Midterm Exam: Honors 185, Spring 1999

By taking this exam, you agree to abide by the Vanderbilt Honor Code.
All answers will be yours exclusively. You may consult books and webnotes
but you are not to formulate answers by discussing material with people.

STUDENT'S NAME:

1.. Suppose you were designing the visual system for a new animal that was to be active at night. Describe at least three ways you would make that animal's eyes function optimally under those conditions, and give the reason for each of your decisions. (3 pts)

Use rod photoreceptors; have many rods converge onto single ganglion cells (spatial convergence improves sensitivity but sacrifices spatial resolution); large pupil to admit more light; relatively large eye to capture more light; tapetum to reflect light back through the photoreceptors for a second try at capture

2. Why do you suppose "green" sunglasses seemed to help Jonathan - the colorblind painter - see better? (2 pt)

Green sunglasses, which selectively attenuate long and short wavelengths, produce more effective contrast which makes it easier for Jonathan to distinguish borders, as shown by the Mondrian test

3. If you were going to print words in very small letters of a single color, which of the following would be the worst choice of ink color to use (i.e., which would most difficult to read) - blue, yellow, red or green? Why? (2 pts)

Blue ink would be bad for very small letters, because our eyes have no short wavelength receptors in the fovea. One should also take into account the background upon which the letters appear. Yellow on white produces less contrast, but yellow on black produces very good contrast.

4. Circle the arrow(s) that point to the "neutral point (s)" (the wavelength that appears achromatic) in this two-pigment color vision system. Why does this happen? (2 pts)

The middle arrow denotes the neutral point, the wavelength that produces equal responses in both photoreceptor types and, hence, the same response as white light.

5. What is color constancy and what does it imply about the way surface colors are registered by the eye and brain? (3 pts)

Color constancy is the invariance in perceived color of a surface with variations in the wavelength distribution in the illumination. It implies that the brain uses more than just the trio of numbers from the three receptors to compute color; also important are the patterns of photoreceptors activity over space, which seems to be computed by V4.

6. Shown below are two versions of a classic visual illusion. Look directly at the center of the figure with dark squares on a white background. Notice the small, black "ghost" dots at the intersections elsewhere in the figure? When you locate directly at one of the ghosts, it disappears - it's not really there. The same occurs with the figure on the right, the one with white squares on a black background, only now the ghosts are white. Based on what we've discussed (and what you've read), why would this occur? What, in other words, occurs in our visual system to create these illusory spots? Why are they sometimes black and sometimes white? And why don't you see the ghost when you look directly at it? (4 pts)

The ghost spots arise because of lateral inhibition within the center/surround receptive fields of neurons. Cells with on-center/off-surround receptive fields produce the white ghosts; off-center/on-surround cells cause the black ghosts. We don't see spots at the point of fixation because the receptive fields are too small to be differentially stimulated by the black and white portions of the figures.

7. You may experience trouble figuring out the picture shown below. Make the image blurry by taking off your glasses or by squinting your eyes. What does "blur" do to the image, and why do you think this might actually help you recognize what you're looking at? (3 pts)

Blur selectively removes (filters) high spatial frequencies from the image, leaving just low spatial frequencies. The sharp edges in the unblurred picture are not naturally part of the image of the face; they are masking the facial details - removing them "unmasks" the face.

8. Why don't you normally see your blind spot? (1 pt)

We don't see the blind spot because a) information in surrounding portions is used by the brain to fill in the blindspot, and b) the corresponding area of the retina of the other eye provides filling input, too.

9. This question is about color matching. Suppose you adjust the relative amounts of 680 nm light and 520 nm light to produce a "yellow" that is indistinguishable from the "yellow" associated with 570 nm light. The additive pair of lights, in other words, is metameric with the single wavelength of light. Now suppose you put on a pair of sunglasses with lenses tinted red. Will your metameric matches be preserved? Describe why or why not. (3 pts)

Red-tinted sunglasses remove nearly all the short wavelengths, some of the medium wavelengths and very little of the long wavelengths. Thus the three wavelengths of light viewed through the glasses will be differentially affected (520 the most and 680 the least), and the metameric match won't be maintained. The yellow will probably take on a reddish tint (because of shifts in the balance of activation in the three cone types), but it won't look identical to the 680 red.

10. The optics of the human eye have a cut-off spatial frequency around 60 cycles/deg, which corresponds closely to human visual acuity. What might happen if the optics were twice as good, meaning that the cut-off frequency were 120 c/deg, but the rest of the eye remained structually the same? (2 pts)

Without improvements in the neural part of the visual system, this improvement in the optics would have no useful benefits - to achieve greater acuity, the photoreceptor mosaic would have to be more dense, and ganglion receptive fields would have to be smaller. In fact, one possible consequence of improving the optics only would be aliasing - we might mistake very high spatial frequencies for lower spatial frequencies.

11. Naive realism is the view that the world actually is just as it appears to us. Using any of the sources we've discussed in class, write a compelling rebuttal to this position. (5 pts)

A compelling essay would draw upon at least some of the following:

different animal species see the world differently

different people can see the world differently depending on genetic differences (color anomolous vision) and brain damage

visual illusions show that our brains "interpret" the world, sometimes getting things wrong

vision changes throughout lifetime, as the visual system matures and then ages