Knowledge Base

Problem

After placing a 96-well plate in DanioVision, the EthoVision detection algorithm produces poor tracking results — many fish trajectories do not match the animals' actual movements, appearing unstable or incorrect.

Details

Tracking inaccuracy in DanioVision multi-well experiments is commonly caused by suboptimal detection settings or an inappropriate detection algorithm for the specific imaging conditions (e.g., lighting, contrast, fish pigmentation). Affects DanioVision setups using EthoVision XT for acquisition and analysis.

Solution

1. Open the **Detection Settings** in EthoVision XT for the DanioVision experiment.
2. Try switching the detection algorithm. If using the default (subtraction-based) method, switch to the **Dynamic** detection method, which adapts to local background changes and is often better suited to multi-well plate setups.
3. Adjust the **detection threshold** incrementally and preview the result — the detected subject should closely match the actual fish position and size in each well.
4. Check that the arena definition (well boundaries) is correctly set for each well to avoid cross-well tracking errors.
5. Verify that the DanioVision lighting is uniform and the plate is seated flat and consistently.
6. Once tracking appears accurate in the preview, save the detection settings and proceed with the acquisition or re-analysis.