Things You Should Know About Your Lenses, But May Not

Posted by

More Elements Aren’t Always Better

Today’s lenses are marvels of design with low dispersion and aspheric elements, supersecret submolecular whatchamacallit coatings etc. I notice more and more the marketing people touting stuff like 63 elements in 45 groups!!!!! in their brochures. One thing we sometimes forget is that every air-glass interface in a lens is going to reflect a fraction of light (usually less than 1%, often as little as 0.2%). That stray light bounces around in the lens and can cause flare, loss of contrast, and ghosting (reflected sunspots in the picture).

The top lens above has 23 elements, the one below it has 12. Below are the flare tests done by Bryan Carnathan at The Digital Picture for those same lenses, 23 element lens on the left, 12 element lens on the right.

Of course lens coatings help minimize flare and the manufacturers are quick to tell you the new versions of some lenses have “supernano” or “subwavelength” coatings that are better than old coatings. They aren’t quite so quick to mention that usually just a single surface in the lens has the new coatings. The other lens surfaces are coated the old way.

Only the back of the front element in this lens actually has the new subwavelength coating. The other 15 glass-air interfaces have standard coating.

It doesn’t mean lenses with more elements always have more flare and ghosting. But there is a general tendency that they do. Older lens designs with older coatings have the same tendency. So an old lens design with old coatings and lots of elements can have major problems shooting with the sun in the field.

One other point about all that reflection that may be happening: reflected light isn’t getting to the sensor, so the f/# (which is a theoretical calculation) may be quite different than the T/# which measures actual transmission. An f/1.4 lens may, or may not, actually let f/1.4 worth of light hit the sensor. (Actually, it definitely won’t, but more on that later.)

  • Cfreak

    I know a lot about cine lenses and less about still lenses. It was a very informative read: confirming things I do know, confirming things I suspected and informing me still more. Using still lenses to shot HD video I realize how spoilt I was in the movie business and why the true cine lenses COST SO MUCH :-) -Thank you.

  • Mini

    This article is also available in German.

    Thanks for mentioning after i have already read THE WHOLE ARTICLE in English!
    Anyway there are some interesting facts in here.

  • Jacek Z.

    One thing that seems closely connected, but might be missing on page 4 is "Aperture and Shutter Values Aren’t Totally Accurate". For example, switching a 50 mm prime from f/4 and 1/500s to f/5.6 and 1/250s on the same object the amount of light coming to the sensor - at least theoretically - should be equal. But often IT ISN'T.

    So... either something is not too accurate with the shutter speed, or the lens diaphragm mechanism doesn't work perfectly, letting more or less light than declared. Let it be just 1/3 EV. Small difference, but a difference, especially when this "pleasing circular aperture" gets a bit oval at some values.

  • David

    That bit about the magical wizard's tube made me laugh. Oddly, my TC is the only lens I regret buying. Never seem to use the thing; never willing to sully the performance of the base lens.

  • Walter

    Ugh, the rattling IS.

    It turns out that Olympus sensor-shift IS systems calibrate/reset themselves every time you turn the camera off, which makes a rattling noise. Easily the most common newbie question on the 4/3 forums over at DPReview is "I got this E-whatever and it's great, but it makes this horrid rattling noise every time I turn it off, is it broken?"