Chromatic Aberration-My Understanding
 

So I bought this cheesy little Retinar .6X wide angle adapter for my camera because I am tired of leaning out of windows to get everyone in frame.

I attach it and start zooming around and see how awful the chromatic aberration is and it occurs to me that all that chromatic aberration is is the curve of the lens splitting up light like a prism.

So I think about it for a second and I start to realize that I am zooming in and getting this problem, when I bought the lens to actually "zoom out". So why do people complain about C.A., when the lenses they are getting it on were not designed to zoom in on in the first place? It's perfect when I just screw it in and go wide, like I had originally intended for its use.

Anybody else think we're pressuring lens makers to overcome a petty gripe?

"I want the lens curved but I don't want it to bend light."

Sheesh!

Some reasons I can think of:

A- You want to do run and gun shooting. The wide-angle can help in tight spaces, and to make the camera more stable (since shakes are less visible when zoomed out). But then you might have times when you want to zoom... in which case, the CA won't be good.

B- A lot of HD lenses have some trouble with CA... even/especially with the "cheap" HD lenses that run thousands of dollars. It's just harder to design HD lenses well, and low volume drives prices up.

Stephen...

Glenn's pretty much got the idea: Some people (like me) prefer to have a lens which allows wide coverage for restrictive spaces but also need the ability to zoom in to a tighter shot.

If you're only shooting wide then a fixed-focal-length wide lens would be the best option and provide the best performance for the dollar. However, you give up the flexibility of a zoom lens.

Getting a clean, sharp image isn't an unreasonable request for any optics or camera manufacturer. It's an essential part of the business. There's even a difference in quality from the higher-end companies. For example, my own experience shows that Schneider made better lenses for my 4x5 camera than Nikon. And that Leica made better compact binoculars than Zeiss.

Unfortunately what we're asking for is something that can be added to a fully engineered system which will make it do something it may not have been intended to do. In this case it's covering a wider field of view.

Adding on optics to any lens is a compromise. Even adding filters can have its own set of problems. Wide angle adapters are often expected to work with not just one lens design but several. Ideally you'd create a matched adapter for a specific lens. Century, to a certain extent, did that for the Canon XL1 lenses. The zoom-through adapter retails for about $700.

So as far as wide angle adapters are concerned, "cheap" and "good" are two terms that aren't going to be found together at all. Cheap lenses don't get the benefit of either good design or careful manufacture. Good lenses require both careful engineering and precise manufacturing, all of which can cost big bucks. Especially if the market is relatively small.

So then is it a bad coating or lens design that creates CA or is it how closely it matches the focal length of a camera's built-in lens? Or both? Is there a way to calculate which lens would best fit your camera so that way you don't need to buy an outrageously expensive one?

Coatings are primarily to reduce internal reflections and minimize loss of contrast.

Chromatic abberation is controlled by using "corrective" elements in a lens (made of slightly different types of glass or other material) which are intended to compensate for CA. That's one of the reasons why expensive lenses have so many glass elements in them.

Some of them deal with focal length, image distortion, and image circle issues. Some of them correct for chromatic abberation.

CA and other faults can be attributed to many reasons: bad lens design, poor or improper choice of glass, inaccurate grinding and polishing of the lens (shape), imprecise placement of the lens elements, off-axis and/or non-parallel element mounting, failure to accurately cement multi-element lenses together, and so on.

It would be way too difficult to calculate a good fit. The only option, really, is to test a lens and see if it works with what you have. Getting optics from a known source can also help (Century, for example, has a good reputation).

Otherwise, you'd have to actually mount the optics to your lens and give it a try to see if there's barrel or pincushion distortion, vignetting, resolution losses, chromatic abberation, etc.

So are there any cheap options? In a sense, no. While a bargain lens in itself might not cost much, trying to find a really good one can cost a lot of time. But either way you go, the up-front cost is worthwhile given the value of images (and the versatility) that a good piece of glass can provide.

Chromatic aberation is caused by "dispersion" in the glass from which a lens is made which means that the refractive index of the glass depends on the wavelength. This in turn means that the strength (focal length) of the lens is different for blue light from what it is for green light and red light. As a consequnce if a camera is adjusted to focus the green part of the spectrum the green part of an image will be sharp but the red and blue images won't because their focal planes are not coincident with the film or CCD. Furthermore, the sizes of the images depend on the focal length so that the three color images will be different sizes. This is what causes the fringing. In a compound lens there are tricks that can be used to mimimize these effects. Arranging a negative lens between two positive lenses, for example (the famous Cooke triplet) results in reduction of CA because the increase in negative focal lenth of the center element cancells the increase in positive focal length of the other two (if the indices and strengths are chosen properly). Another approach is to use lens materials that have low dispersion across the visible spectrum such as fluorite which is used in may lenses but is soft and expensive. Obviously an screw-on auxialiary lens can't use either of these tricks because multiple elements are not practical and fluorite elements must be protected (not to mention the expense).

Here is one of the best explanations of CA on the web. CA is caused by the lenses and imaging chips. If you have additional questions after reading Van Walree's explanations, post back and I'll try to answer them.