Calibration questions

Calibration tells HVS Image the size and position of your pool in the camera picture and where every platform position is, so that paths, quadrants and platform measures are calculated correctly. Click Calibrate, then click the four pool edges (North, South, West and East); a dashed guide circle shows the result, and the platform buttons stay off until all four edges are clicked. Then calibrate every platform position you may use: click them on screen, use 'Suggested MWM Platform Positions', or enter them with the Cartesian or Polar boxes. Click Save and Return, then check lighting and set the tracking threshold before running trials. The Barnes maze has its own calibration, kept separately, so calibrating one maze never changes the other.

Going deeper

Recommended order of work: pool and platform calibration, then 'Check Lighting', then tracking threshold and subject size, then practice trials. The first platform position calibrated becomes platform 0, so calibrating the centre first lets platform 1 be quadrant 1 (north-east). Target diameter (Atlantis) sets the area that triggers platform rise and may differ from the physical platform. Keep camera settings unchanged for the whole experiment.

Calibrating tells HVS Image exactly where the pool and each platform position are. First click Calibrate and click the four pool edges (North, South, West and East). The calibration oval passes through your four clicks and a dashed guide circle shows it; a pool that goes beyond the Run screen area is refused, so adjust the camera, and the platform buttons stay off until all four edges are clicked. Then calibrate all the platform positions you may use, for example one in each quadrant. Add the centre first: the first position calibrated is 'platform 0', so platform 1 can then be in quadrant 1 (north-east). You can click the targets on screen, click 'Suggested MWM Platform Positions' (one per quadrant plus position 0 in the centre), or enter positions numerically using either the Cartesian or the Polar boxes. Follow the white text instructions at each step, then click Save and Return.

Going deeper

Positions are stored by number and the trial's 'Platform Number' refers to them. Cartesian entries are proportions of pool diameter from the North and West edges (e.g. centre 0.5, 0.5; quadrant positions 0.67/0.33, 0.67/0.67, 0.33/0.67, 0.33/0.33, each followed by Add). Polar entries are angle from North in degrees plus proportion of pool radius. If the waterline is hard to see, float a ball or similar object on the water.

No, you don't need to set it: the Barnes box shows which maze the screen is running, and it always matches the button you used to open the maze. It is ticked when you open the Barnes maze and unticked when you open the water maze. The two mazes are kept fully separate: each has its own calibration, so calibrating one never changes the other; Barnes trials save to the Barnes folders; and each maze's SUM files go in that maze's own SUM folder. To change maze, close the screen and open the other maze from its own button.

Going deeper

The Barnes screens use Barnes wording throughout (for example 'Hole Number' and 'Auto-end on escape hole'), and the Analyse heading and title bar name the maze, so you can always see which one you are in. On the Barnes maze, the calibrated holes are masked on screen during trials.

Cartesian entries are proportions of the pool diameter measured from the North and West edges. Click the Add button nearest the Cartesian boxes after each position. For one platform in the centre and one in each quadrant, halfway between centre and edge, enter: 0.5/0.5, 0.67/0.33, 0.67/0.67, 0.33/0.67, 0.33/0.33, clicking Add each time, then Save and Return.

Going deeper

Enter the centre first so that it becomes platform 0 and the quadrant positions become platforms 1–4, with platform 1 in the north-east quadrant. Use either the Cartesian or the Polar boxes for a given set, not both.

Polar entries give each platform position as an angle from North (degrees) and a proportion of the pool radius (decimal). Click the Add button nearest the Polar boxes after each one, and Save and Return when all positions are added.

Going deeper

Polar entry is convenient for platform sets defined by bearing. The radius proportion is of the radius (Cartesian uses proportion of diameter).

Click the video preview button towards the bottom right of the calibration screen and select a video recorded with the same on-screen position as those you will track. The pool or maze surface must fit within and almost fill the view; you may need to re-size videos if the image is too large. With a student licence where codecs were omitted at installation, install ffdshow_rev4532_20140717_clsid and MatroskaSplitter from C:\HVS Image 2027\Helper Apps\.

Going deeper

Calibration is geometric, so every video in the set must share the calibrated camera position; if the view differs you must recalibrate.

Click Check Lighting. On a white background, white dots show shadow; on a dark background, red spots show glare. Adjust your lighting and repeat. If you cannot remove shadow or glare completely, adjust the camera brightness (lighter for a white background, darker for a dark one) or the lens aperture (the O–C ring).

Going deeper

The aim is consistent colour and brightness across the surface with good contrast between surface and subject. Keep camera settings the same throughout an experiment.

Start with the tracking threshold at 100 and the Subject Size slider at the left; for water maze leave the slider at the left. Right-click the left or right arrows for large adjustments and left-click to fine-tune. Once the cross tracks the subject consistently, move on to running trials. For dry mazes such as Barnes, use the Subject Size slider to ignore faeces: with no faeces present, move it right until the cross encloses an area larger than any expected faeces but slightly smaller than the animal.

Going deeper

For dry mazes: with no faeces present, move the slider right until the cross and graticules appear and enclose an area larger than any expected faeces but slightly smaller than the animal. If the cross is not central, lower the threshold for dark animals on light backgrounds, raise it for white animals on dark backgrounds. Run practice trials and analyses first.