How to remove a frame with ffmpeg without re-encoding? - ffmpeg

I am making a datamoshing program in C++, and I need to find a way to remove one frame from a video (specifically, the p-frame right after a sequence jump) without re-encoding the video. I am currently using h.264 but would like to be able to do this with VP9 and AV1 as well.
I have one way of going about it, but it doesn't work for one frustrating reason (mentioned later). I can turn the original video into two intermediate videos - one with just the i-frame before the sequence jump, and one with the p-frame that was two frames later. I then create a concat.txt file with the following contents:
file video.mkv
file video1.mkv
And run ffmpeg -y -f concat -i concat.txt -c copy output.mp4. This produces the expected output, although is of course not as efficient as I would like since it requires creating intermediate files and reading the .txt file from disk (performance is very important in this project).
But worse yet, I couldn't generate the intermediate videos with ffmpeg, I had to use avidemux. I tried all sorts of variations on ffmpeg -y -ss 00:00:00 -i video.mp4 -t 0.04 -codec copy video.mkv, but that command seems to really bug out with videos of length 1-2 frames - while it works for longer videos no problem. My best guess is that there is some internal checker to ensure the output video is not corrupt (which, unfortunately, is exactly what I want it to be!).
Maybe there's a way to do it this way that gets around that problem, or better yet, a more elegant solution to the problem in the first place.
Thanks!

If you know the PTS or data offset or packet index of the target frame, then you can use the noise bitstream filter. This is codec-agnostic.
ffmpeg -copyts -i input -c copy -enc_time_base -1 -bsf:v:0 noise=drop=eq(pos\,11291) out
This will drop the packet from the first video stream stored at offset 11291 in the input file. See other available variables at http://www.ffmpeg.org/ffmpeg-bitstream-filters.html#noise

Related

How to re-derive keyframes in damaged AVI using ffmpeg -vcodec copy

I discovered some damaged AVI files that VLC complains about broken index when I try to play them. I can play directly without ability to scroll the timeline or wait...wait... for the index to be built (but not saved) and play normally. Some other players can play them without complaining, others refuse to play.
I can solve the problem seamlessly in VirtualDub by opening the .avi with "extended options" in Open with "re-derive keyframe flags" and then saving a new .AVI file with
direct-stream-copy for video and audio. The resulting file plays perfectly.
I can also solve the problem with ffmpeg but not without problems.
ffmpeg -i INFILE -vcodec copy -acodec copy OUTFILE
Important: only stream copy and same container are of interest.
The resulting file plays in VLC without complaints or the next problem, but in many other players when jumping on the timeline the video gets distorted immediately at the jump destination and stays heavily distorted until the next I frame in the stream. All this doesn't happen when it was processed with VirtualDub.
ffmpeg is faster but most importantly it is scriptable and one could make automation for many files. With VirtualDub one has to manually process each file and wait a looooooong time for the open process to re-derive keyframe flags first. Wouldn't mind if ffmpeg speed was lost because of the automation it can provide.
So far I only found a very old unanswered mailing list post here
Can ffmpeg fix such files, without the afore mentioned problem? If yes, how?
Thank you.
AVI file indexes contain all frames (key or not), but they have a flags field (which FFmpeg fills in) which should help players seek only to keyframes. I don't have access to your exact file (ffprobe information would be helpful), but we can assume the flags field is not written correctly, e.g. it might be set for every frame or for none at all.
VLC likely parses the codec packets to derive the keyframe flag if absent in the container, but other players might not. I think what you're looking for is to derive keyframe flags while stream-copying. The exact commandline depends a bit on the codec. For example, for H264 you'd want to dump to annex-B as intermediate file format, and then re-read that so the H264 parser is invoked, which sets the keyframe flag, and then re-mux that into AVI - but H264 in AVI is rare so that's probably not what's happening here.
So for a solution, I will need the output of ffprobe $file so I know what codec the AVI file contains.

How to grab specific segments of M3U8 via ffmpeg?

