Diffraction Calculator

A long-running commentary, I think some of you have seen me talk about, is that high MP cameras are not as diffraction-limited as you may first think. That's because a lot of you have been using a very simplistic calculator to determine if your f-stops are diffraction-limited on your sensor.

As Canon may be approaching 40MP on a crop camera, I know there's going to be some stating that you cannot shoot at F/5.6 or above, or you will be diffraction-limited. So, since the R7 Mark II may cause some of you some angst, here we are.

But that's only part of the equation. It really depends on what you are doing with this image, as diffraction only matters if you can see it.

We have all seen this calculator.

Basically, this calculator calculates the airy disc caused by diffraction and compares it to the size of the pixel. If it's larger than the pixel, then it's diffraction-limited.

But this is only suitable if you are inspecting each pixel at optimum viewing distance away from your monitor, at 100% magnification on the screen. So let's assume you have a 27″ 4K monitor. That's the diffraction that you can detect if you printed that image 4 feet wide, put it up on the wall, and viewed it from approximately 12 to 18 inches away.

But normally, to view that 48″ wide print, you'd have to stand back, approximately somewhere between 1.5x to 2x the diagonal length of the image, or around, in this case, 7 to 9 feet away. Once you do that, you'll no longer be able to see the details that you did when you were 12″ or 18″ away. This changes the circle of confusion for the observer as the farther back you stand, the less you'll be able to see fine details.

What this mathematically comes down to is when the airy disc of diffraction is large enough for us to see it. We see the airy disc of diffraction when it's larger than our observable circle of confusion.

So we get this calculator, which better represents sensor megapixels and how you view an image.

Now, for some people, you are going to use a combination of the two, and find a happy medium between what the observer diffraction is, versus the pixel level difraction. This is meant to illustrate that there are no absolutes, and being DLA at the pixel level isn't necessarily a problem.