Ewald Alshuth, Wilhelm Bernhard Kloke, Stephanie Jainta, Jörg Hoormann, Wolfgang Jaschinski

The calibration of eye movement measurements using binocular targets may be adversely affected by a fixation disparity. Therefore, we generally use a monocular calibration routine either with shutter glasses or at a mirror stereoscope. In order to optimize the calibration procedure for high precision and reliability of vergence measurements, we conducted an experiment using Eyelink II equipment at a mirror stereoscope.

Monocular and binocular presentations of calibration targets were interleaved in two different ways: 1) Blockwise interleave: monocular calibration blocks were made before and after a binocular test phase comprising two binocular calibration blocks; 2) Targetwise interleave: two blocks were made up of a random series of binocular and left and right eye monocular presentations of all targets In the second way, it is expected that the fixation disparity in the binocular presentations is estimated with better reliability than in the first. The quantitative results of 20 subjects and 2 replications suggest that using a mixed calibration strategy may be advantageous, where most calibration targets are displayed binocularly and only a subset monocularly, because less presentations are needed without deteriorating the information regarding fixation disparity.