MOL's Head
Stuff from old Mols Head about technology

 
 Misc. Technique  

  Sell your images 1 - resolution

Microscopy images can be worth a tidy sum to science libraries, magazines, and other publishers. Whenever there is topical debate on some 'news-worthy' issue, the media fly into action with picture researchers buzzing the internet looking for prime examples of whatever it is causing the public storm. With global warming increasingly on the media agenda, and armed with the knowledge that the planet is more controlled at all levels by microscopic organisms (algae?), your microphotographs may be worth a small fortune in the future.

There is only one small problem:-

If your image has been taken using a consumer digital camera, the publishers will not consider the resolution to be of sufficient quality for their printing methods!

Consumer Market Digital Cameras
Most digital cameras, designed for the home consumer market, up until recently (2004) would deliver a 300 DPI image at 6x4 inches.
A 3.4 Megpix camera is typical of home consumer digitals. The internet uses images around 72 dpi. Magazine Publishers typical use A4 size images at a minimum of 300 dpi.

The means that although your images are suitable for internet (72 dpi) use and small printed photos (300 dpi) at 6x4 inches, a full page version for a professional magazine requires larger sizes at 300 dpi or higher. It is possible to resize your image using Adobe Photoshop or other imaging software, but the process used decreases the quality of your picture by increasing blur and pixellation. There is no way out of this using conventional image techniques to manage image size and resolution, except to purchase a more expensive camera with a higher definition ccd - a 6 megpixel camera for example!

To the rescue - 'Fractal Technology!'
I did say it was impossible to solve the problem using 'conventional techniques' but it is solvable by using fractal technology. Any image is naturally made up of parts of itself. Nature has a way of reiterating its shapes and forms, sometimes bigger or smaller, but always similar in its constructions. The curvature of your shoulder or arm will be repeated else-where on your own body, possibly in your legs, the curve of your cheek, the heel of your foot - albeit, at a different scale.

Fractal imaging software exists which can analysis your picture, break it down into a list of reiterated forms and then use these forms to reconstruct your image faithfully at higher resolutions and sizes. It is important that you don't use your image editing software to sharpen your image prior to using fractal software to analyze it. Doing so, would introduce false artifacts (fringing) into your image, which would be greatly amplified and thus more visible in scaled up versions.


Genuine Fractals
This is the brand name for an excellent software package that will provide fractal enlargements and higher resolution versions of your images acceptable to magazine publishers. A trial version is available so you can see exactly what I mean before purchasing a copy.
Take a look at the product here:
http://www.lizardtech.com/solutions/gf/


An example:-


This is an image taken from this month's micscape. Its at a resolution of 72 dpi for the web. If we take a small section of it and blow it up using conventional software, we get a result similar to the image below left.

As these are jpeg images, there are disturbances in the image where color and shapes touch each other, around the petals for example. This is caused by the compression process in the jpeg file itself.
   



The image on the right has been produced from the image above. It is a blown up section of a petal. This one has been done using Genuine Fractals by Lizard Software. Notice how the edges of the petals have kept their precise form and shape!
The small blow-up below is from the image on the left. It has been created using photoshop to increase the image size.





Links
Off-site review of 'genuine fractals'

Terminology
dpi / DPI - dots per inch
pixellation - see the dots or blocks that form the image

 Mol