Filter question

Apologies for an unusual and perhaps nerdy sounding question, but I was wondering if anyone knows if there is a graduated filter available that will progressively remove blue, and increase contrast, towards the top of the image? Does anyone know if such a thing exists? Or any ideas how I could make one? Ideally, I would like to find a filter that is graduated from the middle to both edges, so the sky and foreground are not affected.

A consistent problem with aerial photography is that the image gets more blue, and loses contrast, as the scenery recedes into the distance, due to the increasing amount of atmosphere between the camera and the ground. Also, if I select a part of the image from the horizon to halfway to the foregound, the histogram only covers perhaps a third of the range from 0 t0 255, for example from 80 to 160, sometimes much less than that. I can't use a polariser because the windows are laminated (it causes prismatic effects). A graduated orange/amber filter would block blue light, but also green light, which is not desirable. Tiffen do colour grad nd filters, but they are incredibly expensive (£257.99) for the square ones, which is what I would need, and I am not sure exactly what colour or density is required, so it would be a very expensive experiment. Apart from that, I can't find anything online. At the moment, I use Lightroom grad filters with a light orange hue, which helps.

Am I right in thinking that a filter on the camera would allow more information to reach the sensor, and therefore allow more latitude to tweak the picture in PP?

The filter would end up allowing less info getting to the sensor. But your intent is to reduce only unwanted light, while having the sensor work with only those frequencies you are wishing to capture. That means you can expose specifically for that desirable part of the spectrum, without blowing the image with light from the unwanted parts of it.

As to how you would make your own, not sure - sorry!

You raise a good point

Adobe Lightroom ( LR ) allows the variability of various colour saturations and luminance.

However while it's graduated filters do allow contrast adjustments they do not allow graduated colour adjustments.

I can't see why any LR parameters can't be applied to a graduated filter, or a localised adjustment brush. It's only software after all.

Lets hope Lightroom 4 does allow this.

Graduated filters do allow for progressive reduction in part of the colour spectrum, that's how filters work. For example, a graduated warming filter will remove more blue at its densest part, this making the sky progressively warmer.

There are likewise graduated colour filters for black and white photography that increase contrast in the sky, but leave the foreground relatively untouched.

I would suggest you search eBay and camera fairs for what you want at low prices and old and new filter catalogues for ideas as to what has been/is available. Pentax made some oddd double-coloour filters at one time for special effects, so all sorts of oddments can be unearthed.

Pity you didn't post this when I've had lots of odd filters on eBay, but sadly they are all gone now.

Thanks for the replies! I will trawl ebay as you suggest John. A shame that I missed your sale of filers on there. I only thought about using a colour grad filter the other day while messing around with colour filters in Lightroom. I have used a standard nd grad with some success but a tinted one, if I can get the right colour, would I think give much better results. Aerial haze actually looks much worse on a photograph than to the naked eye, I guess because our eyes pick up a much greater range of frequencies. However, if I can find a way of reducing the blue tint caused by this haze, I would imagine I would get more contrast and detail, as there should be less scattering of light.

I should have an IR camera soon, and am looking forward to trying that out, aerial haze does not affect IR anywhere near as much as it does the visible spectrum. The only downside is the pics will all be mono. I did have the probably completely nutty idea of trying to take an IR and a colour picture of the same scene, converting them both to Lab Colour, and taking the Lightness channel from the IR picture and replacing the same channel in the colour pic with it. I think it would be extremely difficult to get the replacement channel to line up though, especially as the cameras have different resolutions. I would need two identical lenses too. You reading this, Richard? If anyone here could work out if this would be possible, my money would be on you

Polarisation filter might do the trick since it also cuts the haze somewhat making things more vibrant with more contrast.
But sadly the effect is not so even over the image...

As I said, Stefan, I can't use a polarising filter as I am shooting through laminated glass; the lamination causes coloured bands that look like oil on water. For the same reason I am not allowed to wear polarised sunglasses at work.

Sorry to have read over that part.
Haze filter is the next best thing then, they cut out some of the blue light i believe and a big chunk of the UV.
But i'm not sure what the effect really is on digital... it might not be worth it.

And they aren't graduated but that isn't so much a problem.

Just found this, maybe it helps.
Note that it's also possible to mitigate haze to some extent in post-processing, with or without an optical filter. Applying unsharp masking repeatedly to all channels or to the blue channel alone with very large radius and very low intensity (amount) settings can be surprisingly effective.

And there are more things you can do in PP.

If you use filters on digital remember not to use AWB, otherwise whatever the filter colour is may be "corrected" by the camera. Use daylight or another setting that suits your style.

Lee Filters are well known filter manufacturers and do supply the movie industry. They have an "Ask A Question" section on their website.

Thanks for that tip Stefan, I will give it a try, never came across that before.

John, I don't bother setting the WB at all, I leave it in auto all the time, as the colour temperature is so variable, anything from 4500 to 9500, depending on time of day, amount of haze, amount of water or sky in the shot, and no doubt many other factors too. I shoot in RAW and then set the WB in Lightroom. Often I have to select the further scenery as a heavily feathered seperate layer (or even several layers) and adjust the colour balance, usually adding red and reducing blue, to get the colour looking right. Would it be better to set a mid-range colour temperture on the camera?

Thanks also Graham, I was looking on their site but didn't realise that they had a question and answer section.

This thread was brought about by my incresing frustration at having hundreds (probably thousands) of pictures of amazing scenery, especially in Turkey, Iran, Asia and eastern Europe, and not being able to get rid of the haze and variable blue cast satisfactorily unless it's a particularly clear day, which is not that common. It's also my main motivation for trying infrared.

For contrast it's very easy to stretch the 80-160 to 0-255, it can be done with level or with the curve tool you find in many programs but you probably already knew that.

As regards WB, I would be aiming to relect the nature of the ambienmt light and would shoot JPEG with a fixed Daylight WB.

If you use AWB the camera will try and correct to what it sees as a neutral grey overall, so the effect of any dramatic natural lighting and indeed of a filter will be reduced or even removed.

If AWB can correct the filter then surely the white balance can also mimic the filter so why use a filter to begin with?

Jpegs don't work well at all from high altitude, there's so little dynamic range recorded and far too much processing involved. I have tried them and they just get messed up in processing. As Stefan said, if WB can reduce the effect of a filter, can it then mimic a filter? If so, what colour temp would be worth experimenting with? I would guess towards red? What's the highest colour temp i could realistically use and still get full dynamic range? I might try that, then adjust it afterwards in PP.