February 2020 - ffmpeg-user - lists.ffmpeg.org

Re: [FFmpeg-user] Submitting a Feature Request
by Chris 08 Feb '20

08 Feb '20

-----Original Message----- From: Carl Eugen Hoyos <ceffmpeg(a)gmail.com> To: FFmpeg user questions <ffmpeg-user(a)ffmpeg.org> Sent: Sun, Jan 19, 2020 11:34 am Subject: Re: [FFmpeg-user] Submitting a Feature Request > ffmpeg with the command-line interface forces video levels to either 0 - 255 or 16 - 235 > Ok. > Could you 1) repeat the usecase for other levels and 2) post the command line > that produces a file for this use-case but with wrong video levels. > Carl Eugen Here is the code that invokes ffmpeg (ffmpeg is invoked from a C language program). // Open an input pipe from ffmpeg and an output pipe to a second instance of ffmpeg FILE *pipein = popen("ffmpeg -i C0015.MP4 -f image2pipe -vcodec rawvideo -pix_fmt rgb24 -s 1280x720 -", "r"); FILE *pipeout = popen("ffmpeg -y -f rawvideo -vcodec rawvideo -pix_fmt rgb24 -s 1280x720 -r 59.94 -i - -f mp4 -q:v 5 -vcodec rawvideo -pix_fmt rgb24 -an output.avi", "w"); Again, this code works perfectly fine, exactly as expected. I want to add a feature where ffmpeg does not force the levels to 0 - 255 or 16 - 235: out_range=unity.

2 15

US Consultants
by tmason＠trueprobity.com 07 Feb '20

07 Feb '20

Are there any US based FFMPEG consultants available?

3 2

Where can I find source code?
by Mark Filipak 07 Feb '20

07 Feb '20

I keep winding up here: git.videolan.org Where can I find source code? Thanks, Mark.

5 5

What does ffmpeg do when packs switch?
by Mark Filipak 07 Feb '20

07 Feb '20

I guess I need some schooling... What does ffmpeg do when packs switch from hard-telecine to soft-telecine right in the middle of a GOP? Of course, a link would be a fine response, but I've not found such a link (and I've been researching this stuff for months). I certainly hope that the answer isn't that ffmpeg doesn't care about it. I hope the answer isn't that ffmpeg just takes what the decoder outputs because in one case the decoder outputs 5 frames (3 frame pictures and 2 field pictures that are sequential, not concurrent) and in the other case the decoder output 4 frame pictures. Regards, and Thanks, Mark.

2 3

Dejudder
by Mark Filipak 07 Feb '20

07 Feb '20

This is a most important filter. I thoroughly understand judder caused by hard-telecine. I understand pulldown. I simply seek clarification of the dejudder filter's behavior. Re: https://ffmpeg.org/ffmpeg-filters.html#dejudder "If the original source was partially telecined content then the output of pullup,dejudder will have a variable frame rate." What does the author mean by "pullup,dejudder"? That is: Why did the author jamb the words together separated by a comma? What is the author trying to 'say'? I suspect that "of pullup,dejudder" would best be left out of the text, but I can't know because the current text somewhat mystifies me due to errors in language usage. "May change the recorded frame rate of the container." Under what specific circumstances? And (with tongue in cheek) how is the "recorded frame rate of the container" changed when the container is on a DVD-ROM? Should the sentence read: "Will change the frame rate of the output"? "Aside from that change, this filter will not affect constant frame rate video." Does that mean, 1, That the dejudder filter is ignored for CFR video (and can thereby be specified universally and indiscriminately)? or 2, That dejudder is incapable of changing CFR video? In other words, why wouldn't I simply always specify "dejudder"? Is anyone interested in pictures that visually display what telecine & pulldown are and what what they do? I have such pictures. They're very helpful and work much better than words. Is anyone interested in me acting as a critical editor for ffmpeg documentation? I would function like a book editor by suggesting textual changes to improve English language usage and meaning. Of course, all suggestions would be subject to approval by the original authors. This is a serious offer. My qualifications are as a documentarian and as an electronic hardware engineer (retired). Simply put, I'd be another pair of eyes, eyes that care. Regards, Mark.

