend=all_mode=average after gives you 240-1=239 frames. This is [D] . Of course. [C] is the other frame selection set; 600-120=480 Interleaving selection [C] with [D] should give you (600-120) + (240-1) = 719 The 716 might be from interleave dropping frames at the end when it hangs. Expect it to hang because there is no end-of-stream signal when used with select. https://ffmpeg.org/ffmpeg-filters.html#interleave_002c-ainterleave You have confirmed everything that I assert. Thank you. ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-user] Decombing via screening - 'tblend' bug (?)
This: ffmpeg -i IN -filter_complex "telecine=pattern=5,split[A][B],[A]select='not(eq(mod(n+1\,5)\,3))'[C],[B]select='eq(mod(n+1\,5)\,3)',split[E][F],[E][F]blend[D],[C][D]interleave" OUT outputs 598 frames. 'blend' outputs as expected. This: ffmpeg -i IN -filter_complex "telecine=pattern=5,split[A][B],[A]select='not(eq(mod(n+1\,5)\,3))'[C],[B]select='eq(mod(n+1\,5)\,3)',separatefields,scale=height=2*in_h:sws_flags=neighbor,setsar=1,tblend[D],[C][D]interleave" OUT outputs 716 frames. 'tblend' (documented in the same article) outputs extra frames. Now, does that look consistent to you? But of course, since I can't read, and I'm always wrong... ___ You would expect 719 frame output if you started with a 240 frame, 23.976p clip tblend changes the frame count by -1 . Default mode doesn't appear to do anything except drop the original frame zero. When you use all_mode=average, it blends adjacent frames. The new frame zero becomes 50/50 mix of (old frame 0,1). New frame one becomes a 50/50 mix of (old frame 1,2). You can test the filter by itself to verify this When you separate fields, you have 2 times the number of original frames. If you resize them to full height and treat them as frames, you still have 2 times the number of frames on that selection set If you started with a 240 frame clip , you should end up with a 600 frame clip after that telecine filter only with those settings. If you take every 5th frame from the telecine output ; 600/5 =120 . Separating fields give you 120*2 =240. Applying tblend=all_mode=average after gives you 240-1=239 frames. This is [D] . [C] is the other frame selection set; 600-120=480 Interleaving selection [C] with [D] should give you (600-120) + (240-1) = 719 The 716 might be from interleave dropping frames at the end when it hangs. Expect it to hang because there is no end-of-stream signal when used with select. https://ffmpeg.org/ffmpeg-filters.html#interleave_002c-ainterleave -- Sent from: http://www.ffmpeg-archive.org/ ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-user] concat demuxer filter_complex (fade)
‐‐‐ Original Message ‐‐‐ On Wednesday, April 15, 2020 10:44 AM, Paul B Mahol wrote: > On 4/15/20, Mark Filipak markfilipak.windows+ffm...@gmail.com wrote: > > > On 04/14/2020 02:40 PM, Ted Park wrote: > > On 04/14/2020 09:43 AM, atticus via ffmpeg-user wrote: > > > > > ‐‐‐ Original Message ‐‐‐ > > > On Tuesday, April 14, 2020 11:07 AM, Paul B Mahol one...@gmail.com > > > wrote: > > > > > > > > I thought about this yesterday and came up with something like this: > > > > > ffmpeg -i in.JPG -filter_complex "[0:0]loop=loop=-1:start=0:size=100 > > > > > [looped] ; [looped] trim=start=0:end=10 [trimmed] ; [trimmed] > > > > > fade=type=in:start_frame=0:duration=3:color=black [fadeIn]" -map > > > > > [fadeIn] > > > > > > > > -c:v h264 -r 60 out.mkv > > > > > or this > > > > > ffmpeg -loop 1 -i in.JPG -filter_complex "[0:0] trim=start=0:end=200 > > > > > [trimmed] ; [trimmed] > > > > > fade=type=in:start_frame=0:duration=3:color=black > > > > > [fadeIn]" -map [fadeIn] -c:v h264 out2.mkv > > > > > (I'd just have to add a concat filter to the filter chain and an audio > > > > > stream). I'm just not quite sure if there is a more smart way to do > > > > > this > > > > > (which for example would be a bit faster, since this is (in my opinion > > > > > a bit > > > > > slow for just duplicating a single frame). Well is there a smarter > > > > > and/or faster way? > > > > > And can you recommend which of these two commands above might be the > > > > > better one? > > > Hi, > > > The filter is not in the version of ffmpeg you are using, you will need to > > > get a more recent version for it to be there. Try downloading a recent > > > static build from the website or try compiling it yourself. (It's not in > > > any release version as far as I can tell) > > > Regards, > > > Ted Park > > > > May I suggest that the most expeditious thing to do is to simply address the > > original question? I would do it myself if I only had a brain... a heart... > > the nerve. > > Original approach is very fragile and extremely complicated to work with. Can you please elaborate what you mean by fragile and complicated? I for my part find it very easy to work with, since you easily can split up the big file into multiple smaller ones and concat them with the concat demuxer in the end. So you don't need to know whether and if there is, which stream comes after to create one single filter. In addition, in my option relying on a ffmpeg build that is not in any release version yes is not to be considered more stable. But please, I don't work that long with ffmpeg up to now and maybe I'm missing something. So please explain further what you mean ;) ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-user] Flutter FFmpeg
Am Mi., 15. Apr. 2020 um 20:14 Uhr schrieb cowlol : > ./ffmpeg -noautorotate -i input1.mp4 -movflags +faststart > -vcodec h264 -crf 18 -preset:v veryfast output.mp4 If you want a smaller filesize you have to change the crf value. If you chose a slower preset, you get either better quality (and can use a bigger crf) or smaller file size but the connection is not always obvious. Carl Eugen PS: Not everybody wants noautorotate... ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-user] Flutter FFmpeg
Thank you for the reply. I uploaded the compressed file to firebase, downloaded it and ran the following command on my mac. The original file is 20.6 and the compressed version (Done by the flutter picker) is 3.5mb. They both look pretty similar in terms of quality. When i run the following on the compressed file ./ffmpeg -noautorotate -i input1.mp4 -movflags +faststart -vcodec h264 -crf 18 -preset:v veryfast output.mp4 I get an output file that is 5.4mb 2mb larger. Here are the logs.. ffmpeg version N-97344-g51db0a472a-tessus https://evermeet.cx/ffmpeg/ Copyright (c) 2000-2020 the FFmpeg developers built with Apple clang version 11.0.0 (clang-1100.0.33.17) configuration: --cc=/usr/bin/clang --prefix=/opt/ffmpeg --extra-version=tessus --enable-avisynth --enable-fontconfig --enable-gpl --enable-libaom --enable-libass --enable-libbluray --enable-libdav1d --enable-libfreetype --enable-libgsm --enable-libmodplug --enable-libmp3lame --enable-libmysofa --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libopenh264 --enable-libopenjpeg --enable-libopus --enable-librubberband --enable-libshine --enable-libsnappy --enable-libsoxr --enable-libspeex --enable-libtheora --enable-libtwolame --enable-libvidstab --enable-libvmaf --enable-libvo-amrwbenc --enable-libvorbis --enable-libvpx --enable-libwavpack --enable-libwebp --enable-libx264 --enable-libx265 --enable-libxavs --enable-libxvid --enable-libzimg --enable-libzmq --enable-libzvbi --enable-version3 --pkg-config-flags=--static --disable-ffplay libavutil 56. 42.102 / 56. 42.102 libavcodec 58. 78.102 / 58. 78.102 libavformat58. 42.100 / 58. 42.100 libavdevice58. 9.103 / 58. 9.103 libavfilter 7. 77.101 / 7. 77.101 libswscale 5. 6.101 / 5. 6.101 libswresample 3. 6.100 / 3. 6.100 libpostproc55. 6.100 / 55. 6.100 Input #0, mov,mp4,m4a,3gp,3g2,mj2, from 'input1.mp4': Metadata: major_brand : qt minor_version : 0 compatible_brands: qt creation_time : 2020-04-15T17:26:57.00Z com.apple.quicktime.make: Apple com.apple.quicktime.model: iPhone SE com.apple.quicktime.software: 13.4 com.apple.quicktime.creationdate: 2020-04-15T18:23:58+0100 Duration: 00:00:10.54, start: 0.00, bitrate: 2674 kb/s Stream #0:0(und): Audio: aac (LC) (mp4a / 0x6134706D), 44100 Hz, mono, fltp, 93 kb/s (default) Metadata: creation_time : 2020-04-15T17:26:57.00Z handler_name: Core Media Audio Stream #0:1(und): Video: h264 (High) (avc1 / 0x31637661), yuv420p(tv, bt709), 1280x720, 2575 kb/s, 29.98 fps, 29.97 tbr, 600 tbn, 1200 tbc (default) Metadata: rotate : 90 creation_time : 2020-04-15T17:26:57.00Z handler_name: Core Media Video encoder : H.264 Side data: displaymatrix: rotation of -90.00 degrees Stream mapping: Stream #0:1 -> #0:0 (h264 (native) -> h264 (libx264)) Stream #0:0 -> #0:1 (aac (native) -> aac (native)) Press [q] to stop, [?] for help [libx264 @ 0x7fa4c3009400] using cpu capabilities: MMX2 SSE2Fast SSSE3 SSE4.2 AVX FMA3 BMI2 AVX2 [libx264 @ 0x7fa4c3009400] profile High, level 3.1, 4:2:0, 8-bit [libx264 @ 0x7fa4c3009400] 264 - core 160 r3000 33f9e14 - H.264/MPEG-4 AVC codec - Copyleft 2003-2020 - http://www.videolan.org/x264.html - options: cabac=1 ref=1 deblock=1:0:0 analyse=0x3:0x113 me=hex subme=2 psy=1 psy_rd=1.00:0.00 mixed_ref=0 me_range=16 chroma_me=1 trellis=0 8x8dct=1 cqm=0 deadzone=21,11 fast_pskip=1 chroma_qp_offset=0 threads=6 lookahead_threads=2 sliced_threads=0 nr=0 decimate=1 interlaced=0 bluray_compat=0 constrained_intra=0 bframes=3 b_pyramid=2 b_adapt=1 b_bias=0 direct=1 weightb=1 open_gop=0 weightp=1 keyint=250 keyint_min=25 scenecut=40 intra_refresh=0 rc_lookahead=10 rc=crf mbtree=1 crf=18.0 qcomp=0.60 qpmin=0 qpmax=69 qpstep=4 ip_ratio=1.40 aq=1:1.00 Output #0, mp4, to 'output.mp4': Metadata: major_brand : qt minor_version : 0 compatible_brands: qt com.apple.quicktime.creationdate: 2020-04-15T18:23:58+0100 com.apple.quicktime.make: Apple com.apple.quicktime.model: iPhone SE com.apple.quicktime.software: 13.4 encoder : Lavf58.42.100 Stream #0:0(und): Video: h264 (libx264) (avc1 / 0x31637661), yuv420p, 1280x720, q=-1--1, 29.97 fps, 30k tbn, 29.97 tbc (default) Metadata: encoder : Lavc58.78.102 libx264 creation_time : 2020-04-15T17:26:57.00Z handler_name: Core Media Video Side data: cpb: bitrate max/min/avg: 0/0/0 buffer size: 0 vbv_delay: N/A displaymatrix: rotation of -90.00 degrees Stream #0:1(und): Audio: aac (LC) (mp4a / 0x6134706D), 44100 Hz, mono, fltp, 69 kb/s (default) Metadata: creation_time : 2020-04-15T17:26:57.00Z handler_name: Core Media Audio encoder : Lavc58.78.102 aac frame= 36 fps=0.0 q=24.0 size= 256kB time=00:00:02.29 bitrate= 912.5kbits/frame= 66 fps= 64
Re: [FFmpeg-user] ffplay options for setting audio volume and locating player windows on the screen
On Tue, Apr 14, 2020 at 8:46 PM Michael Glenn Williams < michael.glenn.willi...@smartblondeco.com> wrote: > Hey Ted! > > Thank you for responding. Regarding -volume, i'm just putting 30 > without the %. > Regarding -top and -left, my version of ffplay (the latest MAC OSX download > from the website) > Doesn't seem to know these options. > Have you tried the good ole tried and true X / XQuartz -geometry options? I have not used OSX in a very long time (since the switch to Quartz), so there may be some differences between the old Xorg and XFree nomenclature, but I imagine those types of options would still be there for initial window placements and sizes. > I'm confused why generic options like these would vary from > build/distribution to build? > > > > > On Tue, Apr 14, 2020 at 5:31 PM Ted Park wrote: > > > Hi, > > > > > I tried the -w and -h options, which were not found by ffplay. The -x > and > > > -y options seem to be the options that set the width and heoght, not > the > > > location on the screen. Can anyone tell me how to position the windows > in > > > different places on the screen? > > > > > > I also tried the -volume option to start out at 30% volume, but that > > option > > > also wasn't found. > > I'm not sure if it's the same across all libraries/OS's, but -left and > > -top set the distance from left and top of screen respectively, for me. > And > > when you specify volume, try not including the "%" sign. > > > > Regards, > > Ted Park > > ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-user] Muxing multiple files and concatenating those outputs
On Wed, Apr 15, 2020 at 5:43 AM Ted Park wrote: > Hi, > > > Yeah, when we record the calls, the directory structure preceeding those > > names is /MM/DD/HH and the filenames are MIN_SEC_MSEC.codec(side) > > > I mean some phones specifically put g729a, and I assume it's the same for > g729b. So I started imagining files ending all over the place g729aa, > g729ba, etc. > Ahhh, I see what you are referring to now :) g729b is just an annex to a that only allows for more header enhancements - the raw codec audio is the same though, so we just write g729 for both. > > I was really looking for just syntax to group commands, so I could use > the > > merge filter output as direct input for the concatenation. > > If you mean the afifo I inserted, you could probably get rid of those if > the machine's fast enough, or the calls short enough. Or a different > approach might be padding each file in the same call or using the cue > filter, and mixing them all. But grouping commands isn't really a thing > that ffmpeg does, except for things like image sequences, preparing > segmented delivery media. > I see the restraints I am up against now and I think what I am trying is just not possible, so I will take the path of least resistance and keep it multi-lined. To clarify, the reason I need to merge channels prior is that the 2 sides do not always exactly line up (there almost always is a fraction of a second less on one side of the packet stream) and it causes conversation sync/skew if they are concatenated either prior or simultaneously with the merging. > > The man page and web searches I tried came up empty, but I figured > someone may know some magic sauce I could not find. > > > Maybe you've been looking for the wrong terms? It sounds to me like what > you are looking for is closer to shell features, like parameter > substitution, filename generation, etc. > Yeah, I am currently using mktemp, etc.. I was just hoping to simplify it because IO is the limiting factor when it comes to call recording, so minimizing it is always a prioriry for me. > If portability is not an issue, some shells have more features than > others, though there's a different learning curve to each one. > > Or there's always xargs. Yeah probably xargs. I can't really tell but the > associated filenames seem pretty much arbitrary. Do you parse the date/time > in the filename to find which ones to put together? Or is there a call log > to reference? > > Regards, > Ted Park > Thanks again for all of your input Ted. -Greg ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-user] Encoding Settings
This can now be closed as I have sorted everything and have exactly what I needed. Thank You -- Sent from: http://www.ffmpeg-archive.org/ ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-user] concat demuxer filter_complex (fade)