I recently asked baout how I could download segments of an online m3u8 file, and someone pointed out that this could be accomplished via ffmpeg:
ffmpeg -i [LINK] -codec copy [OUTPUT FILE] #downloads only audio segments;
ffmpeg -i [LINK] -bsf:a aac_adtstoasc -vcodec copy -c copy -crf 50 [OUTPUT] #downloads audio and video segments
For those who aren't familiar, m3u8 is formatted kinda of like a "playlist", with an m3u8 file pointing to a bunch of smaller "segments" which are pieced together to form the whole of the video. As a result, it's completely possible to halt the above commands partway through their execution and still produce a watchable video (i.e. one that will be interpreted correctly by video editors).
I'm wondering if there's a built-in method with ffmpeg that allows me to grab segments N-M of a given m3u8. If there are methods outside of ffmpeg, feel free to mention them as well. Thanks for the help.
After having looked into it, I can say that this isn't possible via ffmpeg. You could theoretically use the -ss and -t parameters to specify a starting point and duration, but ffmpeg appears to look at every clip up until the specified endpoint, making the download process prohibitively long.
If you want to download only a specific number of segments, you need to look at the m3u8 file, find its associated media list, and download segments from that media list.

Scene detection and concat makes my video longer (FFMPEG)

I'm encoding videos by scenes. At this moment I got two solutions in order to do so. The first one is using a Python application which gives me a list of frames that represent scenes. Like this:
285
378
553
1145
...
The first scene begins from the frame 1 to 285, the second from 285 to 378 and so on. So, I made a bash script which encodes all this scenes. Basically what it does is to take the current and previous frames, then convert them to time and finally run the ffmpeg command:
begin=$(awk 'BEGIN{ print "'$previous'"/"'24'" }')
end=$(awk 'BEGIN{ print "'$current'"/"'24'" }')
time=$(awk 'BEGIN{ print "'$end'"-"'$begin'" }')
ffmpeg -i $video -r 24 -c:v libx265 -f mp4 -c:a aac -strict experimental -b:v 1.5M -ss $begin -t $time "output$count.mp4" -nostdin
This works perfect. The second method is using ffmpeg itself. I run this commands and gives me a list of times. Like this:
15.75
23.0417
56.0833
71.2917
...
Again I made a bash script that encodes all these times. In this case I don't have to convert to times because what I got are times:
time=$(awk 'BEGIN{ print "'$current'"-"'$previous'" }')
ffmpeg -i $video -r 24 -c:v libx265 -f mp4 -c:a aac -strict experimental -b:v 1.5M -ss $previous -t $time "output$count.mp4" -nostdin
After all this explained it comes the problem. Once all the scenes are encoded I need to concat them and for that what I do is to create a list with the video names and then run the ffmpeg command.
list.txt
file 'output1.mp4'
file 'output2.mp4'
file 'output3.mp4'
file 'output4.mp4'
command:
ffmpeg -f concat -i list.txt -c copy big_buck_bunny.mp4
The problem is that the "concated" video is longer than the original by 2.11 seconds. The original one lasts 596.45 seconds and the encoded lasts 598.56. I added up every video duration and I got 598.56. So, I think the problem is in the encoding process. Both videos have the same frames number. My goal is to get metrics about the encoding process, when I run VQMT to get the PSNR and SSIM I get weird results, I think is for this problem.
By the way, I'm using the big_buck_bunny video.
The probable difference is due to the copy codec. In the latter case, you tell ffmpeg to copy the segments, but it can't do that based on your input times.
It has to find first the previous I frames (a frame that can be decoded without any reference to any previous frame) and starts from here.
To get what you need, you need to either re-encode the video (like you did in the 2 former examples) or change the times to stop at I frames.
To assert I getting your issue correctly:
You have a source video (that's encoded at variable frame rate, close to 18fps)
You want to split the source video via ffmpeg, by forcing the frame rate to 24 fps.
Then you want to concat each segment.
I think the issue is mainly that you have some discrepancy in the timing (if I divide the frame index by the time you've given, I getting between 16fps to 18fps). When you are converting them in step 2, the output video segment time will be 24fps. ffmpeg does not resample in the time axis, so if you force a video rate, the video will accelerate or slow down.
There is also the issue of consistency for the stream:
Typically, a video stream must start with a I frame, so when splitting, FFMPEG has to locate the previous I frame (when using copy codec, and this changes the duration of the segment).
When you are concatenating, you could also have the issue of consistency (that is, if the segment you are concatenating does end with a I frame, and the next one starts with a I frame, it's possible FFMPEG drops either one, although I don't remember what is the current behavior now)
So, to solve your issue, if I were you, I would avoid step 2 (it's bad for quality anyway). That is, I would use ffmpeg to split the segments of interest based on the frame number (that's the only value that's not approximate in your scheme) in png or ppm frames (or to a pipe if you don't care about keeping them) and then concat all the frames by encoding them at the last step with the expected rate set to totalVideoTime / totalFrameCount.
You'll get a smaller and higher quality final video.
If you can't do what I said for whatever reason, at least for the concat input, you should use the ffconcat format:
ffconcat version 1.0
file segment1
duration 12.2
file segment2
duration 10.3
This will give you the expected duration by cutting each segment if it's longer
For selecting by frame number (instead of time as time is hard to get right on variable frame rate video), you should use the select filter like this:
-vf select=“between(n\,start_frame_num\,end_frame_num),setpts=STARTPTS"
I suggest checking the input and output frame rate and make sure they match. That could be a source of the discrepancy.

