How to add an MJPEG Camera

Before Continuing

If your camera is capable of providing an H.264 Stream I suggest you take a look at these articles first.

• How I Optimized my RTSP Camera
• View H.264 and H.265 streams in the browser with Shinobi and maintain low CPU use

Minor Internal Optimization

Controlling the bitrate is similar to controlling a stream of water. With MJPEG, knowing how to do this is important. Keep in mind that these are recommendations and you don't have to use these settings.

• Framerate (FPS) : High : 10 - 15 FPS, Low : 2-5 FPS
• Bit Rate : between 256kbps and 1Mpbs

Settings in Shinobi

For the Connection section I used:

• Input Type : MJPEG

For the Input section I used:

• Analyzation Duration : 1000000
• Probe Size : 1000000
• Monitor Capture Rate : Should match the stream. Probe the stream if you need to find it.

For the Stream  section I used:

• Stream Type : MJPEG or Base64 over Websocket
• FPS (Frame Rate) : Should match the stream.

For the Record section I used:

• Video Codec : libx264
• Audio Codec : No Audio (MJPEG does not carry Audio)
• FPS (Frame Rate) : Should match the stream.

For  Global Detector Settings section I used:

• Send Frames : Yes
• Save Events to SQL : Yes (if you want to view the dots in Power Video Viewer)
• Feed-in Image Width : 640
• Feed-in Image Height : 480
• How to Record : Traditional Recording
• Trigger Record : Yes
• Recording Timeout : 2

For Motion Detection section I used:

• Use Built-In : yes
• Indifference : Varies based on camera view point
• Full Frame Detection or Regions

For Traditional Recording section I used in combination with  Global Detector Settings:

• Video Codec : libx264
• Audio Codec :  No Audio (MJPEG does not carry Audio)
• Segment Length :  2 (creates 5 second buffer)
• HLS List Size : 3 to 10