On 4/15/20, Mark Filipak wrote: > On 04/14/2020 02:40 PM, Ted Park wrote: >> Hi, >> >> The filter is not in the version of ffmpeg you are using, you will need to >> get a more recent version for it to be there. Try downloading a recent >> static build from the website or try compiling it yourself. (It's not in >> any release version as far as I can tell) >> >> Regards, >> Ted Park > > May I suggest that the most expeditious thing to do is to simply address the > original question? I > would do it myself if I only had a brain... a heart... the nerve. > Original approach is very fragile and extremely complicated to work with. > > On 04/14/2020 09:43 AM, atticus via ffmpeg-user wrote: > > ‐‐‐ Original Message ‐‐‐ > > On Tuesday, April 14, 2020 11:07 AM, Paul B Mahol > wrote: > >>> I thought about this yesterday and came up with something like this: > >>> ffmpeg -i in.JPG -filter_complex "[0:0]loop=loop=-1:start=0:size=100 > >>> [looped] ; [looped] trim=start=0:end=10 [trimmed] ; [trimmed] > >>> fade=type=in:start_frame=0:duration=3:color=black [fadeIn]" -map > [fadeIn] > >>> -c:v h264 -r 60 out.mkv > >>> or this > >>> ffmpeg -loop 1 -i in.JPG -filter_complex "[0:0] trim=start=0:end=200 > >>> [trimmed] ; [trimmed] fade=type=in:start_frame=0:duration=3:color=black > >>> [fadeIn]" -map [fadeIn] -c:v h264 out2.mkv > >>> (I'd just have to add a concat filter to the filter chain and an audio > >>> stream). I'm just not quite sure if there is a more smart way to do > this > >>> (which for example would be a bit faster, since this is (in my opinion > a bit > >>> slow for just duplicating a single frame). Well is there a smarter > and/or > >>> faster way? > >>> And can you recommend which of these two commands above might be the > better > >>> one? > ___ > ffmpeg-user mailing list > ffmpeg-user@ffmpeg.org > https://ffmpeg.org/mailman/listinfo/ffmpeg-user > > To unsubscribe, visit link above, or email > ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe". ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-user] Muxing multiple files and concatenating those outputs
Hi, > Yeah, when we record the calls, the directory structure preceeding those > names is /MM/DD/HH and the filenames are MIN_SEC_MSEC.codec(side) I mean some phones specifically put g729a, and I assume it's the same for g729b. So I started imagining files ending all over the place g729aa, g729ba, etc. > I was really looking for just syntax to group commands, so I could use the > merge filter output as direct input for the concatenation. If you mean the afifo I inserted, you could probably get rid of those if the machine's fast enough, or the calls short enough. Or a different approach might be padding each file in the same call or using the cue filter, and mixing them all. But grouping commands isn't really a thing that ffmpeg does, except for things like image sequences, preparing segmented delivery media. > The man page and web searches I tried came up empty, but I figured someone > may know some magic sauce I could not find. Maybe you've been looking for the wrong terms? It sounds to me like what you are looking for is closer to shell features, like parameter substitution, filename generation, etc. If portability is not an issue, some shells have more features than others, though there's a different learning curve to each one. Or there's always xargs. Yeah probably xargs. I can't really tell but the associated filenames seem pretty much arbitrary. Do you parse the date/time in the filename to find which ones to put together? Or is there a call log to reference? Regards, Ted Park ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-user] Decombing via screening - 'tblend' bug (?)
On 04/15/2020 06:16 AM, Paul B Mahol wrote: On 4/15/20, Mark Filipak wrote: On 04/15/2020 05:38 AM, Paul B Mahol wrote: On 4/15/20, Mark Filipak wrote: On 04/15/2020 05:00 AM, Paul B Mahol wrote: On 4/15/20, Mark Filipak wrote: On 04/15/2020 04:25 AM, Paul B Mahol wrote: On 4/15/20, Mark Filipak wrote: Request for bug confirmation is included. Correction: I had datascope in #2 just 1 time. I don't know how it got into the original post twice. Sorry. - M. On 04/14/2020 10:24 AM, Paul B Mahol wrote: On 4/14/20, Mark Filipak wrote: In the command line below, I will eventually insert a screen maker --+ ¦ separatefields,scale=height=2*in_h:sws_flags=neighbor,--HERE--,tblend=normal[D] but first, I need to persuade 'tblend' to accept the 2 frames made by 'separatefields' (and scaled by 'scale'). But as you can see in the next section (extracted from the log), though the 'height=2*in_h' directive works (i.e., the 1920x540 frames from 'separatefields' are scaled to 1920x1080), ffmpeg is not happy. Is it because the SARs don't match? How can I overcome that? Thanks! Insert setsar=1 after scale? Thank you, Paul. It works. I apologize for not finding that solution on my own ...how embarrassing. Next problem: 'tblend' appears to be malfunctioning. "The tblend (time blend) filter takes two consecutive frames from one single stream, and outputs the result obtained by blending the new frame on top of the old frame." #1 - This works as expected: ffmpeg -i IN -filter_complex "telecine=pattern=5, split[A][B], [A]select='not(eq(mod(n+1\,5)\,3))'[C], [B]select='eq(mod(n+1\,5)\,3)', datascope=size=1920x1080:x=45:y=340:mode=color2[D], [C][D]interleave" OUT #2 - This malfunctions: ffmpeg -i IN -filter_complex "telecine=pattern=5, split[A][B], [A]select='not(eq(mod(n+1\,5)\,3))'[C], [B]select='eq(mod(n+1\,5)\,3)', separatefields, scale=height=2*in_h:sws_flags=neighbor, setsar=1, tblend, datascope=size=1920x1080:x=45:y=340:mode=color2, [C][D]interleave" OUT #1 brings up datascope in frames (n, zero-based): 2 7 12 17 etc., as expected. #2 brings up datascope in frames (n, zero-based): 4 7 10 13 etc. Another clue is: #1 OUT has 598 frames. #2 OUT has 716 frames. Would someone who has the latest nightly build kindly confirm this. I can supply the test video -- contact me off-list. You still can not read documentation? Why? interleave filter simply picks frames from several streams. This is obviously explained in documentation. tblend filter blend successive frames, this is obviously different from interlace filter as that will not drop any frame like interlace does. Look at my command lines again. Here, I'll give you filtergraphs: #1 telecine -> split -> select (n+1%5!=3) > interleave -> select (n+1%5==3) > 598 frames (correct) #2 telecine -> split -> select (n+1%5!=3) > interleave -> select (n+1%5==3) -> separatefields...tblend -> 716 frames (malfunction) 598*4/5 = 478 frames (upper 'select') 598*1/5 = 119 frames (lower 'select') 119*2 = 238 frames (output of 'separatefields') 238/2 = 119 frames (what output of 'tblend' is supposed to be) 238 frames (what output of 'tblend' actually is) 478+238 = 716 frames (output of 'interleave' showing that 'tblend' malfunctions) tblend works fine, you obviously do not know what you doing. Given frames: A B C D E F G: tblend gives this output: AB BC CD DE EF FG, so original number of frames in input minus one. Are you even looking at the command lines? #2 telecine -> split -> select (n+1%5!=3) > interleave -> select (n+1%5==3) -> separatefields...tblend -> 'telecine' input (IN) is A B C... 'telecine' output is A A AB B B ... (modulo 5) ...frame 2 is combed upper 'select' output is A A __ B B (n+1%5!=3) ...progressive frames lower 'select' output is _ _ AB _ _ (n+1%5==3) ...combed frame 'separatefields' output isAB AB ...2 frames (1/2 height) 'scale=height=2*in_h' output is AB AB ...2 frames (full height) 'tblend' output is supposed to be AB...1 frame OUT is supposed to be A A AB B B...598 frames OUT isA A AB AB B B ...716 frames Conclusion: 'tblend' is outputting 2 frames instead of 1 frame. And that is completely correct for filter to do. If you not happy with that, write own code instead. From here: https://ffmpeg.org/ffmpeg-filters.html#blend "The tblend (time blend) filter takes two consecutive frames from one single stream, and outputs the result obtained by blending the new frame on top of the old frame." Seems to me that 'tblend' is supposed to output 1 frame. Nope. Enjoy living in your own world. This: ffmpeg -i IN
Re: [FFmpeg-user] Decombing via screening - 'tblend' bug (?)
