T-stops vs. F-stops

[Uli Plank]

I had a bit of time today and thought I'd put a cine Distagon, the Arri/Zeiss 40mm 2.1 T, against a few of my stills lenses to check them for exposure.

The results are interesting: while I always thought that all marked F-stops would be slower than the marked T-stops, this is wrong!

All lenses were slower in T-stops when wide open, but starting from the first stop closed down, the values were constantly identical to the F-stops (within an error margin well below one third stop).

One example: the Minolta Rokkor 58mm f1.2 (aka Bokeh-Monster) has a T-stop of 1.4 (true speed) wide open. At f2.0 the value is identical to the lens with the marks in T-stops. This was the same on all the lenses I tested, a few Minoltas, Zeiss Contax and one Voigtländer (compared on a GH2, not RED). OK, not a fully scientific test, but good enough for me. For most practical purposes I can consider my lenses closed down as marked like T-stops, only the WO values are exaggerated.

Voigtländer 25mm @ f0.95 is a T 1.1.5, Minolta Zoom 35-70mm @ f3.5 is a T 3.8, both Rokkor PG and Contax Zeiss @ f1.4 are T 1.6, even the Rokkor 28mm @2.0 is a T 2.1.

Thought some might find this useful…

 

[Martin Stevens]

Very useful and interesting Uli.

I always use F-Stops for depth-of-field, and T-Stops for metering, and since they are so close together
I have not paid the difference much attention.....especially now that I have the ability to shot with the
very variable ISO of the Epic.

 

[Shervin Mandgaryan]

Thanks for the info Uli!

To those who don't know, it all has to do with how the glass is coated and with what formula of coating. Zeiss has higher light transmission rates in their lenses than Canon or other still lens manufacturers from what I've seen. Most stills glass is F-stop rated because it's simple math that all manufacturers can adhere to... T-stop is calculated and I'm assuming more embarrassing for manufactures to post as it all has to do with how good your coatings are for anti-reflectivity.

That being said, the same F-Stop on Contax lenses vs. my other lenses shows that Zeiss has a bit more luminosity than others. No real surprise there as Zeiss arguably has the most matured coating than other lens manufacturers.

 

[Timur Civan]

Tstops are not calculated, they are measured. It is a transmission of light in finite amounts that can be accurately measured.

In theory Tstops should be absolutely consistent even among different manufacturers.

 

[Shervin Mandgaryan]

Tstops are not calculated, they are measured. It is a transmission of light in finite amounts that can be accurately measured.

In theory Tstops should be absolutely consistent even among different manufacturers.

That's what I meant to say, wrong choice of words on my part. Cheers.

 

[Uli Plank]

Yes, and T-stops are not dependent on coating only, the number of lenses and their thickness plays a role too, even the surface and the black paint on the inside of the tubes.
T-stops are measured, while F-stops are calculated from the focal length of the lens and the diameter of the aperture.

The only surprise to me was that they don't differ much any more if the lens is stopped down just one step.

 

[David Mullen ASC]

The difference between T-stops and f-stops matters more in cinematography than in still photography where several shots in a sequence may use different lenses, or maybe the same set-up will use multiple cameras with different lenses, so setting them all to the same f-stop will produce some slight variations in exposure. It's become less of an issue with digital cameras since you are likely to set final exposure based on how it looks on the monitor and with exposure aids such as a histogram or waveform so you are matching set-ups that way rather than by strictly metering.

 

[Ryan Patrick O'Hara]

Wait... what T-stop lenses were you comparing the photography lenses with?

-ryan

 

[Uli Plank]

Didn't I mention? Arri/Zeiss Distagons, they are very precise in both their T-stops and witness marks in my experience.