Many people tell me that they love the photographs that I have taken and the one main question they have is where those skills came from.
In 2000, I was trying to take a picture of a building for a paid website project but I could not get the image to look how I wanted it to look. It just would not look right, no matter how I took the picture. I am often my own worst critic when I am working on something because I believe that if you do something, then it has to be correct otherwise there is no point in doing it.
I went to the library, got some books on photography, and spent a whole day making notes, reading tips and tricks, but it did not help.
I considered abandoning the whole project and forgetting the work. Then I remembered Tony Hart. He was one of those decent people whom I could trust and respected as an educator. I gave him a call and he gave me a date and time to see him.
I showed him a painting of a building, and then I showed him a photograph I had taken and asked how I could make the photograph look like the painting.
His first reaction was, "Well, I'm not a photographer, but what I think you're trying to achieve is correct perspective."
He spent an hour explaining to me in very simple terms, and with the aid of sketches how impressionistic painters draw, and how to correctly set the perspective. After an hour, I had a very good idea and knew exactly how to point the camera to get the correct shot. He was an amazing teacher...
What camera do you use?
I use a number of digital cameras from various manufacturers. I normally take them apart to find out which chip and controller it uses. If it is using an ARM processor based on RISC architecture then there is a good chance that I can hack it and make it work how I want it to. I only use cameras where I can change their firmware and preferably rewrite their code. I have opened up a few in my time to learn about digital camera technology and engineering, so I have a good idea of camera basics and some low level principles.
I possess good mathematics knowledge and some microcontroller skills mainly through ATMega32 Microcontroller Projects. Having worked with commercial microcontrollers, I enjoy rewriting their firmware.
Originally, I was using the discrete Fourier algorithm, as shown above, for the CCD imaging and sub-sampling, but realised that Laplace and the Z-transform were giving me just as good results. I used the Z-transform for some of my earlier work, but being a perfectionist I was not happy with it and decided to change the algorithm.
I later decided to shift to the Laplace transform, but the colours were too light for my needs. It was almost as if there was a temporal shift in the spectrum. After a lot of experimentation, I decided to settle on the fast Fourier because it offered the kind of images that I was looking for, and I was able to reuse my assembly library that I had previously developed for other projects.
What kind of lighting do you use?
Whenever possible I use a tungsten filament bulb indoors. I reflect the light using two long mirror strips and focus them onto a single point through a diffuser assembly. The diffuser assembly consists of white curtain netting layers rotated by 90 to depolarise the reflected light. It seems to work... It will do for the time being anyway.
Are you a professional photographer?
I do not know what a professional photographer is. Is it someone with a lofty title who belongs to a photography guild? Is it someone with the right connections in the photography industry? Is it someone who possesses fancy expensive SLR cameras with big lenses? Is it someone who attends Pro Photography events and gives speeches on the subject? I do not do any of those things. I am only interested in the engineering principles and the basic mathematics involved in photography, as a hobby and just for fun of course.
The way I see it, any fool can take some holiday snaps, and upload his 300 MB GIFs onto a website. Google will even make such a person a 10 for 10 publisher. However, what I do is more difficult to achieve and takes great skill. It takes even greater skill to recognise, and there are very few people in the world to recognise it! Due to the prejudice and corruption, even fewer to give credit where it is due.
I also use my own JPEG and GIF CODECs including the core algorithms that I wrote myself in C++ back in the 1990s. I was experimenting with my own pallet optimisation algorithms before GIF had become a well-known standard. Back then I was working on getting the best surfing speed on a mere 32 kbit/s dial-up modem and the idea of removing un-necessary colour data was the first thing most people thought of including pallet optimisation.
Many people even wrote their degree thesis on fractal encryption and image processing - which I read and had a good laugh. It is always nice to see people following my footsteps and getting degrees for it!
You will probably notice that my graphics and images are loading faster than most sites... My CODECs will not be made available to the public just yet though. :-)