On 4/15/20, Mark Filipak wrote: > On 04/15/2020 05:38 AM, Paul B Mahol wrote: >> On 4/15/20, Mark Filipak wrote: >>> On 04/15/2020 05:00 AM, Paul B Mahol wrote: On 4/15/20, Mark Filipak wrote: > On 04/15/2020 04:25 AM, Paul B Mahol wrote: >> On 4/15/20, Mark Filipak wrote: >>> Request for bug confirmation is included. >>> >>> Correction: I had datascope in #2 just 1 time. I don't know how it >>> got >>> into >>> the original post twice. >>> Sorry. - M. >>> >>> On 04/14/2020 10:24 AM, Paul B Mahol wrote: On 4/14/20, Mark Filipak wrote: > In the command line below, > I will eventually insert a screen maker --+ > ¦ > separatefields,scale=height=2*in_h:sws_flags=neighbor,--HERE--,tblend=normal[D] > > but first, I need to persuade 'tblend' to accept the 2 frames made > by > 'separatefields' (and scaled > by 'scale'). > But as you can see in the next section (extracted from the log), > though > the > 'height=2*in_h' > directive works (i.e., the 1920x540 frames from 'separatefields' > are > scaled > to 1920x1080), ffmpeg is > not happy. Is it because the SARs don't match? How can I overcome > that? > Thanks! Insert setsar=1 after scale? >>> >>> Thank you, Paul. It works. I apologize for not finding that solution >>> on >>> my >>> own ...how embarrassing. >>> >>> Next problem: 'tblend' appears to be malfunctioning. >>> >>> "The tblend (time blend) filter takes two consecutive frames from one >>> single >>> stream, and outputs the >>> result obtained by blending the new frame on top of the old frame." >>> >>> #1 - This works as expected: >>> ffmpeg -i IN -filter_complex "telecine=pattern=5, split[A][B], >>> [A]select='not(eq(mod(n+1\,5)\,3))'[C], >>> [B]select='eq(mod(n+1\,5)\,3)', >>> datascope=size=1920x1080:x=45:y=340:mode=color2[D], [C][D]interleave" >>> OUT >>> >>> #2 - This malfunctions: >>> ffmpeg -i IN -filter_complex "telecine=pattern=5, split[A][B], >>> [A]select='not(eq(mod(n+1\,5)\,3))'[C], >>> [B]select='eq(mod(n+1\,5)\,3)', >>> separatefields, >>> scale=height=2*in_h:sws_flags=neighbor, setsar=1, tblend, >>> datascope=size=1920x1080:x=45:y=340:mode=color2, [C][D]interleave" >>> OUT >>> >>> #1 brings up datascope in frames (n, zero-based): 2 7 12 17 etc., as >>> expected. >>> #2 brings up datascope in frames (n, zero-based): 4 7 10 13 etc. >>> >>> Another clue is: >>> #1 OUT has 598 frames. >>> #2 OUT has 716 frames. >>> >>> Would someone who has the latest nightly build kindly confirm this. I >>> can >>> supply the test video -- >>> contact me off-list. >> >> You still can not read documentation? Why? >> >> interleave filter simply picks frames from several streams. >> This is obviously explained in documentation. >> tblend filter blend successive frames, this is obviously different >> from interlace filter as that will not drop any frame like interlace >> does. > > Look at my command lines again. Here, I'll give you filtergraphs: > > #1 > telecine -> split -> select (n+1%5!=3) > > interleave > -> select (n+1%5==3) > > 598 frames (correct) > > #2 > telecine -> split -> select (n+1%5!=3) > > interleave > -> select (n+1%5==3) -> separatefields...tblend -> > 716 frames (malfunction) > > 598*4/5 = 478 frames (upper 'select') > 598*1/5 = 119 frames (lower 'select') > 119*2 = 238 frames (output of 'separatefields') > 238/2 = 119 frames (what output of 'tblend' is supposed to be) > 238 frames (what output of 'tblend' actually is) > 478+238 = 716 frames (output of 'interleave' showing that 'tblend' > malfunctions) > tblend works fine, you obviously do not know what you doing. Given frames: A B C D E F G: tblend gives this output: AB BC CD DE EF FG, so original number of frames in input minus one. >>> >>> Are you even looking at the command lines? >>> >>> #2 >>> telecine -> split -> select (n+1%5!=3) > >>> interleave >>> -> select (n+1%5==3) -> separatefields...tblend -> >>> >>> 'telecine' input (IN) is A B C... >>> 'telecine' output is A A AB B B ... (modulo 5) ...frame 2 is >>> combed >>> upper 'select' output is A A __ B B (n+1%5!=3) ...progressive >>> frames >>> lower 'select' output is _ _ AB _ _ (n+1%5==3) ...combed frame >>> 'separatefields' output is
Re: [FFmpeg-user] Decombing via screening - 'tblend' bug (?)
This: ffmpeg -i IN -filter_complex "telecine=pattern=5,split[A][B],[A]select='not(eq(mod(n+1\,5)\,3))'[C],[B]select='eq(mod(n+1\,5)\,3)',split[E][F],[E][F]blend[D],[C][D]interleave" OUT outputs 598 frames This: ffmpeg -i IN -filter_complex "telecine=pattern=5,split[A][B],[A]select='not(eq(mod(n+1\,5)\,3))'[C],[B]select='eq(mod(n+1\,5)\,3)',separatefields,scale=height=2*in_h:sws_flags=neighbor,setsar=1,tblend[D],[C][D]interleave" OUT outputs 716 frames Now, does that look consistent to you? But of course, since I can't read, and I'm always wrong... ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-user] Decombing via screening - 'tblend' bug (?)