2 1

avformat_open_input with custom AVIOContext (duration incorrect)
by tmason＠trueprobity.com 06 Feb '20

06 Feb '20

I am trying to use a custom AVIOContext to open a wav audio file buffer and read the duration along with other audio characteristics. The example code posted blow shows 2 methods of opening the same file. In the case where I call avformat_open_input with the audio.wav file saved on disk, I consistently get the correct duration. When I call avformat_open_input with the audio.wav file saved as a file buffer using a custom AVIOContext, I am not able to get a reasonable value for duration. >From the API documentation, I have tried 3 methods I can find for determining duration without luck. I have tried many different test files to verify the issue and the duration is either zero or a very large negative number. Are there any other methods of getting the duration of an audio stream or am I missing a step before trying to get the duration? m_audioInfo->duration = static_cast<double>(pFormatContext->duration) / AV_TIME_BASE; m_audioInfo->duration = (pStream->duration / (1/av_q2d(pStream->time_base))); m_audioInfo->duration = pFormatContext->duration; Any help you can provide would be greatly appreciated. Thank you! Below is the code that I am using: #include "AsyncInspectionWorker.h" using namespace v8; using namespace std; const size_t EMPTY = 0; /* User defined data holder that will be passed to avio_alloc_context() */ struct buffer_in_data { uint8_t* ptr; size_t size; }; /* Function for reading the user defined buffer_in_data structure passed to avio_alloc_context() */ static int read_packet(void* opaque, uint8_t* buf, int buf_size) { struct buffer_in_data* bd = (struct buffer_in_data*)opaque; buf_size = FFMIN(buf_size, bd->size); memcpy(buf, bd->ptr, buf_size); bd->ptr += buf_size; bd->size -= buf_size; return buf_size; } AsyncInspectionWorker::AsyncInspectionWorker(string filename, Nan::Callback *callback) : Nan::AsyncWorker(callback) { m_filename = filename; m_buffer_in_size = EMPTY; } AsyncInspectionWorker::AsyncInspectionWorker(char* buffer, size_t bufferSize, Nan::Callback* callback) : Nan::AsyncWorker(callback) { m_buffer_in = buffer; m_buffer_in_size = bufferSize; } void AsyncInspectionWorker::Execute() { AVCodec* pCodec; AVStream* pStream = NULL; AVIOContext* avio_ctx = NULL; AVCodecContext* pCodecContext; AVCodecContext* pCodecContextOrig; AVFormatContext* pFormatContext = avformat_alloc_context(); struct buffer_in_data bd = { 0, 0}; if (m_buffer_in_size != EMPTY) { // This Buffer reading code based on ffmpeg project's doc/examples/avio_reading.c uint8_t* avio_ctx_buffer = NULL; size_t avio_ctx_buffer_size = 4096; bd.ptr = reinterpret_cast<uint8_t*>(m_buffer_in); bd.size = m_buffer_in_size; if (!(avio_ctx_buffer = static_cast<uint8_t*>(av_malloc(avio_ctx_buffer_size)))) { SetErrorMessage("Unable to allocate enough memory for processing"); return; } if (!