Hi All,
I'm working on a Video Application and would like to hear some opinionenter code here
of other developer about the logic for this section of the application.
The application I'm working on will allow two live video streams to be recorded, and combined into a single FLV file. The application will have two "channels", I will have a waiting list for each channel that you need to be in in order to participate. When it's your turn I will stream videos and audio (Flash Appl) to a FMS where I'm going to record the FLV. The name of the FLV will be the timestamp of moment the streams starts. The two channes may have concurrent participants or one channel may be with no participant for a while.
Let's say the application runs for 1 hour. At the end, I will have two list of FLV files, one for each channel. The name of each videos will have this format: c1_timestamohere.flv for channel 1 and c2_timestamphere.flv for the other channel. Each video will be 320x240 and all videos will have audio.
This is what I need to do: I need to create a single file, 640x240 that is the result of what happened in the application during the time it was running. Like having a screen recording software (of course this would have to be a server solution, since I need it to be automatic).
I have manage to do this: * Export all frames of each video to images so I can re-compose them later. (FFMPEG) * Extract the audio from each FLV (FFMPEG)
This is where I would like to get help: 1.) What logic can I use to compose the videos. I think I can extract the duration of each video and create a "timeline", generate black images 320x240 when there is a hole in the timeline and use ImageMagick to generate 640x320 images, and then use FFMPE to generate the video from the large images. 2.) What logic and tools can I use for composing the audio? I need help here. Probably generate empty audio files of different durations, and add them to the holes in the timeline as weel??
If you have any other idea, even if it a totally different approach, I would like to hear it.
Thanks in advance!