On 04/15/2020 05:38 AM, Paul B Mahol wrote: On 4/15/20, Mark Filipak wrote: On 04/15/2020 05:00 AM, Paul B Mahol wrote: On 4/15/20, Mark Filipak wrote: On 04/15/2020 04:25 AM, Paul B Mahol wrote: On 4/15/20, Mark Filipak wrote: Request for bug confirmation is included. Correction: I had datascope in #2 just 1 time. I don't know how it got into the original post twice. Sorry. - M. On 04/14/2020 10:24 AM, Paul B Mahol wrote: On 4/14/20, Mark Filipak wrote: In the command line below, I will eventually insert a screen maker --+ ¦ separatefields,scale=height=2*in_h:sws_flags=neighbor,--HERE--,tblend=normal[D] but first, I need to persuade 'tblend' to accept the 2 frames made by 'separatefields' (and scaled by 'scale'). But as you can see in the next section (extracted from the log), though the 'height=2*in_h' directive works (i.e., the 1920x540 frames from 'separatefields' are scaled to 1920x1080), ffmpeg is not happy. Is it because the SARs don't match? How can I overcome that? Thanks! Insert setsar=1 after scale? Thank you, Paul. It works. I apologize for not finding that solution on my own ...how embarrassing. Next problem: 'tblend' appears to be malfunctioning. "The tblend (time blend) filter takes two consecutive frames from one single stream, and outputs the result obtained by blending the new frame on top of the old frame." #1 - This works as expected: ffmpeg -i IN -filter_complex "telecine=pattern=5, split[A][B], [A]select='not(eq(mod(n+1\,5)\,3))'[C], [B]select='eq(mod(n+1\,5)\,3)', datascope=size=1920x1080:x=45:y=340:mode=color2[D], [C][D]interleave" OUT #2 - This malfunctions: ffmpeg -i IN -filter_complex "telecine=pattern=5, split[A][B], [A]select='not(eq(mod(n+1\,5)\,3))'[C], [B]select='eq(mod(n+1\,5)\,3)', separatefields, scale=height=2*in_h:sws_flags=neighbor, setsar=1, tblend, datascope=size=1920x1080:x=45:y=340:mode=color2, [C][D]interleave" OUT #1 brings up datascope in frames (n, zero-based): 2 7 12 17 etc., as expected. #2 brings up datascope in frames (n, zero-based): 4 7 10 13 etc. Another clue is: #1 OUT has 598 frames. #2 OUT has 716 frames. Would someone who has the latest nightly build kindly confirm this. I can supply the test video -- contact me off-list. You still can not read documentation? Why? interleave filter simply picks frames from several streams. This is obviously explained in documentation. tblend filter blend successive frames, this is obviously different from interlace filter as that will not drop any frame like interlace does. Look at my command lines again. Here, I'll give you filtergraphs: #1 telecine -> split -> select (n+1%5!=3) > interleave -> select (n+1%5==3) > 598 frames (correct) #2 telecine -> split -> select (n+1%5!=3) > interleave -> select (n+1%5==3) -> separatefields...tblend -> 716 frames (malfunction) 598*4/5 = 478 frames (upper 'select') 598*1/5 = 119 frames (lower 'select') 119*2 = 238 frames (output of 'separatefields') 238/2 = 119 frames (what output of 'tblend' is supposed to be) 238 frames (what output of 'tblend' actually is) 478+238 = 716 frames (output of 'interleave' showing that 'tblend' malfunctions) tblend works fine, you obviously do not know what you doing. Given frames: A B C D E F G: tblend gives this output: AB BC CD DE EF FG, so original number of frames in input minus one. Are you even looking at the command lines? #2 telecine -> split -> select (n+1%5!=3) > interleave -> select (n+1%5==3) -> separatefields...tblend -> 'telecine' input (IN) is A B C... 'telecine' output is A A AB B B ... (modulo 5) ...frame 2 is combed upper 'select' output is A A __ B B (n+1%5!=3) ...progressive frames lower 'select' output is _ _ AB _ _ (n+1%5==3) ...combed frame 'separatefields' output isAB AB ...2 frames (1/2 height) 'scale=height=2*in_h' output is AB AB ...2 frames (full height) 'tblend' output is supposed to be AB...1 frame OUT is supposed to be A A AB B B...598 frames OUT isA A AB AB B B ...716 frames Conclusion: 'tblend' is outputting 2 frames instead of 1 frame. And that is completely correct for filter to do. If you not happy with that, write own code instead. From here: https://ffmpeg.org/ffmpeg-filters.html#blend "The tblend (time blend) filter takes two consecutive frames from one single stream, and outputs the result obtained by blending the new frame on top of the old frame." Seems to me that 'tblend' is supposed to output 1 frame. ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To
Re: [FFmpeg-user] Decombing via screening - 'tblend' bug (?)
On 4/15/20, Mark Filipak wrote: > On 04/15/2020 05:00 AM, Paul B Mahol wrote: >> On 4/15/20, Mark Filipak wrote: >>> On 04/15/2020 04:25 AM, Paul B Mahol wrote: On 4/15/20, Mark Filipak wrote: > Request for bug confirmation is included. > > Correction: I had datascope in #2 just 1 time. I don't know how it got > into > the original post twice. > Sorry. - M. > > On 04/14/2020 10:24 AM, Paul B Mahol wrote: >> On 4/14/20, Mark Filipak wrote: >>> In the command line below, >>>I will eventually insert a screen maker --+ >>> ¦ >>> separatefields,scale=height=2*in_h:sws_flags=neighbor,--HERE--,tblend=normal[D] >>> >>> but first, I need to persuade 'tblend' to accept the 2 frames made by >>> 'separatefields' (and scaled >>> by 'scale'). >>> But as you can see in the next section (extracted from the log), >>> though >>> the >>> 'height=2*in_h' >>> directive works (i.e., the 1920x540 frames from 'separatefields' are >>> scaled >>> to 1920x1080), ffmpeg is >>> not happy. Is it because the SARs don't match? How can I overcome >>> that? >>> Thanks! >> >> Insert setsar=1 after scale? > > Thank you, Paul. It works. I apologize for not finding that solution on > my > own ...how embarrassing. > > Next problem: 'tblend' appears to be malfunctioning. > > "The tblend (time blend) filter takes two consecutive frames from one > single > stream, and outputs the > result obtained by blending the new frame on top of the old frame." > > #1 - This works as expected: > ffmpeg -i IN -filter_complex "telecine=pattern=5, split[A][B], > [A]select='not(eq(mod(n+1\,5)\,3))'[C], [B]select='eq(mod(n+1\,5)\,3)', > datascope=size=1920x1080:x=45:y=340:mode=color2[D], [C][D]interleave" > OUT > > #2 - This malfunctions: > ffmpeg -i IN -filter_complex "telecine=pattern=5, split[A][B], > [A]select='not(eq(mod(n+1\,5)\,3))'[C], [B]select='eq(mod(n+1\,5)\,3)', > separatefields, > scale=height=2*in_h:sws_flags=neighbor, setsar=1, tblend, > datascope=size=1920x1080:x=45:y=340:mode=color2, [C][D]interleave" OUT > > #1 brings up datascope in frames (n, zero-based): 2 7 12 17 etc., as > expected. > #2 brings up datascope in frames (n, zero-based): 4 7 10 13 etc. > > Another clue is: > #1 OUT has 598 frames. > #2 OUT has 716 frames. > > Would someone who has the latest nightly build kindly confirm this. I > can > supply the test video -- > contact me off-list. You still can not read documentation? Why? interleave filter simply picks frames from several streams. This is obviously explained in documentation. tblend filter blend successive frames, this is obviously different from interlace filter as that will not drop any frame like interlace does. >>> >>> Look at my command lines again. Here, I'll give you filtergraphs: >>> >>> #1 >>> telecine -> split -> select (n+1%5!=3) > >>> interleave >>> -> select (n+1%5==3) > >>> 598 frames (correct) >>> >>> #2 >>> telecine -> split -> select (n+1%5!=3) > >>> interleave >>> -> select (n+1%5==3) -> separatefields...tblend -> >>> 716 frames (malfunction) >>> >>> 598*4/5 = 478 frames (upper 'select') >>> 598*1/5 = 119 frames (lower 'select') >>> 119*2 = 238 frames (output of 'separatefields') >>> 238/2 = 119 frames (what output of 'tblend' is supposed to be) >>> 238 frames (what output of 'tblend' actually is) >>> 478+238 = 716 frames (output of 'interleave' showing that 'tblend' >>> malfunctions) >>> >> >> tblend works fine, you obviously do not know what you doing. >> >> Given frames: A B C D E F G: >> >> tblend gives this output: AB BC CD DE EF FG, so original number of >> frames in input minus one. > > Are you even looking at the command lines? > > #2 > telecine -> split -> select (n+1%5!=3) > > interleave >-> select (n+1%5==3) -> separatefields...tblend -> > > 'telecine' input (IN) is A B C... > 'telecine' output is A A AB B B ... (modulo 5) ...frame 2 is combed > upper 'select' output is A A __ B B (n+1%5!=3) ...progressive > frames > lower 'select' output is _ _ AB _ _ (n+1%5==3) ...combed frame > 'separatefields' output isAB AB ...2 frames (1/2 > height) > 'scale=height=2*in_h' output is AB AB ...2 frames (full > height) > 'tblend' output is supposed to be AB...1 frame > OUT is supposed to be A A AB B B...598 frames > OUT isA A AB AB B B ...716 frames > Conclusion: 'tblend' is outputting 2 frames instead
Re: [FFmpeg-user] Decombing via screening - 'tblend' bug (?)
On 04/15/2020 05:00 AM, Paul B Mahol wrote: On 4/15/20, Mark Filipak wrote: On 04/15/2020 04:25 AM, Paul B Mahol wrote: On 4/15/20, Mark Filipak wrote: Request for bug confirmation is included. Correction: I had datascope in #2 just 1 time. I don't know how it got into the original post twice. Sorry. - M. On 04/14/2020 10:24 AM, Paul B Mahol wrote: On 4/14/20, Mark Filipak wrote: In the command line below, I will eventually insert a screen maker --+ ¦ separatefields,scale=height=2*in_h:sws_flags=neighbor,--HERE--,tblend=normal[D] but first, I need to persuade 'tblend' to accept the 2 frames made by 'separatefields' (and scaled by 'scale'). But as you can see in the next section (extracted from the log), though the 'height=2*in_h' directive works (i.e., the 1920x540 frames from 'separatefields' are scaled to 1920x1080), ffmpeg is not happy. Is it because the SARs don't match? How can I overcome that? Thanks! Insert setsar=1 after scale? Thank you, Paul. It works. I apologize for not finding that solution on my own ...how embarrassing. Next problem: 'tblend' appears to be malfunctioning. "The tblend (time blend) filter takes two consecutive frames from one single stream, and outputs the result obtained by blending the new frame on top of the old frame." #1 - This works as expected: ffmpeg -i IN -filter_complex "telecine=pattern=5, split[A][B], [A]select='not(eq(mod(n+1\,5)\,3))'[C], [B]select='eq(mod(n+1\,5)\,3)', datascope=size=1920x1080:x=45:y=340:mode=color2[D], [C][D]interleave" OUT #2 - This malfunctions: ffmpeg -i IN -filter_complex "telecine=pattern=5, split[A][B], [A]select='not(eq(mod(n+1\,5)\,3))'[C], [B]select='eq(mod(n+1\,5)\,3)', separatefields, scale=height=2*in_h:sws_flags=neighbor, setsar=1, tblend, datascope=size=1920x1080:x=45:y=340:mode=color2, [C][D]interleave" OUT #1 brings up datascope in frames (n, zero-based): 2 7 12 17 etc., as expected. #2 brings up datascope in frames (n, zero-based): 4 7 10 13 etc. Another clue is: #1 OUT has 598 frames. #2 OUT has 716 frames. Would someone who has the latest nightly build kindly confirm this. I can supply the test video -- contact me off-list. You still can not read documentation? Why? interleave filter simply picks frames from several streams. This is obviously explained in documentation. tblend filter blend successive frames, this is obviously different from interlace filter as that will not drop any frame like interlace does. Look at my command lines again. Here, I'll give you filtergraphs: #1 telecine -> split -> select (n+1%5!=3) > interleave -> select (n+1%5==3) > 598 frames (correct) #2 telecine -> split -> select (n+1%5!=3) > interleave -> select (n+1%5==3) -> separatefields...tblend -> 716 frames (malfunction) 598*4/5 = 478 frames (upper 'select') 598*1/5 = 119 frames (lower 'select') 119*2 = 238 frames (output of 'separatefields') 238/2 = 119 frames (what output of 'tblend' is supposed to be) 238 frames (what output of 'tblend' actually is) 478+238 = 716 frames (output of 'interleave' showing that 'tblend' malfunctions) tblend works fine, you obviously do not know what you doing. Given frames: A B C D E F G: tblend gives this output: AB BC CD DE EF FG, so original number of frames in input minus one. Are you even looking at the command lines? #2 telecine -> split -> select (n+1%5!=3) > interleave -> select (n+1%5==3) -> separatefields...tblend -> 'telecine' input (IN) is A B C... 'telecine' output is A A AB B B ... (modulo 5) ...frame 2 is combed upper 'select' output is A A __ B B (n+1%5!=3) ...progressive frames lower 'select' output is _ _ AB _ _ (n+1%5==3) ...combed frame 'separatefields' output isAB AB ...2 frames (1/2 height) 'scale=height=2*in_h' output is AB AB ...2 frames (full height) 'tblend' output is supposed to be AB...1 frame OUT is supposed to be A A AB B B...598 frames OUT isA A AB AB B B ...716 frames Conclusion: 'tblend' is outputting 2 frames instead of 1 frame. Of course, since I can't read and I'm always wrong... ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-user] Decombing via screening - 'tblend' bug (?)