(avio_ctx = avio_alloc_context(avio_ctx_buffer, avio_ctx_buffer_size, 0, &bd, &read_packet, NULL, NULL))) { SetErrorMessage("Unable to allocate enough memory for processing"); return; } pFormatContext->pb = avio_ctx; if (avformat_open_input(&pFormatContext, NULL, NULL, NULL) < 0) { SetErrorMessage("Unable to read Buffer data"); return; } } else { if (avformat_open_input(&pFormatContext, m_filename.c_str(), NULL, NULL) != 0) { SetErrorMessage("Unable to open file"); return; } } if (avformat_find_stream_info(pFormatContext, NULL) != 0) { SetErrorMessage("Unable to find stream info"); return; } for (unsigned int i = 0; i < pFormatContext->nb_streams; i++) { if (pFormatContext->streams[i]->codec->codec_type == AVMEDIA_TYPE_AUDIO) { pStream = pFormatContext->streams[i]; break; } } if (pStream == NULL) { SetErrorMessage("Unable to find audio stream"); return; } pCodecContextOrig = pStream->codec; pCodec = avcodec_find_decoder(pCodecContextOrig->codec_id); if (pCodec == NULL) { SetErrorMessage("Unsupported codec"); return; } pCodecContext = avcodec_alloc_context3(pCodec); if (avcodec_copy_context(pCodecContext, pCodecContextOrig) != 0) { SetErrorMessage("Unable to copy codec context"); return; } avcodec_open2(pCodecContext, pCodec, NULL); m_audioInfo = new AudioInfo(); //m_audioInfo->duration = static_cast<double>(pFormatContext->duration) / AV_TIME_BASE; //m_audioInfo->duration = (pStream->duration / (1/av_q2d(pStream->time_base))); m_audioInfo->duration = pFormatContext->duration; m_audioInfo->sample_rate = pCodecContext->sample_rate; m_audioInfo->channels = pCodecContext->channels; avcodec_close(pCodecContext); avcodec_close(pCodecContextOrig); avformat_close_input(&pFormatContext); pFormatContext = NULL; if (avio_ctx && avio_ctx != NULL) { av_freep(&avio_ctx->buffer); avio_context_free(&avio_ctx); } } void AsyncInspectionWorker::HandleOKCallback() { Nan::HandleScope scope; const int argc = 1; Local<Object> audioInfo = Nan::New<Object>(); if(m_audioInfo != NULL) { Nan::Set(audioInfo, Nan::New<String>("duration").ToLocalChecked(), Nan::New<Number>(m_audioInfo->duration)); Nan::Set(audioInfo, Nan::New<String>("sampleRate").ToLocalChecked(), Nan::New<Number>(m_audioInfo->sample_rate)); Nan::Set(audioInfo, Nan::New<String>("channels").ToLocalChecked(), Nan::New<Number>(m_audioInfo->channels)); //This inspection reports how the data will look after transcoding, everything is converted to 16 bit (2 bytes) Nan::Set(audioInfo, Nan::New<String>("bitRate").ToLocalChecked(), Nan::New<Number>(16)); Nan::Set(audioInfo, Nan::New<String>("sampleSize").ToLocalChecked(), Nan::New<Number>(2 * m_audioInfo->channels)); delete m_audioInfo; } Local<Value> argv[argc] = {audioInfo}; callback->Call(argc, argv); } void AsyncInspectionWorker::HandleErrorCallback() { Nan::HandleScope scope; const int argc = 1; Local<Object> returnObject = Nan::New<Object>(); Nan::Set(returnObject, Nan::New<String>("error").ToLocalChecked(), Nan::New<v8::String>(ErrorMessage()).ToLocalChecked()); Local<Value> argv[argc] = {returnObject}; callback->Call(argc, argv); }

