Routine for Eye tracker calibration

1
Hello,

I'm actually running some experiments using an Eyetracker (via PTB +
Eyelink), and I'd like to know if one of you developed a smart routine
to let the subject start his calibration/validation/drift correction
easily, I mean automatically, with a signal (a sound or something
else) to the experimentator if one of the step failed ? Or do you have
any reasons continuing to start each calibration step by your self ?
There is an old topic on this forum about that question, but no answer
since 2007...
Thanks,
Nathan
2
Hello Jorge,
How would you automatize things in the first case ?
The second point is also my problem. I can't find in EyelinkDoTrackerSetup.m the mean error rate...is it retrieved only from the host computer ? I'm sure though that the Eyelink toolbox get this information somewhere !
Did you try to teach your subjects to lauch calibration steps by themselves ? It may be a bit risky...

Thanks,

Nathan

--- In psychtoolbox@yahoogroups.com, "jorge7582" <jorge7582@...> wrote:
>
> This is an interesting point I was also thinking about.
>
> Testing the code I arrived to the conclusion that it cannot be completely done.
>
> I think there are two types of failure during the calibration:
>
> 1 - the subject cannot finish with all the fixation points required
> 2 - the average error in the fixations is too high
>
> The first one could be automatized but the second one I don´t think is possible. I didn´t find a way to obtain the average error from eyelink to decide if the calibration has failed or not.
>
> Anybody has another ideas about this issue?
>
> Thanks!
>
> Jorge.
>
>
>
> --- In psychtoolbox@yahoogroups.com, "idiot_eque" <nathan.faivre@> wrote:
> >
> > No one for this issue ?
> >
> >
> > --- In psychtoolbox@yahoogroups.com, "idiot_eque" <nathan.faivre@> wrote:
> > >
> > > Hello,
> > >
> > > I'm actually running some experiments using an Eyetracker (via PTB +
> > > Eyelink), and I'd like to know if one of you developed a smart routine
> > > to let the subject start his calibration/validation/drift correction
> > > easily, I mean automatically, with a signal (a sound or something
> > > else) to the experimentator if one of the step failed ? Or do you have
> > > any reasons continuing to start each calibration step by your self ?
> > > There is an old topic on this forum about that question, but no answer
> > > since 2007...
> > > Thanks,
> > > Nathan
> > >
> >
>