Continuous recording

[swex]

0. Reviewed guide and contributing documents? (Yes/No): yes
1. current master
2. git
3. standalone or part of third party [motion, MotionEyeOS, other]: motion
4. video stream source [V4L (card or USB), net cam (mjpeg, rtsp, other), mmal]: NA
5. hardware [x86, ARM, other]: NA
6. operating system [16.04, Stretch, etc, FreeBSD, other]: NA

Motion events and recordings are good, but most surveillance solutions are also provides a way to continuous recordings.
It would be great to have a way to break output videos in some customizable segments and have like motion event chapters inside(as in mkv container).
There was an old issue: #278 but it turns out it was motioneyeos related and closed after.

[usernamepi]

FYI:
The current workaround I use for for this is to use the timelapse features in the config files.
So, configure both motion capture and timelapse for the same device.
Disadvantages compared to the requested solution:

• the best fps for timelapse videos you can get is 1 fps
• limited codec/container support for timelapse videos
  All motion video configuration capabilties stay the same (including fps).

[MortenVinding]

I have been looking for a solution like that for some time.

With the current price on harddrives, I just don’t think it makes sense to only record motion events anymore.

What I want to do is stream all my cameras to my cloud server, and either let motion on my server add index points when it detects motion, or maybe even just use the build-in motion detection in the cameras to trigger index generation, for minimum processor load and power consumption.

Currently I have build my own simple system usen ffmpeg to generate .mp4 segments of 10min lenght, and a simple REST call from the camera to trigger a index pointing (URL on a list) to the segment with motion.

But I would like to use some more polished solution.