2 1

'print:file=foo.txt' fails
by Mark Filipak 06 Feb '20

06 Feb '20

The command line below creates 'foo.txt' but the file is empty (i.e., zero length). Any clues why? Thanks, Mark C:\CMD & tiny apps\ffmpeg>ffmpeg -i G:\VIDEO_TS\VTS_02_2.VOB -vf metadata=mode=print:file=foo.txt -f rawvideo -y NUL ffmpeg version N-94664-g0821bc4eee Copyright (c) 2000-2019 the FFmpeg developers built with gcc 9.1.1 (GCC) 20190807 configuration: --enable-gpl --enable-version3 --enable-sdl2 --enable-fontconfig --enable-gnutls --enable-iconv --enable-libass --enable-libdav1d --enable-libbluray --enable-libfreetype --enable-libmp3lame --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libopenjpeg --enable-libopus --enable-libshine --enable-libsnappy --enable-libsoxr --enable-libtheora --enable-libtwolame --enable-libvpx --enable-libwavpack --enable-libwebp --enable-libx264 --enable-libx265 --enable-libxml2 --enable-libzimg --enable-lzma --enable-zlib --enable-gmp --enable-libvidstab --enable-libvorbis --enable-libvo-amrwbenc --enable-libmysofa --enable-libspeex --enable-libxvid --enable-libaom --enable-libmfx --enable-amf --enable-ffnvcodec --enable-cuvid --enable-d3d11va --enable-nvenc --enable-nvdec --enable-dxva2 --enable-avisynth --enable-libopenmpt libavutil 56. 33.100 / 56. 33.100 libavcodec 58. 55.101 / 58. 55.101 libavformat 58. 31.104 / 58. 31.104 libavdevice 58. 9.100 / 58. 9.100 libavfilter 7. 58.101 / 7. 58.101 libswscale 5. 6.100 / 5. 6.100 libswresample 3. 6.100 / 3. 6.100 libpostproc 55. 6.100 / 55. 6.100 [mpeg2video @ 000001b88fa7a240] Invalid frame dimensions 0x0. Last message repeated 2 times Input #0, mpeg, from 'G:\VIDEO_TS\VTS_02_2.VOB': Duration: 00:24:02.08, start: 1159.049500, bitrate: 5956 kb/s Stream #0:0[0x81]: Audio: ac3, 48000 Hz, 5.1(side), fltp, 448 kb/s Stream #0:1[0x1e0]: Video: mpeg2video (Main), yuv420p(tv, top first), 720x480 [SAR 32:27 DAR 16:9], Closed Captions, 29.97 fps, 29.97 tbr, 90k tbn, 59.94 tbc Stream #0:2[0x82]: Audio: ac3, 48000 Hz, stereo, fltp, 192 kb/s Stream #0:3[0x80]: Audio: ac3, 48000 Hz, stereo, fltp, 192 kb/s Stream #0:4[0x83]: Audio: ac3, 48000 Hz, 5.1(side), fltp, 448 kb/s Stream #0:5[0x1bf]: Data: dvd_nav_packet Stream mapping: Stream #0:1 -> #0:0 (mpeg2video (native) -> rawvideo (native)) Press [q] to stop, [?] for help Output #0, rawvideo, to 'NUL': Metadata: encoder : Lavf58.31.104 Stream #0:0: Video: rawvideo (I420 / 0x30323449), yuv420p, 720x480 [SAR 32:27 DAR 16:9], q=2-31, 124291 kb/s, 29.97 fps, 29.97 tbn, 29.97 tbc Metadata: encoder : Lavc58.55.101 rawvideo [mpeg @ 000001b88fa79000] New subtitle stream 0:6 at pos:5988366 and DTS:1166.24s More than 1000 frames duplicated= 9974528kB time=00:10:57.42 bitrate=124290.3kbits/s dup=884 drop=0 speed=62.5x [mpeg2video @ 000001b88fad6c00] ac-tex damaged at 12 1640.35 bitrate=124290.4kbits/s dup=5326 drop=0 speed=69.2x [mpeg2video @ 000001b88fad6c00] Warning MVs not available [mpeg2video @ 000001b88fad6c00] concealing 630 DC, 630 AC, 630 MV errors in B frame G:\VIDEO_TS\VTS_02_2.VOB: corrupt decoded frame in stream 1 frame=43213 fps=2080 q=-0.0 Lsize=21876581kB time=00:24:01.87 bitrate=124291.7kbits/s dup=5455 drop=0 speed=69.4x video:21876581kB audio:0kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: 0.000000%

2 4

Seeking VOB metadata
by Mark Filipak 06 Feb '20

06 Feb '20

I'm sorry I have to ask such a basic question. I use VOBEdit to examine VOBs, but it is not batch and it will not dump an entire 'VTS_xx_x.VOB' file and it doesn't allow for selective metadata sampling. Is there anything in ffmpeg or ffprobe that will, for every frame in a single VOB (or preferably, all VOBs in an entire TS: 'VTS_xx_1.VOB'+'VTS_xx_2.VOB'+'VTS_xx_3.VOB'+...), list the following? SCR progressive_sequence picture_structure top_field_first repeat_first_field progressive_frame Thanks, Mark.