FFmpeg segmentation and inaccurate/wrong framerate

I use ffmpeg to save on files an RTSP stream at 15 fps. The command is similar to this (I've simplified it):
ffmpeg -y -i rtsp://IP/media.amp -c copy -r 15 -f segment -segment_time 60 -reset_timestamps 1 -segment_atclocktime 1 -strftime 1 outputFile%Y-%m-%d_%H-%M-%S.mp4
It basically creates 1 minute long files from the stream, but the problem is that the framerate of every segmented file is NEVER 15fps.
The values that I get are something like this.
14.99874
15.00031
This is a huge problem for me because I need to merge these files with other 15fps videos and the result is not good. The merged file is unstable, the image crashes and sometimes even VLC crashes if I randomly click on the time bar.
If I just merge the stream files all is well, when I try it to mix it with something else, there is nothing I can do to have a video file that is watchable and stable.
Is this normal? What can I do to have segments with a fixed 15fps without re-encoding?
Thanks in advance.
As Mulvya pointed out, ffmpeg truncates the last frame.
There are two ways to solve this:
1) Save the files to another container other than mp4, it can be TS
2) Remove the last frame of the video also works but you have to use a filter which means re-encoding which can be long and heavy on the cpu/ram

ffmpeg convert without loss quality

I need convert all videos to my video player (in website) when file type is other than flv/mp4/webm.
When I use: ffmpeg -i filename.mkv -sameq -ar 22050 filename.mp4 :
[h264 # 0x645ee0] error while decoding MB 22 1, bytestream (8786)
My point is, what I should do, when I need convert file type: .mkv and other(not supported by jwplayer) to flv/mp4 without quality loss.
Instead of -sameq (removed by FFMpeg), use -qscale 0 : the file size will increase but it will preserve the quality.
Do not use -sameq, it does not mean "same quality"
This option has been removed from FFmpeg a while ago. This means you are using an outdated build.
Use the -crf option instead when encoding with libx264. This is the H.264 video encoder used by ffmepg and, if available, is the default encoder for MP4 output. See the FFmpeg H.264 Video Encoding Guide for more info on that.
Get a recent ffmpeg
Go to the FFmpeg Download page and get a build there. There are options for Linux, OS X, and Windows. Or you can follow one of the FFmpeg Compile Guides. Because FFmpeg development is so active it is always recommended that you use the newest version that is practical for you to use.
You're going to have to accept some quality loss
You can produce a lossless output with libx264, but that will likely create absolutely huge files and may not be decodeable by the browser and/or be supported by JW Player (I've never tried).
The good news is that you can create a video that is roughly visually lossless. Again, the files may be somewhat large, but you need to make a choice between quality and file size.
With -crf choose a value between 18 to around 29. Choose the highest number that still gives an acceptable quality. Use that value for your videos.
Other things
Add -movflags +faststart. This will relocate the moov atom from the end of the file to the beginning. This will allow the video to begin playback while it is still being downloaded. Otherwise the whole video must be completely downloaded before it can begin playing.
Add -pix_fmt yuv420p. This will ensure a chroma subsampling that is compatible for all players. Otherwise, ffmpeg, by default and depending on several factors, will attempt to minimize or avoid chroma subsampling and the result is often not playable by non-FFmpeg based players.
convert all mkv to mp4 without quality loss (actually it is only re-packaging):
for %a in ("*.mkv") do ffmpeg.exe -i "%a" -vcodec copy -acodec copy -scodec mov_text "%~na.mp4"
For me that was the best way to convert it.
ffmpeg -i {input} -vcodec copy {output}
I am writing a script in python that appends multiple .webm files to one .mp4. It was taking me 10 to 20 seconds to convert one chunk of 5 seconds using:
ffmpeg -i {input} -qscale 0 copy {output}
There's some folders with more than 500 chunks.
Now it takes less than a second per chunk. It took me 5 minutes to convert a 1:20:00 long video.
For MP3, the best is to use -q:a 0 (same as '-qscale 0'), but MP3 has always loss quality.
To have less loss quality, use FLAC
See this documentation link

Resources