Shooting through angled glass.

I am wondering if we have any experts here on the optical qualities of glass? More specifically, the effect of shooting through laminated glass at different angles, and how to mitigate these effects? As some of you are aware, I shoot a lot of photographs from aircraft, and although I am getting there as far as PP work goes, I still have several things I would like to improve. One of the main problems is that I am often shooting through the front windows, which are generally angled at about 45-70 degrees to the plane of the lens, depending on if I am shooting the sky or the ground. I can generally get around the problem of reflections by shielding the lens, and sometiimes by placing the sun-visor over reflecting objects. A polariser is no use as the glass is laminated. However, pictures taken at these acute angles generally look quite soft, especially at longer focal lengths (at extreme focal lengths (200mm+)focus is not possible at all unless the lens is perpendicular to the glass, as is the case with the shot at 250mm of another aircraft in my portfolio here). I have experimented with apertures, but it seems wider apertures help with removing the effects of dirt and scratches on the glass while smaller apertures generally seem to work better with longer focal lengths. I use single point focusing or manual focus. I have a UV filter on the lens, mainly for protection (the windows filter UV anyway). I was wondering if anyone here has any ideas how I can get better quality shots: one thing I have been wondering is whether the size of the the front element of the lens will affect the quality of the photograph. At the moment I use a Sigma 18-250, which has some compromises obviously with it's large range, but it's that range that makes it such a good lens to use on the aircraft (changing lenses is not really an option). I've not noticed much improvement in quality using different lenses anyway. Shooting through the side windows I don't have the problem of angles as much, but the glass is not as good optically.

Another problem is that pictures taken from the aircraft (especially in haze and at high altitude) have a very short dynamic range, often only from about 50 to 200 when viewed on the histogram, or 70-230 when I expose to the right. At the moment I get round this by use of selective levels adjustments in CS5, but am wondering if there is anything else that can improve the dynamic range? I used to have a compact that shot TIFF's that seemed to handle dynamic range in these conditions better than my K7 does with RAW.

If anyone has any ideas I would love to hear them! At the moment I am not sure if I need to change my technique, buy a new lens, or maybe even a new camera! Or perhaps just accept that what I am getting is a good as I can get? However, I am sure I can improve somehow, and would appreciate any input.

I would try two things. First use a ND2 (or higher if possible) filter for getting extra contrast and then use a rubber lens hood that you an put flush to the window, this may help with the angling and the reflections.
However, the glass you are shooting through is under stress and this will probably show up in one way or the other and even by using a polariser you may find that you get 'stress lines'. Using the rubber hood directly against the window may reduce this effect.

An ND filter simply reduces the overall amount of light, so has no effect on contrast.

However, the glass is the problem and it may be scratched and dirty and will certainly be an optical mess. So, wide apertures will defocus the glass and keeping the camera wello away from touching the plane will reduce vibration to the minimum. High shutter speeds may dictate high ISO values.

I'd remove that "protection" filter as well as it's not helping quality and is unlikely to afford much protection. The lens hood is always essential. Failing that - small plane, fly low and open a window....

On a Cessna plane. you can undo the metal retainer that permits only partial window opening. That way, the window will open fully and allow you much more latitude when it comes to shooting. If you do remove that metal strap, be careful when you open the window so it doesn't hit the bottom of the wing. After it's open, the window will stay open by air pressure at about 1 inch from the wing. Be aware that rear passenger will hate you...

Thanks John, Yves and SmutjeUK for the replies!

As it's a commercial airliner, opening the window is not an option. The windows are generally quite clean, new and fairly blemish free, it's the angle and the thickness of the glass (about 3/4 of an inch) that seems to cause the most problems. It's also laminated, in three layers. Using large apertures to reduce the effects of the glass may be a bit of a Catch 22 situation, as the Sigma 18-250 is quite soft wide open. However, I'll do some more experimenting. I will also try removing the filter and see how much difference that makes.

I have done quite a bit of research on PP techniques, but can't find anything about correcting refraction. I don't know much about optics, but one thing I have wondered is if a tilt-shift lens could reduce refraction by reducing the angle between the lens and glass? It would be an expensive experiment so I'd rather understand more about how such a lens actually works before investing in one. I am not sure how quickly they can be adjusted either.

