1. In MPEG streams lost packets will show up as in this example. Even though data is lost there will be an image in the player.
2. 
   

   Squares and blocks. These areas could be a much larger part of the screen. This is an example of about 1% packet loss. Jagged edges or picture elements out of phase.

3. In IR-only mode (available in FLIR IR Camera Player) lost packets will cause the whole frame to drop. Each frame consists of approximately 50 to 150 packets so a Packet-Error-Rate of 1% will cause the frame to drop. Dropped frame causes the image to freeze.
4. Decreasing frames per second will generally not help a lot, since this is not a bandwidth issue. Increasing the FPS-rate means that the number of successfully transferred frames should increase.

How can I detect lost packets?

1. Generally lost packets will cause problems or even severe problems. See above.
2. Use managed switches so you can monitor and detect layer 2-problems like CRC-faults and collisions. Be careful mixing old and new technologies such as hubs and new equipment with Fast Ethernet and duplex mode.
3. On a more technical level there is a possibility to telnet to the camera and check the RTCP-log. You can also use a network sniffer like Wireshark and analyze the stream.
4. The bandwidth needed for the stream depends on image type (MPEG, IR-only), the mode (e.g., rainbow), screen size and fps (frames per second). Generally the stream needs between 0.1 to 14 Mbps.
5. Packets will be lost if the settings produces a stream larger than the bandwidth that can be delivered through the network end-to-end. The video stream is based on UDP so packets will not be resent.
   An example: IR-monitor 15 fps, normal resolution, inverted rainbow will produce about 1–2 Mbps. (With high resolution the bandwidth could triple.)
6. Hostile environment: Generally it is a good idea to test the camera in an office environment when you suspect problems with power drops, strong electromagnetic fields, grounding issues, etc.