On 4/15/20, Mark Filipak wrote: > On 04/15/2020 04:25 AM, Paul B Mahol wrote: >> On 4/15/20, Mark Filipak wrote: >>> Request for bug confirmation is included. >>> >>> Correction: I had datascope in #2 just 1 time. I don't know how it got >>> into >>> the original post twice. >>> Sorry. - M. >>> >>> On 04/14/2020 10:24 AM, Paul B Mahol wrote: On 4/14/20, Mark Filipak wrote: > In the command line below, > I will eventually insert a screen maker --+ > ¦ > separatefields,scale=height=2*in_h:sws_flags=neighbor,--HERE--,tblend=normal[D] > > but first, I need to persuade 'tblend' to accept the 2 frames made by > 'separatefields' (and scaled > by 'scale'). > But as you can see in the next section (extracted from the log), though > the > 'height=2*in_h' > directive works (i.e., the 1920x540 frames from 'separatefields' are > scaled > to 1920x1080), ffmpeg is > not happy. Is it because the SARs don't match? How can I overcome that? > Thanks! Insert setsar=1 after scale? >>> >>> Thank you, Paul. It works. I apologize for not finding that solution on >>> my >>> own ...how embarrassing. >>> >>> Next problem: 'tblend' appears to be malfunctioning. >>> >>> "The tblend (time blend) filter takes two consecutive frames from one >>> single >>> stream, and outputs the >>> result obtained by blending the new frame on top of the old frame." >>> >>> #1 - This works as expected: >>> ffmpeg -i IN -filter_complex "telecine=pattern=5, split[A][B], >>> [A]select='not(eq(mod(n+1\,5)\,3))'[C], [B]select='eq(mod(n+1\,5)\,3)', >>> datascope=size=1920x1080:x=45:y=340:mode=color2[D], [C][D]interleave" OUT >>> >>> #2 - This malfunctions: >>> ffmpeg -i IN -filter_complex "telecine=pattern=5, split[A][B], >>> [A]select='not(eq(mod(n+1\,5)\,3))'[C], [B]select='eq(mod(n+1\,5)\,3)', >>> separatefields, >>> scale=height=2*in_h:sws_flags=neighbor, setsar=1, tblend, >>> datascope=size=1920x1080:x=45:y=340:mode=color2, [C][D]interleave" OUT >>> >>> #1 brings up datascope in frames (n, zero-based): 2 7 12 17 etc., as >>> expected. >>> #2 brings up datascope in frames (n, zero-based): 4 7 10 13 etc. >>> >>> Another clue is: >>> #1 OUT has 598 frames. >>> #2 OUT has 716 frames. >>> >>> Would someone who has the latest nightly build kindly confirm this. I can >>> supply the test video -- >>> contact me off-list. >> >> You still can not read documentation? Why? >> >> interleave filter simply picks frames from several streams. >> This is obviously explained in documentation. >> tblend filter blend successive frames, this is obviously different >> from interlace filter as that will not drop any frame like interlace >> does. > > Look at my command lines again. Here, I'll give you filtergraphs: > > #1 > telecine -> split -> select (n+1%5!=3) > > interleave >-> select (n+1%5==3) > > 598 frames (correct) > > #2 > telecine -> split -> select (n+1%5!=3) > > interleave >-> select (n+1%5==3) -> separatefields...tblend -> > 716 frames (malfunction) > > 598*4/5 = 478 frames (upper 'select') > 598*1/5 = 119 frames (lower 'select') > 119*2 = 238 frames (output of 'separatefields') > 238/2 = 119 frames (what output of 'tblend' is supposed to be) > 238 frames (what output of 'tblend' actually is) > 478+238 = 716 frames (output of 'interleave' showing that 'tblend' > malfunctions) > tblend works fine, you obviously do not know what you doing. Given frames: A B C D E F G: tblend gives this output: AB BC CD DE EF FG, so original number of frames in input minus one. ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-user] Decombing via screening - 'tblend' bug (?)
On 04/15/2020 04:25 AM, Paul B Mahol wrote: On 4/15/20, Mark Filipak wrote: Request for bug confirmation is included. Correction: I had datascope in #2 just 1 time. I don't know how it got into the original post twice. Sorry. - M. On 04/14/2020 10:24 AM, Paul B Mahol wrote: On 4/14/20, Mark Filipak wrote: In the command line below, I will eventually insert a screen maker --+ ¦ separatefields,scale=height=2*in_h:sws_flags=neighbor,--HERE--,tblend=normal[D] but first, I need to persuade 'tblend' to accept the 2 frames made by 'separatefields' (and scaled by 'scale'). But as you can see in the next section (extracted from the log), though the 'height=2*in_h' directive works (i.e., the 1920x540 frames from 'separatefields' are scaled to 1920x1080), ffmpeg is not happy. Is it because the SARs don't match? How can I overcome that? Thanks! Insert setsar=1 after scale? Thank you, Paul. It works. I apologize for not finding that solution on my own ...how embarrassing. Next problem: 'tblend' appears to be malfunctioning. "The tblend (time blend) filter takes two consecutive frames from one single stream, and outputs the result obtained by blending the new frame on top of the old frame." #1 - This works as expected: ffmpeg -i IN -filter_complex "telecine=pattern=5, split[A][B], [A]select='not(eq(mod(n+1\,5)\,3))'[C], [B]select='eq(mod(n+1\,5)\,3)', datascope=size=1920x1080:x=45:y=340:mode=color2[D], [C][D]interleave" OUT #2 - This malfunctions: ffmpeg -i IN -filter_complex "telecine=pattern=5, split[A][B], [A]select='not(eq(mod(n+1\,5)\,3))'[C], [B]select='eq(mod(n+1\,5)\,3)', separatefields, scale=height=2*in_h:sws_flags=neighbor, setsar=1, tblend, datascope=size=1920x1080:x=45:y=340:mode=color2, [C][D]interleave" OUT #1 brings up datascope in frames (n, zero-based): 2 7 12 17 etc., as expected. #2 brings up datascope in frames (n, zero-based): 4 7 10 13 etc. Another clue is: #1 OUT has 598 frames. #2 OUT has 716 frames. Would someone who has the latest nightly build kindly confirm this. I can supply the test video -- contact me off-list. You still can not read documentation? Why? interleave filter simply picks frames from several streams. This is obviously explained in documentation. tblend filter blend successive frames, this is obviously different from interlace filter as that will not drop any frame like interlace does. Look at my command lines again. Here, I'll give you filtergraphs: #1 telecine -> split -> select (n+1%5!=3) > interleave -> select (n+1%5==3) > 598 frames (correct) #2 telecine -> split -> select (n+1%5!=3) > interleave -> select (n+1%5==3) -> separatefields...tblend -> 716 frames (malfunction) 598*4/5 = 478 frames (upper 'select') 598*1/5 = 119 frames (lower 'select') 119*2 = 238 frames (output of 'separatefields') 238/2 = 119 frames (what output of 'tblend' is supposed to be) 238 frames (what output of 'tblend' actually is) 478+238 = 716 frames (output of 'interleave' showing that 'tblend' malfunctions) ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-user] Decombing via screening - 'tblend' bug (?)