2 6

How can I pipe ffmpeg listing?
by Mark Filipak 05 Feb '20

05 Feb '20

This: >ffmpeg -ss 2:00 -i g:\VIDEO_TS\VTS_02_2.VOB -filter:v idet -frames:v 900 -an -f rawvideo -y NUL produces the listing at the bottom -- it is that listing that I'm trying to pipe. However, adding 'pipe:' & 'pipe:1', thus: >ffmpeg -ss 2:00 -i g:\VIDEO_TS\VTS_02_2.VOB -filter:v idet -frames:v 900 -an -f rawvideo -y NUL pipe:1 | FINDSTR /L "Parsed_idet" doesn't work and produces this: [NULL @ 000002a346af4a80] Unable to find a suitable output format for 'pipe:1' pipe:1: Invalid argument Redirection to a file doesn't work, either. Operating system is Windows 10. I've searched for a solution for the last few hours. I'd really like to do some post-processing, but how? Thanks for any help, Mark. ffmpeg version N-94664-g0821bc4eee Copyright (c) 2000-2019 the FFmpeg developers built with gcc 9.1.1 (GCC) 20190807 configuration: --enable-gpl --enable-version3 --enable-sdl2 --enable-fontconfig --enable-gnutls --enable-iconv --enable-libass --enable-libdav1d --enable-libbluray --enable-libfreetype --enable-libmp3lame --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libopenjpeg --enable-libopus --enable-libshine --enable-libsnappy --enable-libsoxr --enable-libtheora --enable-libtwolame --enable-libvpx --enable-libwavpack --enable-libwebp --enable-libx264 --enable-libx265 --enable-libxml2 --enable-libzimg --enable-lzma --enable-zlib --enable-gmp --enable-libvidstab --enable-libvorbis --enable-libvo-amrwbenc --enable-libmysofa --enable-libspeex --enable-libxvid --enable-libaom --enable-libmfx --enable-amf --enable-ffnvcodec --enable-cuvid --enable-d3d11va --enable-nvenc --enable-nvdec --enable-dxva2 --enable-avisynth --enable-libopenmpt libavutil 56. 33.100 / 56. 33.100 libavcodec 58. 55.101 / 58. 55.101 libavformat 58. 31.104 / 58. 31.104 libavdevice 58. 9.100 / 58. 9.100 libavfilter 7. 58.101 / 7. 58.101 libswscale 5. 6.100 / 5. 6.100 libswresample 3. 6.100 / 3. 6.100 libpostproc 55. 6.100 / 55. 6.100 [mpeg2video @ 0000015ef0c6acc0] Invalid frame dimensions 0x0. Last message repeated 2 times Input #0, mpeg, from 'g:\VIDEO_TS\VTS_02_2.VOB': Duration: 00:24:02.08, start: 1159.049500, bitrate: 5956 kb/s Stream #0:0[0x81]: Audio: ac3, 48000 Hz, 5.1(side), fltp, 448 kb/s Stream #0:1[0x1e0]: Video: mpeg2video (Main), yuv420p(tv, top first), 720x480 [SAR 32:27 DAR 16:9], Closed Captions, 29.97 fps, 29.97 tbr, 90k tbn, 59.94 tbc Stream #0:2[0x82]: Audio: ac3, 48000 Hz, stereo, fltp, 192 kb/s Stream #0:3[0x80]: Audio: ac3, 48000 Hz, stereo, fltp, 192 kb/s Stream #0:4[0x83]: Audio: ac3, 48000 Hz, 5.1(side), fltp, 448 kb/s Stream #0:5[0x1bf]: Data: dvd_nav_packet Stream mapping: Stream #0:1 -> #0:0 (mpeg2video (native) -> rawvideo (native)) Press [q] to stop, [?] for help Output #0, rawvideo, to 'NUL': Metadata: encoder : Lavf58.31.104 Stream #0:0: Video: rawvideo (I420 / 0x30323449), yuv420p, 720x480 [SAR 32:27 DAR 16:9], q=2-31, 124291 kb/s, 29.97 fps, 29.97 tbn, 29.97 tbc Metadata: encoder : Lavc58.55.101 rawvideo [mpeg @ 0000015ef0c69580] New subtitle stream 0:6 at pos:106573838 and DTS:1279.62s frame= 900 fps=0.0 q=-0.0 Lsize= 455625kB time=00:00:30.03 bitrate=124291.7kbits/s speed=38.9x video:455625kB audio:0kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: 0.000000% [Parsed_idet_0 @ 0000015ef0c72840] Repeated Fields: Neither: 772 Top: 66 Bottom: 63 [Parsed_idet_0 @ 0000015ef0c72840] Single frame detection: TFF: 687 BFF: 1 Progressive: 159 Undetermined: 54 [Parsed_idet_0 @ 0000015ef0c72840] Multi frame detection: TFF: 901 BFF: 0 Progressive: 0 Undetermined: 0

4 4

Error implementing the AFFTDN filter with a noise type of w
by tmason＠trueprobity.com 05 Feb '20

05 Feb '20

Hello, I am trying to clean up an audio file which contains background noise using the AFFTDN filter with a noise type of white noise. I have not been able to find an example of the AFFTDN filter implemented and working. Based on other examples I tried this command line code. ffmpeg -i C:\Media\test.wav afftdn="'nt=w':om='o'" C:\media\test_clean.wav When I run this Unalbe to find a suitable output format for afftdn="'nt=w':om='o'": Invalid argument. Thank you in advance for your help. Tim Mason

2 2