Help Clarifying Focal Lengths

[Eric Dennis]

Ok I really like my 80mm Atlas Anamorphic lens and I’m specifically referring to the focal length in this case. And I’ve been picking up some Cooke minis and want to grab one in that same focal length. I understand the squeeze and I’m looking more to replicate the field of view I get on the vertical plane not horizontal. Since these are both S35 lenses would I be looking at the Cooke 75mm or are these focal lengths being calculated differently. I’ve just confused myself from comparing full frame lenses and throw in anamorphic and I’d hope someone could just clarify. Thanks!

 

[Tom Gleeson]

Eric,

Assuming the same sensor size your anamorphic is squeezing the horizontal X2 so 80 divide by 2 and a 40mm spherical should match FOV. Assuming then the spherical and anamorphic shots are both being cropped equally to 2.40 you should then have matching vertical FOV with 40mm spherical.

 

[M Harvey]

Yes, the 80mm Atlas and 75mm Cooke should have a similar vertical FOV.

Possibly helpful: keep in mind that you might USE an anamorphic and a spherical lens differently. See this comparison from American Cinematographer, which I think demonstrates the similarity in horizontal angle of view while showing how the feel is different.

 

[Eric Dennis]

Great, thanks for clarifying. I’ll check that out.

 

[Björn Benckert]

A 80mm lens is always a 80mm lens at least on the vertical plane. So if you want the same vertical FOV you should get a 80mm COOKE. That is true as long as you shoot your 80mm Atlas in sensor full hight.


So yes, the focal length of all lenses is calculated the same, Then they might differ some between actual focal length and what the manufacturers mark them as, they tend to round of the numbers a bit and yes some measure at infinity and some at other distances and the breathing makes the lens slightly shift in terms of FOV.

 

[David Mullen ASC]

Yes, you have to compare the same sensor areas -- if, for example, you are using a 15mm height of a sensor for the image, then a 40mm 2X anamorphic and a 40mm spherical have the same vertical view but the 2X anamorphic has twice the horizontal view. But most people aren't comparing the same height because if you are aiming for a 2.40 image, often on a digital sensor you are cropping the sides to use a 1.20 area when using a 2X anamorphic lens... but you are cropping some of the height to get a 2.40 image on a spherical lens.

So in that case, you need to know the exact sensor area for 2X anamorphic versus 2.40 crop in spherical to compare field of views in both directions.

 

[Björn Benckert]

Yes, you have to compare the same sensor areas -- if, for example, you are using a 15mm height of a sensor for the image, then a 40mm 2X anamorphic and a 40mm spherical have the same vertical view but the 2X anamorphic has twice the horizontal view. But most people aren't comparing the same height because if you are aiming for a 2.40 image, often on a digital sensor you are cropping the sides to use a 1.20 area when using a 2X anamorphic lens... but you are cropping some of the height to get a 2.40 image on a spherical lens.

So in that case, you need to know the exact sensor area for 2X anamorphic versus 2.40 crop in spherical to compare field of views in both directions.

Eric / thread starter writes "I understand the squeeze and I’m looking more to replicate the field of view I get on the vertical plane not horizontal." So if you for example want a 16:9 images then 40mm anamorphic and spherical 40mm will give the exact same FOV if the full sensor hight is used.

Exmaple: Monstro 8k 2x ANA 8:9 with a 40mm anamorphic Gives the exact same FOV as Monstro 8kHD / 16:9 with a 40mm spherical.

Then sure, its only half the amount of pixels used in the anamorphic capture. But fov remains the same and you get the oval focus falloff, flares etc that comes with ana´s while the 16:9 crop gets rid of the worst distortions and edge softens and vingeting. Just shot a quick test in that format for a TV series. The DP wanted Ana windscreen but of course the network wants "full picture" so then he wants to meet them half way / 16:9 ana. And the producers like the idea of shooting only half the amount of pixels.

On red, to me it actually not such a bad idea. You can double your data rate and cut the amount of pixels in half. Also did a test with 8:9 in 8,7 and 6k. Basically just using the 40,65,100mm Orions and by shooting with the 3 different crop factors get a ladder of 9 different crops out of the 3 lenses. Sure it's always better to shot as many K's as you can but 6k is still plenty and monstro at 6k still looks stellar and there is a lot of time saved staying with 3 primes instead of 9.

here is links to those tests, not the most beautiful imagery, but you get the idea.

Sensor crop on 40,65,100mm ana: https://f.io/1g7nfPWx

8:9 8k 2x ana: https://f.io/e8_WHigC



To me 8:9 2xana looks way more cool than shooting for example 4:3 1.3xana.

 

[Eric Dennis]

All makes sense. That’s what I thought but started confusing myself quite a bit after obsessing over it all. Thanks for the tests!