On 4/15/20, Mark Filipak wrote: > Request for bug confirmation is included. > > Correction: I had datascope in #2 just 1 time. I don't know how it got into > the original post twice. > Sorry. - M. > > On 04/14/2020 10:24 AM, Paul B Mahol wrote: >> On 4/14/20, Mark Filipak wrote: >>> In the command line below, >>> I will eventually insert a screen maker --+ >>>¦ >>> separatefields,scale=height=2*in_h:sws_flags=neighbor,--HERE--,tblend=normal[D] >>> >>> but first, I need to persuade 'tblend' to accept the 2 frames made by >>> 'separatefields' (and scaled >>> by 'scale'). >>> But as you can see in the next section (extracted from the log), though >>> the >>> 'height=2*in_h' >>> directive works (i.e., the 1920x540 frames from 'separatefields' are >>> scaled >>> to 1920x1080), ffmpeg is >>> not happy. Is it because the SARs don't match? How can I overcome that? >>> Thanks! >> >> Insert setsar=1 after scale? > > Thank you, Paul. It works. I apologize for not finding that solution on my > own ...how embarrassing. > > Next problem: 'tblend' appears to be malfunctioning. > > "The tblend (time blend) filter takes two consecutive frames from one single > stream, and outputs the > result obtained by blending the new frame on top of the old frame." > > #1 - This works as expected: > ffmpeg -i IN -filter_complex "telecine=pattern=5, split[A][B], > [A]select='not(eq(mod(n+1\,5)\,3))'[C], [B]select='eq(mod(n+1\,5)\,3)', > datascope=size=1920x1080:x=45:y=340:mode=color2[D], [C][D]interleave" OUT > > #2 - This malfunctions: > ffmpeg -i IN -filter_complex "telecine=pattern=5, split[A][B], > [A]select='not(eq(mod(n+1\,5)\,3))'[C], [B]select='eq(mod(n+1\,5)\,3)', > separatefields, > scale=height=2*in_h:sws_flags=neighbor, setsar=1, tblend, > datascope=size=1920x1080:x=45:y=340:mode=color2, [C][D]interleave" OUT > > #1 brings up datascope in frames (n, zero-based): 2 7 12 17 etc., as > expected. > #2 brings up datascope in frames (n, zero-based): 4 7 10 13 etc. > > Another clue is: > #1 OUT has 598 frames. > #2 OUT has 716 frames. > > Would someone who has the latest nightly build kindly confirm this. I can > supply the test video -- > contact me off-list. You still can not read documentation? Why? interleave filter simply picks frames from several streams. This is obviously explained in documentation. tblend filter blend successive frames, this is obviously different from interlace filter as that will not drop any frame like interlace does. > > Thanks, > Mark. > ___ > ffmpeg-user mailing list > ffmpeg-user@ffmpeg.org > https://ffmpeg.org/mailman/listinfo/ffmpeg-user > > To unsubscribe, visit link above, or email > ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe". ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-user] ffplay options for setting audio volume and locating player windows on the screen
Hey, > Thank you for responding. Regarding -volume, i'm just putting 30 > without the %. > Regarding -top and -left, my version of ffplay (the latest MAC OSX download > from the website) > Doesn't seem to know these options. > > I'm confused why generic options like these would vary from > build/distribution to build? Woah that's really weird. You could say that they vary because they're such generic parameters: ffplay mostly doesn't interact with the hardware. Drawing the frames in a window, playing the sound through the system audio server, is done by an external library, my build uses SDL2 for this. I thought that was the most common case when building ffplay for macOS. But anyway, a lot of the common parameters are usually "mapped" to the same ffmpeg option, so the commands are as consistent as possible even if you use different components in your build. The options do vary from build to build, because each build is probably going to have some meaningful difference in the source code, the project is active and keeps changing. The banner that shows up by default when you type any ffmpeg command indicates almost everything there is to know about each build configuration, since there are practically infinite variations of the software, people tend to assume you're running a fairly recent build unless you provide that. A lot of people think they are downloading the latest version as they download the latest release, but it's actually pretty old (relatively) since by nature there is only a few a year, if that. The options I mentioned go back a couple years though, so I'm not sure, maybe it wasn't working for some time? Do post the output when you get the chance (from ffplay command) Regards, Ted Park ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-user] ffplay options for setting audio volume and locating player windows on the screen
On 04/15/2020 03:40 AM, Moritz Barsnick wrote: On Tue, Apr 14, 2020 at 18:46:27 -0700, Michael Glenn Williams wrote: -snip- ffplay doesn't support position coordinates, as far as I know. While it is a good player, it is quite basic regarding some features. You may want to choose a more powerful player such as mpv. Thanks, Moritz Michael Glenn, I second Moritz's suggestion. MPV is a joy. It has just the right amount of user interface and uses the keyboard as a remote control -- no mouse needed. It will accurately single step through frames, forward or backward. Awesome. Mark. ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-user] ffplay options for setting audio volume and locating player windows on the screen
On Tue, Apr 14, 2020 at 18:46:27 -0700, Michael Glenn Williams wrote: > Hey Ted! > > Thank you for responding. Regarding -volume, i'm just putting 30 > without the %. > Regarding -top and -left, my version of ffplay (the latest MAC OSX download > from the website) > Doesn't seem to know these options. > > I'm confused why generic options like these would vary from > build/distribution to build? To help us understand, you should quote the *actual* command line used (see Ted guessing?), and its complete, uncut console output. ffplay doesn't support position coordinates, as far as I know. While it is a good player, it is quite basic regarding some features. You may want to choose a more powerful player such as mpv. Thanks, Moritz ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-user] "Non-monotonous DTS in output stream" error, previous and current values are equal
Try this: Add -dts_delta_threshold 0 *before* the input and retest. On Sun, 23 Feb 2020, 16:27 Crazy Red Elephant via ffmpeg-user, < ffmpeg-user@ffmpeg.org> wrote: > I tried adding "-fflags +ignpts" before input but that didn't help, there > are still DTS warnings, however, this time the values are not equal for > some reason (except the first pair): > > [mp4 @ 01a902a100c0] Non-monotonous DTS in output stream 0:0; > previous: 6030, current: 6030; changing to 6031. This may result in > incorrect timestamps in the output file. > [mp4 @ 01a902a100c0] Non-monotonous DTS in output stream 0:0; > previous: 180204, current: 177177; changing to 180205. This may result in > incorrect timestamps in the output file. > [hls @ 01a901cc9380] Opening 'crypto:part1.ts' for reading > [mp4 @ 01a902a100c0] Non-monotonous DTS in output stream 0:0; > previous: 360384, current: 357357; changing to 360385. This may result in > incorrect timestamps in the output file. > [mp4 @ 01a902a100c0] Non-monotonous DTS in output stream 0:0; > previous: 540565, current: 537538; changing to 540566. This may result in > incorrect timestamps in the output file. > [hls @ 01a901cc9380] Opening 'crypto:part2.ts' for reading > [mp4 @ 01a902a100c0] Non-monotonous DTS in output stream 0:0; > previous: 720745, current: 717718; changing to 720746. This may result in > incorrect timestamps in the output file. > [mp4 @ 01a902a100c0] Non-monotonous DTS in output stream 0:0; > previous: 900925, current: 897898; changing to 900926. This may result in > incorrect timestamps in the output file. > > ‐‐‐ Original Message ‐‐‐ > On Sunday, February 23, 2020 3:15 PM, Crazy Red Elephant via ffmpeg-user < > ffmpeg-user@ffmpeg.org> wrote: > > > Guys please, I still have not found a solution and I'm so desperate :-( > I asked this question on multiple sites and nobody is helping... > > > > ‐‐‐ Original Message ‐‐‐ > > On Wednesday, February 12, 2020 12:44 AM, Ted Park > kumowoon1...@gmail.com wrote: > > > > > > 4. A general question - is there any difference between (a) > decrypting all .ts parts separately, then concatenating them into one big > .ts file and remuxing it into .mp4 and (b) concatenating all encrypted .ts > parts first, then decrypting the big .ts result file and then remuxing it > into .mp4? > > > > > > This makes me think you are not decrypting the segments correctly, I > had understood the specs to mean that if it isn’t specified, the IV needs > to be incremented for each segment (so you would need to use > 0x for the first segment, > 0x0001 for the second, etc. based on the > initial media sequence number of 0). > > > I don’t think the missing picture message at the end is unexpected, > since the stream was cut arbitrarily after all, this is the result of my > attempt at decrypting each segment with openssl and using cat to put them > together (ffmpeg wasn’t used, except to test the file). > > > https://media.kumowoon1025.com/videos/cat.ts > https://media.kumowoon1025.com/videos/cat.ts > > > ffmpeg-user mailing list > > > ffmpeg-user@ffmpeg.org > > > https://ffmpeg.org/mailman/listinfo/ffmpeg-user > > > To unsubscribe, visit link above, or email > > > ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe". > > > > ffmpeg-user mailing list > > ffmpeg-user@ffmpeg.org > > https://ffmpeg.org/mailman/listinfo/ffmpeg-user > > > > To unsubscribe, visit link above, or email > > ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe". > > > ___ > ffmpeg-user mailing list > ffmpeg-user@ffmpeg.org > https://ffmpeg.org/mailman/listinfo/ffmpeg-user > > To unsubscribe, visit link above, or email > ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe". ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
