[FFmpeg-user] Read Exif from file processed by FFMPEG
Eugene Gekhter
egekhter at pixcel.com
Mon Jun 30 17:20:31 CEST 2014
Hi Francois,
Your explanation makes sense. I don't think many of my users will be
uploading files generated in ffmpeg, but it's good to know in order to
handle exceptions better.
Thanks for replying,
*Eugene Gekhter* / CEO
(847) 272-8207/ egekhter at pixcel.com
*Pixcel* Office: (800) 557-3508
3336 Commercial Ave. / Northbrook, IL 60062
http://www.pixcel.com
On Mon, Jun 30, 2014 at 1:29 AM, Francois Visagie <
francois.visagie at gmail.com> wrote:
> > -----Original Message-----
> > From: ffmpeg-user-bounces at ffmpeg.org [mailto:ffmpeg-user-
> > bounces at ffmpeg.org] On Behalf Of Eugene Gekhter
> > Sent: 30 June 2014 04:35
> > To: FFmpeg user questions
> > Subject: [FFmpeg-user] Read Exif from file processed by FFMPEG
> >
> > This might fall into the realm of extreme specificity, but I encountered
> a
> very
> > large discrepancy in performance when reading files generated by FFMPEG
> > and wanted to see if anybody had observed similar behavior in their
> > experience.
> >
> > Situation:
> > Reading exif metadata from video files using exiftool for PHP.
> >
> > Setup:
> > Version A: Original MTS file from camcorder (~ 2.5 minutes) Version B:
> > FFMPEG version of file (ffmpeg -i
> > /Volumes/16GB/PRIVATE/AVCHD/BDMV/STREAM/00013.MTS -c copy
> > ~/Desktop/00013F.MTS)
> > Both versions are in same SSD folder (for unbiased performance readings)
> >
> > Result: Time to read exif data from file (seconds):
> > Version A : 0.49976897239685
> > Version B: 17.803021192551
> >
> > Here's the code I'm using for reading and timing:
> > $logger = new \Monolog\Logger('exiftool');
> >
> > $start = microtime(true);
> >
> > $reader = \PHPExiftool\Reader::create($logger);
> > $file = "/Users/eugenegekhter/Desktop/00013.MTS";
> > $metadatas = $reader->files($file)->first();
> >
> > foreach ($metadatas as $metadata)
> > {
> > var_dump($metadata);
> > }
> >
> > $time_elapsed_us = microtime(true) - $start;
> >
> > var_dump($time_elapsed_us);
> >
> > Here's ffprobe of original file:
> >
> > ffprobe version N-64124-g84de3ed Copyright (c) 2007-2014 the FFmpeg
> > developers
> >
> > built on Jun 21 2014 17:13:02 with Apple LLVM version 5.1
> > (clang-503.0.40) (based on LLVM 3.4svn)
> >
> > configuration: --prefix=/Users/eugenegekhter/ffmpeg_build
> > --extra-cflags=-I/Users/eugenegekhter/ffmpeg_build/include
> > --extra-ldflags=-L/Users/eugenegekhter/ffmpeg_build/lib
> > --bindir=/Users/eugenegekhter/bin --extra-libs=-ldl --enable-gpl
> --enable-
> > libfdk-aac --enable-libx264 --enable-nonfree
> >
> > libavutil 52. 90.100 / 52. 90.100
> >
> > libavcodec 55. 68.100 / 55. 68.100
> >
> > libavformat 55. 44.100 / 55. 44.100
> >
> > libavdevice 55. 13.101 / 55. 13.101
> >
> > libavfilter 4. 9.100 / 4. 9.100
> >
> > libswscale 2. 6.100 / 2. 6.100
> >
> > libswresample 0. 19.100 / 0. 19.100
> >
> > libpostproc 52. 3.100 / 52. 3.100
> >
> > Input #0, mpegts, from '/Users/eugenegekhter/Desktop/00013.MTS':
> >
> > Duration: 00:02:40.19, start: 0.382589, bitrate: 25334 kb/s
> >
> > Program 1
> >
> > Stream #0:0[0x1011]: Video: h264 (High) (HDMV / 0x564D4448), yuv420p,
> > 1920x1080 [SAR 1:1 DAR 16:9], 59.94 fps, 59.94 tbr, 90k tbn, 119.88 tbc
> >
> > Stream #0:1[0x1100]: Audio: ac3 (AC-3 / 0x332D4341), 48000 Hz,
> 5.1(side),
> > fltp, 384 kb/s
> >
> > Stream #0:2[0x1200]: Subtitle: hdmv_pgs_subtitle ([144][0][0][0] /
> 0x0090),
> > 1920x1080
> >
> >
> > And probe of filed generated via ffmpeg copy command:
> >
> > ffprobe version N-64124-g84de3ed Copyright (c) 2007-2014 the FFmpeg
> > developers
> >
> > built on Jun 21 2014 17:13:02 with Apple LLVM version 5.1
> > (clang-503.0.40) (based on LLVM 3.4svn)
> >
> > configuration: --prefix=/Users/eugenegekhter/ffmpeg_build
> > --extra-cflags=-I/Users/eugenegekhter/ffmpeg_build/include
> > --extra-ldflags=-L/Users/eugenegekhter/ffmpeg_build/lib
> > --bindir=/Users/eugenegekhter/bin --extra-libs=-ldl --enable-gpl
> --enable-
> > libfdk-aac --enable-libx264 --enable-nonfree
> >
> > libavutil 52. 90.100 / 52. 90.100
> >
> > libavcodec 55. 68.100 / 55. 68.100
> >
> > libavformat 55. 44.100 / 55. 44.100
> >
> > libavdevice 55. 13.101 / 55. 13.101
> >
> > libavfilter 4. 9.100 / 4. 9.100
> >
> > libswscale 2. 6.100 / 2. 6.100
> >
> > libswresample 0. 19.100 / 0. 19.100
> >
> > libpostproc 52. 3.100 / 52. 3.100
> >
> > Input #0, mpegts, from '/Users/eugenegekhter/Desktop/00013F.MTS':
> >
> > Duration: 00:02:40.19, start: 1.416678, bitrate: 26022 kb/s
> >
> > Program 1
> >
> > Metadata:
> >
> > service_name : Service01
> >
> > service_provider: FFmpeg
> >
> > Stream #0:0[0x100]: Video: h264 (High) ([27][0][0][0] / 0x001B),
> yuv420p,
> > 1920x1080 [SAR 1:1 DAR 16:9], 59.94 fps, 59.94 tbr, 90k tbn,
> > 119.88 tbc
> >
> > Stream #0:1[0x101]: Audio: ac3 ([129][0][0][0] / 0x0081), 48000 Hz,
> > 5.1(side), fltp, 384 kb/s
> >
> >
> > If anybody has any input that would be greatly appreciated.
>
> I'm not familiar with the script language you're using, but it seems you're
> calling exiftool without command-line arguments, correct? In that case, it
> won't scan the file for each frame but will only collect container-level
> information.
>
> Seeing that the ffmpeg-muxed version takes so much longer to analyse, my
> guess would be that the ffmpeg muxer moves some container-level information
> to the end of the stream.
>
> Pending the absence of any more authorative explanations, perhaps you could
> try with files of different durations (other parameters remaining the same)
> to test this idea.
>
> >
> >
> > Thanks,
> >
> >
> >
> > *Eugene Gekhter* / CEO
> >
> > (847) 272-8207/ egekhter at pixcel.com
> > *Pixcel* Office: (800) 557-3508
> > 3336 Commercial Ave. / Northbrook, IL 60062 http://www.pixcel.com
> > _______________________________________________
> > ffmpeg-user mailing list
> > ffmpeg-user at ffmpeg.org
> > http://ffmpeg.org/mailman/listinfo/ffmpeg-user
>
> _______________________________________________
> ffmpeg-user mailing list
> ffmpeg-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/ffmpeg-user
>
More information about the ffmpeg-user
mailing list