Is the Sigma 18–35mm ƒ/1.8 really ƒ/1.8?

Ok, this is the third time I’m writing about aperture values trying to figure out something that is puzzling, I reviewed the concept of aperture in lenses and why the Micro Fourth Thirds System has a hard time trying to emulate a full frame result in the two lasts posts:

First, let me show you a picture I have taken using the Sigma 18–35mm ƒ/1.8 with the full frame camera Canon 6D:

Canon 6D image using the Sigma 18–35mm ƒ/1.8 showing different crop factors

As you can see, the vignetting is quite evident, this is because this lens isn’t designed for full frame cameras but for crop cameras, nevertheless it can be used on a full frame camera because of the EF mount, but, if it is a truly ƒ/1.8 lense, Why it has vignetting? Well, because a truly ƒ/1.8 will completely fill a full frame sensor, Sigma decided to make a version for crop cameras using the very same glass measurements but cropping the extra glass that filled a full frame area, thus, making the same result of a truly ƒ/1.8 lens could have but with the extra unused light, in other words, reducing the aperture.

So, if you could make a truly full frame 18–35mm ƒ/1.8 you will see the very same result as the Sigma lens on crop cameras, but you cannot use the Sigma lens on a full frame camera because, after doing some measurements on the vignetting (not scientific calculations) I have find that the real aperture of this lens is ƒ/2.8 on paper.

I’m going to illustrate this idea of “croping” the glass in order to achieve the same results in crop lenses, let’s say you have a full frame lens and camera, a 50mm ƒ/1.8 on a full frame Canon 6D camera, let’s see how the light is used on different sensor sizes:

So, you want to emulate the same result on a crop camera, the Canon 80D for instance, dividing the focal length and the aperture using 1.6x crop factor give us a 31.2mm ƒ/1.12 lens, but that doesn’t exist, so the closer lens is a 30mm ƒ/1.2:

And, as you can see, a full frame lens will project light for a full frame sensor, but this light isn’t being used, so, they crop the glass that isn’t used in a crop camera:

This solution make the lens look like a 30mm ƒ/1.2 but in reality it is a ƒ/1.9 lens, close to a ƒ/1.8 lens, this is what Sigma did and what most manufacturers do with their crop camera lenses, so in reality they are not selling a truly ƒ/1.8 lens but a lens that on a crop camera looks like and works like a ƒ/1.8 lens, what intrigues me is how most manufacturers for the Micro Fourth Thirds System are making their lenses, Are they truly making big aperture lenses or just cropping lenses?

Leave a Reply

Your email address will not be published. Required fields are marked *

Releated

Addressing DaVinci Resolve’s color shift

I should add to the title “Using a Macintosh”, because that’s where the problem is; see, Apple is well known for doing things their way, in many ways it is better, like the DPI scaling using multiple displays, Windows sucks at that, but the way it handles color, well, is quite entangled. Dan Swierenga wrote […]

EOS R5, C-Log 3, Ninja V and Resolve

The Canon EOS R5 and the release of 1.3.1 firmware provide a unique opportunity to try C-Log 3 on a mirrorless camera instead of the cinema line of cameras, and while recoding in RAW, Resolve provides the opportunity to try C-Log 2. Considering the caveat of overheating and the time limit, a workaround that many […]