Skip to content

Camera setup

What a computer needs before Live Tracking works, and a first session to check it. Do this once per computer.

What you need

  • A Windows 10 or 11 computer (64-bit) with TraCell installed. No Python is needed.
  • A Vision Research Phantom camera, connected and set up as for Vision Research's own PCC software. If PCC shows the camera, TraCell finds it too.
  • The file pyphantom-3.11.11.806-py311-none-any.whl. It is part of Vision Research's Phantom SDK, and only Vision Research may hand it out, so TraCell does not contain it. TraCell was tested with this build and refuses another one.

1. Add the camera SDK

  1. Start TraCell and open Edit > Preferences (Ctrl+,).
  2. In the Camera SDK row choose Install from wheel… and pick the .whl file.
  3. The row now reads Installed (3.11.11.806). TraCell copied the file's contents, unchanged, into your user folder; an update of TraCell keeps it.

There is nothing else to install or start. TraCell starts its camera service by itself the first time it needs the camera, and stops it when TraCell closes.

2. See the camera's picture

  1. Close PCC, so that only one program controls the camera.
  2. Tools > Live Tracking. Type the resolution, FPS and exposure you want and press Live View. The first time this takes about ten seconds, while the SDK looks for the camera. If Windows Firewall asks, allow TraCell on private networks.
  3. The camera's picture shows in the main window. Focus and light the channel as usual.
  4. Press CSR for the black reference and cover the lens when TraCell asks (cameras with a shutter do it themselves).

TraCell uses the camera's memory

AutoTrack and Record set the camera's resolution, frame rate, exposure and number of partitions, and record over what is in the camera's memory. Save recordings you still need from the camera first.

3. A first session

A short run that shows whether everything works, before a real experiment. Any flow with something to see in it will do: cells, or beads in water.

  1. With Live View on, set detection as for a video: Mask, Contours and Tracker. Tools > Mask > Show Mask and Show Contours show what TraCell finds.
  2. In Live Tracking choose Frames per partition (the number beside it is the most this computer's memory allows), Stop after partitions: 3, and leave Save results on.
  3. Press AutoTrack. It records a partition, tracks it while the next one records, and stops after the third.
  4. Check: traces sit on the cells; the Dashboard has a point per cell; Open Session Folder shows tables, cells.csv and session.json.

Then use the stop rule of your experiment (Live Tracking) and save the settings with File > Save, so the next session starts from them.

When it does not work

TraCell says What to do
The Phantom SDK is not installed Step 1 above
This wheel holds another SDK build TraCell needs exactly pyphantom-3.11.11.806-py311-none-any.whl
No Phantom camera found Check power, cable and network: does PCC see the camera? Close PCC, restart TraCell and try again
Frames per partition is limited to a small number The partition being recorded is held in memory. Close other programs, or lower the resolution
Anything else Help > Report a Problem saves a zip with the log; send it with a line on what you did