Complex, Pincushion, and Barrel distortion on Oly 14-42IIR Lens when used on E-PM1

Hey everyone. The more lenses I add to my kit the more I realize I am really beginning to dislike my Oly 14-42II Kit Zoom. It dawned on me the other day that I could either have possibly the worst copy of this lens, or the E-PM1 doesn't/can't correct for the lens distortion.

1) I shoot JPEG.

2) "Shading Compensation" is turned on in camera. This is supposed to correct for any vignetting, but I figured perhaps Oly lumped any other lens type corrections into there. It makes no difference on the on-camera-LCD turning it on or off, but I wanted to be sure. There are no other camera menu options for corrections, I combed through it about 15 times.

3) I Process in OV2 and GIMP (if needed). OV2's built in distortion correction can't compensate for the distortion, even when it isn't complex distortion, and even then the auto "use lens info" setting usually makes the distortion worse instead of reducing it. Doing it manually in either program usually ends up cropping a ton of the frame out, which kind of defeats the purpose of using this wide lens.

4) I could not find any firmware updates for either the PM1 or the lens which address any distortion issues.

All images below are SOOC, they've only been resized to save bandwidth.

Images 1-3: Watch the Power Pole and Car Antenna

Images 4-6: Watch the Pine Tree and Power Pole

Images 7-11: Pictures taken of my FLAT LCD Monitor showing a "No Distortion" Grid from a page displaying distortion types.

1

14mm Far Focus - Pincushion

2

14mm Far Focus - Pincushion

3

14mm Far Focus - Pincushion

4

25mm Far Focus - Pincushion

5

25mm Far Focus - Pincushion

6

25mm Far Focus - Pincushion

7

14mm Close Focus - Barrel Distortion

8

25mm Close Focus - Complex Distortion. Notice parallel bottom edge, but distorted all other lines.

9

42mm Close Focus - Complex Distortion. Notice parallel bottom edge, but clipped top right, pinched top left, etc.

10

40-150 4-5.6 ED R Lens: 40mm Close Focus - Mild Barrel Distortion, acceptable.

11

40-150 4-5.6 ED R Lens: 150mm Close Focus - Slight Pincushion, acceptable. Sorry about the slightly rotated image.

Where the lens is focusing, at least in terms of distance, seems to greatly affect which type of distortion is present, and how badly distorted the image is. The real world use images (1-6) could have complex distortion too, but it is hard to tell. I found that I subconsciously frame to minimize distortion when using this lens, not sure if that is good or bad, hah.

Has anyone else encountered these kinds of problems with this lens?

Perhaps it only occurs on older bodies. In another thread a kind fellow checked out both this lens and his 14-150 II on an E-M10 and both lenses showed no noticeable distortion in real world use.

Regarding my 40-150, I barely notice the distortion even when I am looking for it, and never notice it in actual photographs.

Sigurthr wrote:

Hey everyone. The more lenses I add to my kit the more I realize I am really beginning to dislike my Oly 14-42II Kit Zoom. It dawned on me the other day that I could either have possibly the worst copy of this lens, or the E-PM1 doesn't/can't correct for the lens distortion.

Someone is going to tell you that your lens is decentered and someone might be right.

The 14-42 II I had on my E-PL3 was an excellent lens throughout its range and on that basis, I have freely recommended this lens to others.

Clearly yours is not up to snuff.

I don't trust the pix from your screen, though. Are you absolutely positive you got everything centered and the camera sensor and screen absolutely parallel? That is a hard ask in my experience.

Hen3ry wrote:

Sigurthr wrote:

Hey everyone. The more lenses I add to my kit the more I realize I am really beginning to dislike my Oly 14-42II Kit Zoom. It dawned on me the other day that I could either have possibly the worst copy of this lens, or the E-PM1 doesn't/can't correct for the lens distortion.

Someone is going to tell you that your lens is decentered and someone might be right.

The 14-42 II I had on my E-PL3 was an excellent lens throughout its range and on that basis, I have freely recommended this lens to others.

Clearly yours is not up to snuff.

Thanks, I actually hadn't heard of decentering in photographic lenses, and as such had to look it up. In laser optics I've dealt with decentering (but it's called something different), but there we're talking about single element lenses in XY mounts, not complex multi-element imaging lens assemblies. The two optics worlds are vastly different, to say the least.

I don't trust the pix from your screen, though. Are you absolutely positive you got everything centered and the camera sensor and screen absolutely parallel? That is a hard ask in my experience.

-- hide signature --

Geoffrey Heard
Down and out in Rabaul in the South Pacific
http://rabaulpng.com/we-are-all-traveling-throug/i-waited-51-years-for-tavur.html

Quite understandable. I wouldn't call them reliable myself, as they were handheld and thrown in last minute. I was reviewing the first six photos on my pc and noticed the distortion wasn't the same on the left and right sides of the frame, so pulled up a graph out of curiosity. I might try again with tripod and see what I can come up with, but even then I doubt perfect parallelism is achievable.

