40-150 PRO focus breathing?

whumber wrote:

Anders W wrote:

whumber wrote:

Anders W wrote:

Arctra wrote:

Today my Olympus 40-150 PRO lens arrived and while it does feel like a great lens, there's something that I noticed. I was comparing it against my trusty old Minolta Beercan and I was noticing that the Olympus zoomed all the way in at 150mm was producing a similar result as the Beercan at 210mm.

I ran the test again, but this time with APS-C crop mode turned on my Sony cam to simulate ~300ishmm and I had to get significantly closer to my test subject (in this case a old Charmander piggy bank) with the Oly to produce a similar FOV. This isn't a deal-breaking issue for me but it wasn't something I was expecting to see as I hadn't seen any reports of focus breathing issues with this lens.

Is this something that is known about this lens? Is mine the special one on the block? Or am I just fundamentally misunderstanding equivalency here? Like I said, not a huge deal-breaker but was definitely interesting to see.

It's very easy to calculate the focus breathing from the lens specs. The 40-150 PRO has a max reproduction ratio of 0.21 (at 150 mm obviously) and a minimum focus distance of 0.7 meters. The focal length, F, at that reproduction ratio, R, and focus distance, D, is calculated as

F = D/(1/R + R + 2) = 700/(1/0.21 + 0.21 + 2) = 100 mm

so yes, certainly some focus breathing going on, just as one might expect.

This equation is based on a thin lens model and while it may be fairly accurate at infinity focus, it's going to fall apart at close focus distances with thick lenses (i.e. any real lens you can imagine), especially for a lens where the distance between the principal planes is on the order of 10% of the subject distance.

I am well aware that it is based on a thin lens model but that doesn't make it useless for the purpose at hand. Moreover, the formula is useful precisely at close focus distances. At infinity, there is no real need for it.

I think that tells you a lot about the usefulness of that equation for your intended purpose. For some lenses that have principle planes relatively close together and have a long MFD, the equation may be somewhat useful, but for a lens like the 40-150 it completely falls apart. If you want another example where your formula is not useful, look at the Canon 70-200ii. The thin lens approximation equation would suggest that the focal length at MFD decreases to somewhere around 176mm, while direct measurements show that the focal length actual increases to well above 200mm at MFD.

What the results of the formula tells us is that the focal length of the 40-150 when set to 150 mm and shot at its minimum focus distance is equivalent to that of a thin lens (a single lens element) with a focal length of 100 mm. If there were no focus breathing, it would instead be equivalent to a thin lens with a focal length of 150 mm (provided that the 150 mm specification is correct in the first place).

I'm sorry but you're just flat out wrong on this one. The focal length of the 40-150 is nowhere close to 100mm at MFD when set to 150mm. In fact, there's almost no change in focal length at 150mm between infinity and MFD. Below are shots of a test target taken with the 40-150 @ 150mm with the focus set to infinity and MFD. If the thin lens approximation was reasonable, we would see a siginificant decrease in the target size. Instead what we see is almost zero change in the target size, at least to within the precision that the out of focus target will allow us.

Infinity Focus | f/22 | Subject Distance ~8ft

MFD Focus | f/22 | Subject Distance ~8ft

I think the problem is basically that we are talking past each other rather than substantively disagreeing. Let me try to clarify what I am and am not trying to say.

What I am trying to say is that the 40-150/2.8 set to 150 mm and shot at its miniumum focus distance has a reproduction ratio, and thus a field of view (expressed in linear rather than angular terms, e.g. in millimeter à la Christof/CrisPhoto), that is identical to that of a thin lens with a focal length of 100 mm shot at the same focus distance. It also has a reproduction ratio/linear FoV at those settings that is approximately (but not exactly) the same as that of an old-fashioned, reasonably symmetric 100 mm lens that focuses by moving the entire array of lens elements in and out, without any use of floating elements.

What I am not trying to say is that either a thin lens with an FL of 100 mm or the 40-150/2.8 set to 150 mm has an effective focal length of 100 mm with both shot at a focus distance of 0.7 meters. If we define the effective focal length of lens X shot at focus distance Y as the focal length of a lens that gives the same angle of view as X shot at Y when shot at infinity, then the effective focal length of the thin 100 mm lens at 0.7 meters is F(1 + R) = 100(1 + 0.21) = 121 mm. For a thick and potentially asymmetric lens like the 40-150/2.8 set to 150 mm and 0.7 meters, the effective focal length instead becomes F(1 + R/P) where P is the pupil magnification (the ratio of the diameter of the exit pupil to the entrance pupil). But since I don't know what F and P are in this case, I can't calculate the result.

Since the effective focal length is what determines the angle of view, it follows from the above that a thin lens gets a narrower field of view at short focus distances than at infinity. The same is true about an old-fashioned lens of the kind I described above. Your experiment indicates that this is not the case for the 40-150/2.8 set to 150 mm. Instead, its fairly constant angle of view suggests that its effective focal length doesn't change much as you go from infinity focus to minimum focus distance but that the focal length, F, declines (since R/P surely increases).