Forum: help

RE: Different eyes (chin rest) [ Reply ] By: Stuart Gardiner on 2012-08-21 00:46	[forum:13032]
Yes, that sounds right. After tomorrow, there's no subjects scheduled for over a week, so I'll have time to try things out if you've got any suggestions. Or even move to the newer version of the OPI if that one works, although I'd rather not do that part way through data collection.

RE: Different eyes (chin rest) [ Reply ]
By: Andrew Turpin on 2012-08-21 00:41

[forum:13031]

Yes - if you are on the wrong chin rest then EyeSuite returns a message and does not present the stimulus. At the moment, the OPI ignores this chin-rest-message, and pretends everything went well. So as far as the OPI is concerned, the stimulus flashed, the response window elapsed, and they (presumably) didn't press the button.

So it seems in the old OPI I didn't initialise something correctly. Even though it is displaying OS, it still thinks OD for the chin rest. Does that sound like the behaviour?

RE: Different eyes (chin rest) [ Reply ]
By: Stuart Gardiner on 2012-08-21 00:35

[forum:13030]

Just checked it, and OS is correctly displaying after the opiInitialise("left") call (and OD displays if I call "right" instead). So it seems the 900 is getting that message fine.

I also noticed that when you press on the "wrong" (i.e. the right) chin rest, the presentations run a lot quicker, as if it's returning the error message and moving straight on to the next stimulus without waiting for the stimulus presentation time to expire. So whatever is causing the problem must be when the call is made to turn on the stimulus.

RE: Different eyes (chin rest) [ Reply ]
By: Andrew Turpin on 2012-08-21 00:13

[forum:13027]

Hi Stuart,

There is definitely a "feature" in the current OPI that if you have your chin on the wrong chin rest for the eye chosen, opiPresent does not show a stimuli. It is on the "to fix" list (ie get opiPresent to return "wrong chin rest").

When you opiInitialise(..."left"), does "OS" display on the Octopus monitor? If so, the then right chin rest should work. (And vice versa with OD).

If "OS" is not displaying on the O900 monitor, then your observation that the 900 is not getting the "left" message is correct.

You have an old version of the OPI (the initial!) and perhaps I only ever tested it with the right eye. If OS is not displaying, contact me directly and we can try and fix. If it is displaying, I am a little confused.

Cheers,

Andrew

Different eyes [ Reply ]
By: Stuart Gardiner on 2012-08-20 18:08

[forum:12964]

We've got a problem. I'm measuring FOS curves, and the first step before testing is to call a function:

initialize <- function(Eye="right") {
source("octo900_opi.r")
print(paste("Initialise opi:",ret <- opiInitialize(
"C:\\Program Files\\Haag-Streit\\EyeSuite\\", # location of HS*.jar
"C:\\Program Files\\OPIStuff\\", # location of opi folder
Eye # eye to test
)))
if (!is.null(ret)) stop("Cannot initialise octopus")
}

Calling initialize("right") works fine when we're testing a right eye. But today we've had our first left eye patient, and it's not working even when I call initialize("left") before testing.

If there's any weight on the left-hand side of the chin rest (as if you were testing a right eye) it presents stimuli as expected, despite the fact that I called initialize("left"). But if there's weight on the right-hand side of the chin rest (as if you were testing a left eye), it doesn't present any stimuli at all. So it seems the Octopus isn't getting the message about which eye is being tested for some reason.

The work around for today is that we're not using the chin rest, moving it as far out of the way as possible, with the patient just holding his head against the forehead rest and being as still as possible. So we should get reasonable data, but obviously that's not a great solution, and it's only possible because this particular subject is quite steady.

Any ideas? We've got another left-eye subject coming tomorrow!