EDIT: I placed the camera on a tripod, affixed a lens decentering test chat (black and white concentric circles) on the wall and took some test images. The instructions on the chart said to defocus slightly and you'll see an uneven amount of defocusing in the direction of the decentering. At all FLs and aptertures I couldn't see any major difference in focus. Something slight on one side in some conditions, but nothing glaring like the example images. Certainly not a conclusive test, but the absence of a glaring positive result is worthy of note, none the less.

I suppose any additional more-telling tests would have to wait until I have an E-M10 and see how the lens performs on that body. Hopefully Oly ships it out soon.

Sigurthr wrote:

I suppose any additional more-telling tests would have to wait until I have an E-M10 and see how the lens performs on that body. Hopefully Oly ships it out soon.

I doubt that Olympus changed its correction algorithm going from EPM1 to EM10. How a lens is coded for distortion is part of the M43 standard, and I would hope all cameras do the same transformations with that data.

At  focal lengths close to 14mm, it's pretty easy to get into tilt and keystone effects unless one takes the care in composition to minimize them. I should take more care, but I often don't, and if I want to save the picture, then it's some work with photo software.

I like my 14-42R, and I am stubbornly using it, hoping that what skills and knowledge I have will overcome its kit lens pedigree. Still, it's a cheapo lens. I can probably buy a used one for $75. We should expect quality fall off in the edges.

baxters wrote:

Sigurthr wrote:

I suppose any additional more-telling tests would have to wait until I have an E-M10 and see how the lens performs on that body. Hopefully Oly ships it out soon.

I doubt that Olympus changed its correction algorithm going from EPM1 to EM10. How a lens is coded for distortion is part of the M43 standard, and I would hope all cameras do the same transformations with that data.

At focal lengths close to 14mm, it's pretty easy to get into tilt and keystone effects unless one takes the care in composition to minimize them. I should take more care, but I often don't, and if I want to save the picture, then it's some work with photo software.

As do I. I process with PhotoLine with cheap but excellent PS compatible plug-in PTLens. When things are critical, though, I make use of the built-in level indicators in the G6 at picture capture.

I like my 14-42R, and I am stubbornly using it, hoping that what skills and knowledge I have will overcome its kit lens pedigree. Still, it's a cheapo lens. I can probably buy a used one for $75. We should expect quality fall off in the edges.

I found it an excellent lens on the E-PL3, already noted. I haven't noticed many (any?) negative posts about the 14-42 II on this forum apart from suggestions that it is not as sharp at 43 as it is at 14.

Sigurthr, good work on the decentering test. This test should have also shown up the problems you started talking about here. They would have pushed/pulled the circles out of shape.

so why bother about pincushion distortion?

Miron09 wrote:

so why bother about pincushion distortion?

They weren't taken for artistic merit, simply to illustrate an observed problem. In fact they were composed entirely out of the need to clearly display the issues. That is why they were made in P mode, which I pretty much never use, as the exposure settings were irrelevant. In interesting photos that are composed and exposed well it is all too easy to overlook distortions - that's just how our brains work. Fisheye is proof of this. If distortions were an automatic rejection point, all lenses would be rectilinear.

If you don't like the photos, that is fine. They weren't meant to be liked. I have actual real-use photos I can post using this lens if requested. Even if you don't like those, that'd be your prerogative, too. If you don't like that I brought the topic up, well, I apologise? Verifying one's observations with others is the basis for which knowledge is built upon, and the entire point of a forum.

The point is that when a rectilinear lens no longer performs rectilinearly, there is a problem.

Hen3ry wrote:

Sigurthr, good work on the decentering test. This test should have also shown up the problems you started talking about here. They would have pushed/pulled the circles out of shape.

Thanks! I'm going to redo the tests the next sunny day I get a chance on, as I think having to run a higher iso last night might have caused false negatives. If it comes back again inconclusive/negative I wonder if there is even a flaw in the lens, or if it is just all in body. I suppose I should shoot some RAW files as well for a more thorough test.

Side note, I did hear back from Oly today that I should get a shipment notification soon for my M10. It's really tempting to cancel/rebuy with the $50 kit lens, but I'd much rather save that $50 for my next high-IQ lens.

Right, I wasn't implying that Oly would have changed their algorithm, merely that having a second body would allow me to test whether the problem exists in lens or in body. If for some reason my PM1 is no longer applying the correction properly, that would account for this problem entirely. After all, lenses are supposed to be barrel at wide angle, not pincushion. If the problem exists on both bodies, it is highly unlikely the problem doesn't come from the lens.

Lens does not appear to be decentered by this test:

No focal aberrations to be found. SOOC JPEG

Interesting to note though, I did another handheld close-focus edge parallelism test and it came out as wonky as ever. However.... OV3's Auto Distortion Correct w/ Use Lens Info did fix it. Close focus 14mm results in Barrel Dist., which seems correctable at least. In good weather I'll do another far-focus edge parallelism test and see if OV3 knocks that out as well.

The only thing changed here is the body used, which makes this whole situation rather odd. Neither body seems to be applying correction by default, but the embedded data from the new M10 is apparently usable/correct. As for the embedded data of my PM1, well that's never helped distortion correction at all so far.