There was a GQ thread the other day where I got involved in an argument about the spatial extent of colour perception (again). I maintain that under real-life and (thus far) informal experimental conditions, I have been able to perceive colour well toward the fringes of my visual field.
I’m well aware that many consider this flat-out impossible, that I’m deluding myself, that my perceptual system is deceiving me, etc - I’d like to nail this down with some experiments - which will either convince me that they’re right, or provide some measurable evidence to the contrary.
So I need to design an experiment that tests whether I can perceive colour at, say, 90 degrees to the right of where my nose is pointing (and assuming my eyes are pointing there too).
Things to consider:
[ul]
[li]Reflection and radiosity, etc - I need to ensure that the test object/stimulus isn’t visible as a reflection anywhere near the centre of my vision.[/li][li]‘eye flick’ - I need to verify that when the test object is visible, my eyes don’t (unbeknonwst to me) rapidly flick across and sample it.[/li][/ul]
I’m thinking of something along these lines:
[ul]
[li]Two computer screens - one directly in front of me - one at arms length directly to my right.[/li][li]A computer program momentarily displays a small patch of random colour (from a pallete of, say, 6 simple colours - RGBCMY) directly in front of me, at the same time as displaying a large square of a random colour on the VDU to my right.[/li][li]I name the two colours[/li][li]I am filmed face-on by a video camera at 60fps, to verify that my eyes remain fixed ahead during the display period (is 60fps fast enough to catch an ‘eye flick’?)[/li][li]Some sort of conical hood contraption, made from black cloth, surrounds the forward VDU, so there is no way for the right-hand VDU to appear as a reflection on any part of anything in my main field of vision.[/li][li]I am filmed by a second video camera, from a little distance behind my left shoulder, at an angle that can record both screens and my spoken responses in a single narrative.[/li][/ul]
Comments/criticisms/suggestions on this method or another? (Please don’t bother to tell me I’m wrong or wasting my time - it’s not relevant)
This illusion just uses focus on a spot to generate a peripheral vision effect. I’d think something similar where the dropping balls are different colors is all that would be needed. Once you’ve set it up the effect should be obvious and not require all the rigor you propose.
I didn’t spell this out in the prevous post, it’s not the type of thing where you need to check for the possibility of cheating, unless you’re self-delusional. If you test your direct vision by looking through a tube you should identify colors correctly 100% of the time. So you only need to miss one time while doing your best to use only peripheral vision to disprove your point. You’d have to go through a lot of trouble to get all the way to the point of proving you cannot identify colors correctly with only your peripheral vision, it would be easier to just see if your miss rate increases as you move the colors away from the center line of your vision.
It sort of is necessary to eliminate that, because self-delusion is already one of the standard responses to my claim.
A single miss doesn’t constitute failure, as my claim is not that I have perfect visual acuity at the fringes. A failure would be one of the following:
A hit rate no different than that which can be attributed to chance
or
My admission that (under sufficiently controlled conditions) to the effect “Yeah - it does all look grey”
I expect my hit rate to decrease. I just don’t expect it to fail entirely.
You’re in luck. You don’t need the elaborate procedures you describe here, you just need the new Samsung Galaxy S4 phone which can track your eye movements. One of it’s features stops streaming video if you look away. There may be something it does already that would let you know if you’ve moved your eyes. If nothing else you can write your own app for this purpose.
Thanks - I’ll have a look at that - although if I use eye tracking, I need to ensure that the sampling rate is high enough to eliminate the possibility that a ‘quick flick’ occurred.
Two more things to add:
On the walk home from work today, I kept my eyes fixed on the path in front of me and observed the colours of vehicles passing by on the road (from behind me). I’m certain that I am able to discern the colours of the vehicles the moment they entered my peripheral vision. The experiment needs even more than ever to happen…
The other thing: I think I’ve worked out a quite simple setup; it turns out to be possible to carefully position a blinker-type shield to the side of the eye so that when the eye is pointing straight ahead, objects to the side are visible, but if the eye turns to look more directly, the curvature of the eyeball places the pupil behind the blinker.
I tried this and it looks like a shield placed right next to the eye socket provides a fairly large area in which to place the sample object. If the head is braced against turning, it should be impossible to see the sample object with anything but peripheral vision.
I was wondering if a slit, hole, or tube could be used. It will be tricky to get the tolerance needed to get a peripheral view while blocking any eyeball movement. You probably need something very close to your eyeball.
I was going to suggest a bit of black insulation tape attached to the outer edge of the eye socket, but the skin moves with the facial expression. I wear glasses, so it should be possible to clip the shield to those.