Why do our eyes see differently than an optical lens?

pwnell · Post by **pwnell** » Mon Nov 19, 2012 1:07 pm

I have been pondering this question for quite some time now. I have an acrylic tank that is about 30cm front to back. Inside is saltwater. When I view through the front panel, at a perpendicular angle to the front panel at the back panel, I can clearly see tiny flatworms aggregating on the rear panel's inside surface. With my eyes I can see them as clearly and well defined as if there were no water between me and the organisms.

Yet when I take a high quality macro lens (180mm F2.8 ) and push it against the acrylic, 100% perpendicular to the panel, I get this horribly distorted result. Why does the optical path light takes through the lens behave differently than the path it takes through my eyes?

rjlittlefield · Post by **rjlittlefield** » Mon Nov 19, 2012 2:21 pm

It's because the pupil of your eye is much smaller. If you stop down that 180 mm lens to have the same width entrance cone as your eye, it will experience the same aberrations.

The problem is caused by a form of spherical aberration. When a lens looks through water the optical path lengths for on-axis and off-axis rays have a different relationship than they do when it looks through air. As a result you can focus either the center rays or the edge rays or something in between, but you can't focus them all at the same time.

The effect depends strongly on the width of the entrance cone. In the center of the frame, it's roughly angular width raised to the fourth power. So stopping down is a powerful way to reduce it.

The challenge is to balance reduced aberration against diffraction blurring. The pupil of your eye is maybe 4 mm diameter. That would be f/45 on the 180, before taking into account magnification to determine an even larger effective f-number.

--Rik

NikonUser · Post by **NikonUser** » Mon Nov 19, 2012 2:30 pm

Has a lot to do with our brain, it's wonderful at interpretation. Vertebrate eyes are a really bad design, it's a wonder they work at all - it 's all in the brain. Cameras/lenses lack a smart brain.

pwnell · Post by **pwnell** » Mon Nov 19, 2012 3:04 pm

Thank you for the great explanations. I am going to try taking the photo again at higher f-stop (at the risk of introducing diffraction and other aberrations) to see this in action.

Wish I could hook up a CCD to my brain.

pwnell · Post by **pwnell** » Mon Nov 19, 2012 3:23 pm

Well here is a shot at F16. Although far from ideal (as Rik indicated - my lens can only reach F22 and at 1:1 it is far off from the F90 or something it would take), it is clearly better than at F2.8, which was used for the previous image.

And just to show how these actually look like, here is a F2.8 shot of them on the inside of the front panel, therefore there is no water between the lens and the flatworms. Just a 10mm piece of acrylic.

rjlittlefield · Post by **rjlittlefield** » Mon Nov 19, 2012 4:00 pm

As a material, acrylic plastic is actually worse than water (refractive index about 1.49 vs 1.33). But thickness is key. With only 10mm of the stuff to look through, there would be no significant spherical aberration at the magnification and aperture we're talking about here.

By the way, there's much more discussion of this topic (shooting through water) in the thread HERE. It wanders around a bit, but there's solid technical content and image comparisons on the later pages.

One aspect I forgot to mention is that it's important to keep the central rays of all the entrance cones as perpendicular to the glass as possible. That way you only run into spherical aberrations that diminish as aperture to the fourth. Cones whose central rays are not perpendicular to the glass suffer from other aberrations that do not diminish as quickly. So it's better to shoot with a long lens or any other setup that is closer to telecentric. Shooting with a short lens that has a wider angle of view will pick up worse aberrations away from center.

--Rik

pwnell · Post by **pwnell** » Mon Nov 19, 2012 4:05 pm

rjlittlefield wrote:One aspect I forgot to mention is that it's important to keep the central rays of all the entrance cones as perpendicular to the glass as possible. That way you only run into spherical aberrations that diminish as aperture to the fourth. Cones whose central rays are not perpendicular to the glass suffer from other aberrations that do not diminish as quickly. So it's better to shoot with a long lens or any other setup that is closer to telecentric. Shooting with a short lens that has a wider angle of view will pick up worse aberrations away from center.

That makes sense. I generally get better image quality with my 180mm macro lens than I do with my 100mm when photographing subjects in my aquarium.