Oly 24 i.e. my first alt v. Canon 24L TS-E II (no movements) my most recent

SLR/c, tripod, mlu, delay, f8, 1/125, 2.5x angle finder focus @ #5, AB400 bare bulb (mixed with flouescent & window) RAW-jpg @ ACR defaults, no PS adjustments or sharpening, 100% crop screenshots

Focus @ #5 @ appx. 1 foot to film plane.

ruhroh.

The highlited frame is the Canon, what variation is your Olympus lens (mc)?

arstein wrote:
The highlited frame is the Canon, what variation is your Olympus lens (mc)?

Yes, (except for the first one) #1 is the Oly #3 is the Canon (blades are a dead giveaway). Oly is MC version. Top is Oly in the full scene shot ... a touch wider than the TS-E.

I'm continually astonished by the performance of my OM lenses. I'm also continually puzzled that a large number of $1500-2000 lenses from Canon and Zeiss can't maintain edge-to-edge sharpness on a 5D Mark II. The OM lenses are much more consistent in this regard.

Yes, Oly seems to make it a point to have very even edge-to-edge sharpness.

When I compare my Oly's to my Nikons in particular, it is very obvious to me that other companies have intentionally favored central sharpness (which will be sharper than the Oly in Zone A) at the expense of the edges/corners. I think this isn't a "can't", but rather simply a choice regarding their drawing style they incorporate into their lens design ... not unlike other choices, such as Cof vs. flat, etc.

But, as with all optical compromises, the Oly's tend to vignette more than some others until you get stopped down, and have a touch lower contrast. I find the contrast/tonal values between these two rather interesting.

@RustyBug - yes, the Oly wides do have strong vignetting -- the 18mm in particular, even at f/8. The lower contrast or more "open" drawing style is actually pretty nice for post processing flexibility!

+1 @ Mr. Joe ... the 18mm, even at f/11 in the extreme corners "tips". +1 again @ PP flex. Takes a bit different 'recipe' to bring the Oly's into the same contrast realm as others might start out with a bit more sooc, but definitely very maleable as a result.

I DO like the 18mm, even with the pronounced vignetting. I wonder how it compares to the Zeiss ZE 18mm in sharpness near the edge of the frame.

The only Oly I've owned that was an exception to this contrast style was the 28mm f/2 -- that thing had some snap. I only sold it because I typically prefer a wider focal length.

I have an Oly 28 f/2 which is a little gem. I'd expect the 24 f/2 to be good too and with it's floating element do better near mfd than the f/2.8 version.

What was the idea of comparing from two different points of view?

alundeb wrote:
What was the idea of comparing from two different points of view?


Lens change only, no camera/tripod reposition (didn't anticpate the variance).

The TSE is much longer than the Oly, so while the film plane/camera position remained constant, the front element is certainly closer (i.e. greater magnification/different perspective). I thought about reshooting before posting these, but felt it still illustrates the difference between the two lenses (@ unchanged shooting position).

I'll reshoot with the AOV better approximated and/or working distance equalized. Will be interesting to see those as well. It's a starting point for comp of these two lenses, not the final analysis.

RustyBug wrote:
Lens change only, no camera/tripod reposition (didn't anticpate the variance).

The TSE is much longer than the Oly, so while the film plane/camera position remained constant, the front element is certainly closer (i.e. greater magnification/different perspective). I thought about reshooting before posting these, but felt it still illustrates the difference between the two lenses (@ unchanged shooting position).

No, no, no. This is a categorically incorrect interpretation of focal length. A lens's physical dimensions does not directly imply a change in focal length. You should know this already just by looking at how certain zoom lenses like the EF 24-70/2.8L are designed--their shortest focal length corresponds to the longest barrel length.

The framing of these two lenses are different not because the TS-E is a bigger lens than the Olympus. It's because you're shooting both of them near MFD, where the nominal focal length matches actual focal length only when the lenses are focused at infinity. You've forgotten that focal length of an internal-focus lens may change as a function of its focusing distance (focus breathing). Thus, you cannot assume that both lenses are actually shooting at 24mm when you've got them positioned so close to your subject that the resultant change in focus affects the subject magnification, much like a macro lens behaves.

Furthermore, you also have to consider that when the manufacturer states a lens is "24 mm", that doesn't necessarily mean it is exactly 24mm. It could actually be a bit wider or a bit longer, say 24.39mm or 23.94mm, if we were to go by the patent filing for the optical formula. At long focal lengths, this difference is negligible; at shorter focal lengths, a fraction of a millimeter may actually be visible. But this "rounding error" is unlikely to be a significant contributor to the field of view discrepancy you are seeing here.

'I'm also continually puzzled that a large number of $1500-2000 lenses from Canon and Zeiss can't maintain edge-to-edge sharpness on a 5D Mark II.'

Mr Joe, what proportion of photographic compositions do you believe are 'unsuccessful' (a judgment call obviously) as images, due entirely to poor corners? Any examples? Genuinely curious, Phil.

@philip_pj -- I've tried multiple copies of various Canon and Zeiss wide angle lenses on the 5D Mark II with surprisingly poor results. The main problem was lenses that are consistenly soft on one side or the other, even at f/8. More info, links, and examples here: http://www.joereifer.com/words/2010/11/15/canon-5d-mark-ii-wide-angle-lens-sharpness-issues/

Joe - if you found so many 'poor' copies across 2 manufacturers then I'd suggest there is an issue with your setup. Perhaps your camera mount or sensor is slightly off axis, or there is some other issue. To discover so many bad copies of the 2 lenses is very unlikely.

Mr Joe wrote:
@philip_pj -- I've tried multiple copies of various Canon and Zeiss wide angle lenses on the 5D Mark II with surprisingly poor results. The main problem was lenses that are consistenly soft on one side or the other, even at f/8. More info, links, and examples here: http://www.joereifer.com/words/2010/11/15/canon-5d-mark-ii-wide-angle-lens-sharpness-issues/

The issue in your link is not that $1500-2000 lenses from Canon and Zeiss can't maintain edge-to-edge sharpness compared to Oly lenses. It's just an issue with "bad copies" of your lenses or camera. Both those lenses in your test have very nice edge o edge sharpness. At least when using good copies of both the lenses and camera.

Mr Joe wrote:
@philip_pj -- I've tried multiple copies of various Canon and Zeiss wide angle lenses on the 5D Mark II with surprisingly poor results. The main problem was lenses that are consistenly soft on one side or the other, even at f/8. More info, links, and examples here: http://www.joereifer.com/words/2010/11/15/canon-5d-mark-ii-wide-angle-lens-sharpness-issues/

I would also guess that either there is something wrong with your camera, or your testing methods. I have 7 wide angle lenses from Zeiss & Canon, including those two you did test. And all my 7 lenses have been good from the first copy

Mr Joe, thanks for your reply. My inquiry is to do with actual images we shoot, because some of us are much more sensitive to 'poor corners' than others - here at work, I have selections of my images scroll each minute in Win7, and I see very few that actually need perfect corners. Like most, I tend to choose lenses with good corner performance, and I am curious about the experience of others here, since I am familiar with most of their styles.

You, as Lars and Skumar25 point out, have an altogether different problem. I also have very few problems with 'bad' copies, but as Markus from Photozone says in another thread from today, he sees up to 25% as bad copies. And we read about it all the time...so it's another, but (quite separate) interesting issue. I also would look into it further, in your shoes.