12
\$\begingroup\$

I took this photo with a Canon 50mm compact macro lens & Canon "life size converter", and it looked somewhat fuzzy.

After separating out the RGB channels, it seems evident that the different frequencies of light are taking different paths through the lens. I guess that's nothing surprising, but I haven't seen such a difference in the resulting image before -- the focal plane moves significantly, especially between green and blue.

jumping spider RGB focus breakdown

Does this phenomenon affect some lenses more than others? Or am I simply pushing this particular lens past what it's capable of?

Also, if I wanted my next lens purchase to be something that minimizes this effect, what feature am I shopping for? (Or is my best bet to use a filter and shoot in B&W?)

Improve this question
\$\endgroup\$
3
  • 1
    \$\begingroup\$ Before you go out an spend a ton on a macro lens (or if you can't, even if you wanted too), you might want to try what is known as a "close-up filter1" or "two element lens" that screws onto the filter threads of your lens. Some of them, such as the Canon 250D Close-up lens, can be quite effective within the constraints of the overall image quality of consumer grade lenses. Here's a list put together a while back by Fuzzcraft.com. \$\endgroup\$
    – Michael C
    May 21, 2016 at 17:23
  • 1
    \$\begingroup\$ I think that you may try using it to your advantage - use different channels with "high pass filter" applied to get original details back and get additional DOF as well. \$\endgroup\$
    – Euri Pinhollow
    May 21, 2016 at 21:18
  • \$\begingroup\$ It would be "fun" if you were able to display one channel at a time in your camera's monitor, manually focus that channel, shoot, switch to the next channel, etc. and then use post-processing tools to assign R,G,B to these three images as layers in an image :-) \$\endgroup\$
    – Carl Witthoft
    May 23, 2016 at 18:30

3 Answers 3

Reset to default
12
\$\begingroup\$

You are seeing chromatic aberration — a prismatic effect which, as you nicely illustrate, reduces sharpness even in black and white photography.

A lens which has greater correction for this is called an apochromatic lens — often something like "APO" in the lens name.

Note that in lenses for telescopes and microscopes, you'll often also see achromatic lenses — these are simpler and cheaper. You may also find this designation in older camera lenses ("achromatic doublet"), but it's pretty rare to see it advertised in modern ones. Achromatic lenses (or "achromats") correct two wavelengths, rather than the three we need aligned to match human vision and color photography. For telescopes, it's usually blue and red that are brought together; Canon says that for photography it's typically blue-violet and yellow.

I don't see any claims about achromaticity in the description of the EF 50mm f/2.5 compact macro; I think it's fair to say that yes, you're pushing the lens. From reviews at Photozone.de (all with the same 8mp body, so they're comparable), CA is low but higher than some other lenses; compare your Canon 50mm f/2.5; Canon 180mm f/3.5 L; Sigma AF 150mm APO — scroll to the bottom of each page for the charts.

It is possible that the problem is exacerbated by the Life-Size converter. I assume that it's this one sold by Canon to match the lens; a few reviews of that suggest that this does contribute significantly to chromatic aberration. It's probably much better than cheap single- or double-element extenders, but if you're looking to get exceptional results at this level of macro photography, you probably want to invest in a lens that is designed for "true" 1:1 macro natively — and watch the reviews for extremely low CA before buying.

Improve this answer
\$\endgroup\$
8
  • 1
    \$\begingroup\$ Note that Canon uses no specific nomenclature in the model name for lenses that correct more effectively for CA. In terms of lenses within the Canon EF system specified as Macro, any of the "L" series will be very well corrected for CA. In their specifications, Canon uses the nomenclature Achromatic, rather than apochromatic used by some other manufacturers. (It's kind of like the "Macro" vs. "Micro" designation for lenses with a 1:1 reproduction ratio.) \$\endgroup\$
    – Michael C
    May 21, 2016 at 16:26
  • \$\begingroup\$ @Michael: huh, interesting. Does Canon use it differently from the way it's used for telescopes, with achromatic and apochromatic (roughly) meaning different levels of correction, or do they just use it as a blanket term? \$\endgroup\$
    – mattdm
    May 21, 2016 at 16:34
  • \$\begingroup\$ All true but I'd be very wary of the "life size converter" without more detail. \$\endgroup\$
    – Chris H
    May 21, 2016 at 16:36
  • 1
    \$\begingroup\$ @MichaelClark Canon's glossary in this EF Lens Work Book I found online follows the "achromat = 2 / apochromat = 3" definitions. It does appear that they don't use either in their marketing materials, though. \$\endgroup\$
    – mattdm
    May 21, 2016 at 16:46
  • \$\begingroup\$ Ok. I didn't realize they also use the term apochromatic in specifications at times. \$\endgroup\$
    – Michael C
    May 21, 2016 at 17:26
3
\$\begingroup\$

