Home My photos purchase prints What's new Links Search Contents My tools Contact

infrared technique

All images on this website are copyright 2004 Gary & Anna Ayton. Contact me if you want prints from the originals.

 

Infra-red photography with the Olympus C8080 digital processing

The best way to explain what is needed to process infra-red images is to demonstrate it:

Here is a landscape photograph taken in full sun with the sun to the left. A Hoya R72 filter and the camera is set to B&W photo mode & SHQ jpeg, in this case the exposure used was f/4.5 1.6secs at 125 ISO:

Now this is a bit flat in contrast terms, so lets use Adobe Photoshop Elements to adjust the brightness levels histogram in the usual way:

OK, this is looking much better, but in this particular photo, the lens flare of the IR light has become even more prominent with the distracting central bright spot. 

NB. this bright spot in the centre only is noticeable on the this camera on wide angle focal lengths in certain scenes with high IR.

So how to fix this problem?

Well, you could just use a sunburst gradient type mask layer in an image editor and subtract it out, but this may not give the best results as the mask needs to match the lens vignetting reasonable accurately.

So here is a fix taken from the astrophotography guys where image "flat fields" (ie. same brightness throughout the frame) is important and often a problem with telescopes.

What I did is take a photo of an evenly lit surface using similar focal length, focus point and f/ratio as the original image so as to get an accurate image of the true vignetting effects (actually I cheated and didn't use the same focus point for this demo and I used f/3.5).

Here is the resulting image:

Now you can copy this as a layer onto the original image and subtract the layer (or add the invert of the layer - whichever looks better) while experimenting with the layer's opacity level to achieve a reasonable reduction in the uneven brightness.

Finally, I used the brightness levels histogram to get the final image:

This is much better, but there is a problem that all the image manipulations that needed to be done have substantially increased the noise in the image. To minimise this, it would be better that the original was taken as TIFF or RAW and converted to TIFF before performing this image manipulations.

Now I am sure there are many other ways to handle this problem, but I am not the sort of person who will spend hours playing with an image and so I go for the simple things in life.

Hope this helps.

Gary Ayton