I find it quite strange that my old compact seemed to handle haze better, it was nowhere near as good a camera as the the K7. I've not been able to find a way of getting around the limited dynamic range at high altitude, especially when it's hazy. I can only guess that there's something I am missing while processing RAW files: with the compact I just used auto settings in ACDSee, but I can't use the version I have with RAW files unfortunately. At the moment I use layer masks in CS5 and correct the levels selectively in different parts of the picture. However, I often end up with strange colours and lots of noise. If any of you have any tips on a better way to do it I would much appreciate it.

Probably a better lens then is a good place to start. The compact may look better because the image is already quite heavily processed, but with the right lens and Photoshopping your DSLR will be better.

A shift lens will not help. You cannot do anything about the glass other than focus beyond it and boost the contrast in PS. A wider aperture than the Sigma 18-250mm offers will probably help more.

The glass of the window is presumably acting as a prism, slightly dispersing the light from white into its different components (RGB, etc) . Maybe you need to speak with someone who could advise whether another prism could be used to to bring those components back together again.

Or maybe I'm talking a load of cobblers

You just need to tilt the plane so the glass is parallel to your subject

was your compact shooting at 200+mm? could be vibrations and or motion blur... try a gyro stabilizer...
Try any nd grad filters? seems to me something along this line:link or maybe two nd grads one upside down to darken the top and bottom of the frame.... colored grads maybe also of use...
and the above mentioned rubber lens hood... against the glass...

No Don my compact didn't go to 200mm.....that being said the majority of shots are wide angle. However apart from turbulence there is no vibration at all in the flight deck. The stabilisation in the lens seems to work very well, in fact better than the cameras. One of its main advantages is that you can see when the image is steady. I think I might get some funny looks at work if I turned up with a gyro-stabilizer....

I have tried nd grad filters but I try to keep the iso down and shutter speed fast as much as possible. They are sometimes useful when the sky is much brighter than the ground. When I finally get a K5 I guess ISO won't be such a problem. The handheld hdr looks interesting too but it's a shame it's only with jpegs....

I had some success using old hard drives as gyros.... you just need to power them up and tape them to a quickflip type camera bracket... was using it as a robotic experiment....

Have you tried haze/UV filters? I can't say that they will work in your scenario because I've never been at the pointed end of a commercial airliner, except Concord, but had no SLR handy at the time.

Haze/UV filters have to be good at something because they've never been any use to me in a hazy scene.

I've only ever taken pictures through the scratched 'n dirty plastic side windows of a Piper Warrior, so not really a parallel there.

I'd suggest trying a better lens, but do you really need the 18-250 range of the Sigma? Which focal lengths do you find you actually use?

Might be barking up the wrong tree here so dont "shoot me down", so to speak.

When shooting throught the front windows, the surface on top of the flight instruments will reflect on the windows into the lens. Car manufacturers usually select dark, non-relflecting materials for the top surface of dash-boards for exactly this reason. Could you place a matt-black card on top of the instruments before shooting.

Also, place the lens as close as possible to the glass as this will further reduce the effects of any flaws/marks on the windows.

Thanks for the suggestions! I have had a UV filter on the lens which I can't see makes any difference. As I said, the windows filter UV anyway.I think John is right that I should remove it. I like the prism idea but can't see it would be practical! I think adding more glass to the problem is probably not the way to go...

As both Peter and John suggest,a better lens would be a good idea but I do use all the focal lengths. After going through some of my pics I see that some of the better ones were taken with my Sigma 24mm super-wide at f2.8......Maybe I will just have to change lenses when needed. I just need to find a compact 28-200 f2.8 lens...is there such a beast? Thanks for the input too, Rab, the glare-shields are painted with non-reflective paint. I often use the dark plastic sun-visor to shield reflections though.

I think apart from changing lenses improving my Photoshop skills is probably going to reap the most rewards. I guess it is theoretically possible to write some sort of correction software for correcting refraction...any computer gurus out there who fancy designing me a plug-in?

Just a thought about the compact Ihad. It shot Tiff so the pictures had to be processed, which I did in ACD See. Before processing they looked similar to the Raw files I get now. Do any of you use this program? I can't find a Raw dng plug-in for it but would like to try it again. The picture I entered in the Curves comp was processed in ACD See from a Tiff I converted in CS 5. It came out better than a version I did in CS5 but was quite noisy.