fdk-aac | standalone library of the Fraunhofer FDK AAC code | Android library

I am trying to combine two audio files, and delaying the second one. Here's my command

I am converting some old mjpeg videos (stored in .avi container) to h.265 (.mp4 container) but am noticing the colors are smearing. Using the terminal command:

ffmpeg -y -i "input-file.avi" -c:v libx265 -vtag hvc1 "output-file.mp4"

I get the following image (notice how the red and blue are stretched donward). There is a lot of motion in the scene, but the motion is mostly horizontal:

Any idea what might cause this? The detail and resolution seem fine, just the colors are being interpreted weirdly.

Full output:

I'm using ffmpeg to stream raspberry PI cam on rtsp stream.

Before I used this command in combination with OpenCV:

Below is dockerfile with installation details of ffmpeg.

I'm trying to add rotation metadata to the video recorded from RTSP stream. All works fine until I try to run recording with segment format. My command looks like this:

i'm trying to write a shell script to analyse some videos by using the first frame of each second of video as a basis. i'm using ffmpeg to extract the frames. i thought i'd hit gold when this thread came up in my searches, but it doesn't handle the deinterlacing.

i'm using a time-based approach that works well for various formats as long as the video is progressive. it doesn't work so well for interlaced video (only the first field is output in that case, creating a half-height image). i've tried various combinations of deinterlacing (yadif/bwdif) with select but the filter chains i create either cause errors or still return a half-sized image.

here is my call w/a filter that works correctly for progressive video source :
ffmpeg -i $infile -vf "select='if(eq(n\,0),1,floor(t)-floor(prev_selected_t))" -vsync 0 $outfile

the following still return only half-height images for interlaced source :
... -vf "bwdif=0,select='if(eq(n\,0),1,floor(t)-floor(prev_selected_t))'"

... -vf "bwdif=0,select='between(mod(n\,$ips)\,1\,2)'" -- $ips is images per second

... -vf "select='between(mod(n\,$ips)\,1\,2)',bwdif=0"

i've also tried various permutations of the above w/explicit frame reference (-r, -vframes, ...), still no joy.

can someone help me with the syntax ?

--EDIT--
here is the complete output of running the command with the first filter :

I am working on a ffmpeg command to overlay background music to a video which already has audio.

below is the command -

I am trying to add xfade filter and the command is working but audio of second video is missing in output video.

command is -

My objective is to play a 4 video montage of different duration, and loop all the videos until the duration of the longest video. However I couldn't figure out how to do this.

The duration of each of the inputs is:

• input1 (Duration: 00:00:00.24)
• input2 (Duration: 00:00:01.98)
• input3 (Duration: 00:00:04.02)
• input0 (Duration: 00:00:04.02)

The following code produces a video with duration equal to that of input1: