Brick Wall Ahead
Why Moore’s Law Does Not Apply to Digital Photography
By Ray Maxwell
I read Christian Sandstrom’s article titled The Rise of Digital imaging and the Fall of the Old Camera industry. I found the article very interesting and agree with a large part of it. However, I have a problem with one sentence in the article. He states,” Bearing Moore’s Law in mind, there is no reason why this trend should not continue.” He is referring to the rapid improvement in the digital cameras and suggests that much of the progress is because of Moore’s Law. At least with respect to expecting the number of mega pixels to double for a given sensor size (full 35 mm frame), this will not happen. We are very near the limit right now.
Gordon E. Moore originally said that the density of transistors on an integrated circuit would double each year. He later revised that to say that it would double every two years. Moore’s Law is still on track with regard to memory and processing chips today. This will allow extra processing features in our cameras like finding people’s faces and rapidly focusing on them. But it will not help problems with putting more megapixels within a given sensor size.
So why do I say that we are near the limit? The diffraction limit of our lenses is larger than the pixel size on some sensors. This means that the resolution of some current cameras is limited by the diffraction limit of the lens rather than the pixel size on the chip. This means that cramming more, smaller pixels on the chip will not result in a higher resolution image when you make a print.
Someone will immediately ask, “Well design a better lens with a better diffraction limit.” The diffraction limit is a property of light. The only way to get a better diffraction limit in photography is to illuminate your subject with shorter wavelength electromagnetic radiation. There is a process that allows integrated circuits to be made with a smaller feature size that uses X-rays that have a shorter wavelength. Gamma rays could also be used. However, there are some human photo subjects that are not in favor of this and object to be illuminated by these rays.
You can find an excellent article on diffraction limits at this URL:
Be sure and move your cursor over each parameter given in the table.
I recommend this entire site for learning the fine points of photo technology:
As an example, I have a Canon 5D Mk II. If I enter the following values in the Diffraction Limit Calculator I find that when I stop down more than f/11 the Airy disk of the lens exceeds the size of the pixels in my camera.
Resolution = 22 Megapixels
Camera type = 35 mm
Aperture = F/11
Set Circle of Confusion = Twice Pixel Size?
The first three parameters have no effect on the calculation if you check “Set Circle of confusion = Twice Pixel Size?” When you do this you are determining if the sensor or the lens is the limit of your resolution.
So what does this mean in day-to-day use of your digital camera? I have two cameras. One is a Canon 5D Mk II (22 Megapixels). The other is Hasselblad 500 CM with a Leaf Valeo 22 digital back (22 Megapixels). The pixels are larger in the Leaf back than they are in the Canon. This means that I can stop down farther with the Leaf before it becomes diffraction limited. However the depth of field is less at a given F-stop.
Here I would like to tell you about “The Inventor’s Dilemma”. What this says is that you may invent some device that is better than your competitor, but if his product is “Good Enough” for the majority of the market, you are not going to be able to charge a premium for your better product. You won’t get enough market-share to make a profit.
The majority of the time, I make 16” x 20” prints. I have made a lot of prints this size with both of my cameras. The Canon is much easier to use, has a much longer battery life, and is much lighter to carry around. The overwhelming majority of my print viewers cannot tell which camera I used. On one occasion I was ask to supply equipment and technical support to a commercial photographer here in Vancouver. We were asked to produce a city landscape that would be used as a backdrop for a TV studio set. I produced two images that were 16 ft. long by 4 ft. high at 200 dpi. I used the Leaf Valeo and stitched three shots together for each image. Later we repeated this same landscape using the Canon 5D MkII. I turned it to portrait mode and stitched four images together to produce the same final size. Again, no one could tell the difference.
If someone produces a 35 mm full frame camera with 100 Megapixels, beware. Given the limitations of the wavelength of light, no lens can live up to that resolution. I suggest that the Megapixel race is over. Marketing people may produce one of these cameras, but no physics expert will buy one.
Ray Maxwell is the co-host of Maxwell's House on the TWiT network. Maxwell currently resides in Vancouver, British Columbia, Canada. He is an Electronics Engineer, Color Scientist, Teacher, Speaker & Raconteur.