The camera lens is a converging lens. Light rays from the subject enters the lens and the lens, due to the shape and density of the glass lens, emerge tracing out a revised path. This path resembles a cone of light. We focus the camera by moving the lens forward or backward. This action adjusts the position of the apex of the cone. We want the apex to just kiss the surface of the image sensor (or film).

Sorry to report that each color of light traces out a slightly different as to length, cone of image forming rays. Thus each color has a slightly different focal length. The length of the cone determines the size of the image as projected on the image sensor.

Violet light forms an image closest to the lens. Red forms an image the farthest while green, yellow and orange form an image in intermediate positions. Thus the image size (magnification) for each color is different. We see this as a rainbow of colors surrounding the image. This is a lens error called chromatic difference of magnification. We just say chromatic aberration.

A positive lens is convex (bulges outward). A negative lens is concave (bulges inward). Luckily they behave opposing as to chromatic aberrations. The camera lens mounts a strong positive lens and sandwiches in a weak negative lens. This combination midrates to a high degree, chromatic aberration. We call this sandwich an achromatic lens (without color error). No one has ever succeeded to make a lens that truly projects a faithful image. The lens maker is faced with 7 aberrations. They do their best but so far no truly faithful image. You get what you pay for. Countering the 7 aberrations means many lens elements, each with a different shape and some with dense glass and some made with less heavy glass.

Not all lenses are equal when it comes to countering aberrations.

Improve this answer
\$\endgroup\$
3
  • \$\begingroup\$ Where did seven come from? Is it the "seven colours of the rainbow"? In that case, because cameras usually record only Red, Green and Blue light, there are only three. \$\endgroup\$
    – wizzwizz4
    May 22, 2016 at 12:22
  • 3
    \$\begingroup\$ @wizzwizz4 No, the "seven aberrations" are longitudinal and lateral CA plus five others described in the 1800s — spherical aberration, coma, field curvature, astigmatism, and distortion. You can read about them in my answer to What characteristics make a lens good or bad? or in the LensRentals blog post The Seven Deadly Aberrations. \$\endgroup\$
    – mattdm
    May 22, 2016 at 15:00
  • \$\begingroup\$ @mattdm In which case, +1. \$\endgroup\$
    – wizzwizz4
    May 22, 2016 at 15:09
1
\$\begingroup\$

If I understand your example correctly, the focal plane changes with color. I think what you show here is axial (longitudinal) chromatic aberration. APO lenses are not necessarily a solution here. Canon does not mark any lenses as such and as far as I know none of their macro lenses are perfectly corrected for axial CA.

The lens you have is pretty good. Possibly a dedicated 1:1 macro lens like the 100L or the Canon Micro would provide better CA correction at this magnification. But I think the solution here is to use higher f/stop numbers to get everything in focus. I can't help with specific techniques for shooting spiders at 1:1, but perhaps a good macro flash is the feature you are shopping for. And if you absolutely need to shoot handheld, the 100L has AF and IS designed exactly for that.

Improve this answer
\$\endgroup\$
3
  • \$\begingroup\$ More on macro depth of field: How do I get adequate depth of field in macro photography? \$\endgroup\$
    – mattdm
    May 22, 2016 at 18:21
  • \$\begingroup\$ If only these spiders would sit still in well-lit areas long enough for me to adjust the tripod :) \$\endgroup\$
    – Ian
    May 23, 2016 at 14:22
  • \$\begingroup\$ My intuition says that the aberration should be the same for a given lens, regardless of where you're focusing... so if you're focused all the way close up, the effect seems more pronounced. Does that match the reality? \$\endgroup\$
    – Ian
    May 23, 2016 at 14:24

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.