Barrel or Pin Cushion Distortion

[arith]

My lens has some pretty severe barrel distortion at 17mm reported as +3.32% and so am I right to dial in +3.32% correction in PSE7?

Also is this a bit over the top? I find it hard to think of ways to make a scene look interesting and rather than just come away I do bazaar things like this.



cheers

[carregwen]

My lens has some pretty severe barrel distortion at 17mm reported as +3.32% and so am I right to dial in +3.32% correction in PSE7?

Also is this a bit over the top? I find it hard to think of ways to make a scene look interesting and rather than just come away I do bazaar things like this.

Steve

Not sure what's in PSE7, but in CS4 you use lens Correction under the filter menu. You get a grid of lines which helps you decide when you have corrected enough. You can see the full size if you click on it.

The composition needs a figure (or perhaps a dog) with their back to you looking at the scene themselves. Or perhaps a group of three people talking in a circle.

[arith]

Or a gardener maybe, I'm always trying my hardest to avoid people. That utility is exactly the same as in PSE7 and I didn't think of using the grid lines. In fact after saving an image with +3.32% correction added I tried to copy it in Helicon Filter and came out with -12 as the same by superimposing one over the other in Gimp, why; because Helicon Filter is 16 bit and PSE7 isn't.

All lenses are different though. cheers

[Peter Ryan]

Or maybe crop it a little tighter.

[arith]

Or maybe crop it a little tighter.

Was my 17mm day, What I did wrong was to try to keep the lantern on a third and crop the chairs a little, I thought it would be easy to clone them out but it turned out quite hard to do. If I crop too much I'll end up doing the same kind of photo's as everybody else and all I needed to know was if it looked distorted, to me it does.

Whole load of questions involving wide don't get addressed; like decentering, a word I've never seen before. Apparently LW/PH test results are not necessarily from the same image, so I wonder what the point of them is. The Canon ef-s is better but why is it £200 dearer than the 17-40 f4 L?

This is my other effort:
http://public.fotki.com/Arith/nation...87-hf-jpg.html

[Peter Ryan]

I could see you trying to keep the lantern on the thirds but maybe you need to stand back further (if you can) and take a number of shots and stitch them together and then crop for the effect you want.

[arith]

Cheers, I'm trying to avoid stitching since the equipment is too heavy and bulky. I sort of go out with a specific job such as try to think 17mm and avoid doing anything else except in one case where I previously planned to test CA but it was a cloudy day. There shouldn't be any visible CA except at 17mm and even then it isn't much, it looks green against green and what red there is is so little it isn't worth the bother trying to correct it.

cheers

[Colin Southern]

Hi Steve,

Just a quick thought ... when shooting wide-angle light this, be sure to keep the plane of the sensor absolutely vertical to minimise the distortions you have to correct later.

[arith]

cheers Colin.

[Peter Ryan]

Hi Steve,

I understand what you are saying. Sometime I also chose a lens and just go and shoot with it. Good fun trying to make your vision suit the lens.

[arith]

Hi Steve,

Can you let us know what lens you used here? BY the way, I like the mounting effect you use.

cheers; The Tamron SP 17-50mm F2.8 XR Di II VC IF Aspherical LD what a lot of letters but the important one is VC which means it is the latest one that is naff at auto focusing. I just found out that testers don't necessarily use figures from the same image but choose the best from a set of images, when I was looking I was impressed with the figures for CA and compared it to the 17-40mm f4 L for resolution bearing in mind my camera has best resolution of around 2250 LW/PH and I bet you can't easily see differences of 50 or so.
At 1.25% barrel distortion at 24mm this lens performs better all round than the dated canon ef 24mm f2.8 prime although 3.32% is a bit much at 17mm I had to have a wide lens or go mad and I could part exchange for this lens which cannot be more than 6 months old against my old camera that I started to practise wet sensor cleaning on and was likely to break by taking it out in the wet. Is that too much information.

Here's some more; I should have done this earlier but I'm crap at focusing. Brick wall at f2.8 and f6.3 the latter using live view to focus and actually I don't know.

[arith]

Actually when I straightened this up it came out as less than 3.32% distortion.