Difference Between F-Stop and T-Stop

[Vadim Joy]

Is it true that f-stop is actually faster then t-stop? For example T2.8 is same as F2

[Gregg MacPherson]

That's true I guess, but a more usefull way to see it is that T stops are slower. The f stop is a theoretical number based on the lens geometry. Focal length/aperture diameter. The T stop is corrected for light loss, so it's always a bigger number, slower. The exact light loss at each f stop will vary from lens to lens. The lens techs will have a more precise way of describing all this.

[Chris Millar]

Let's see if I can beat Dom, Adrian, Jean-Louis, et al to this :D

 

F-Stop is the ratio between the focal length and the diameter of the entrance pupil .

 

T-Stops are adjusted for light transmission efficiency, and this will always be a higher number.

 

Some confusion may result from the fact that an entrance pupil is in other words an 'effective' aperture - which sounds 't-stoppy'.

 

Hope I didn't just add to the misinformation ;)

[Matthew W. Phillips]

Chris was right on. F-stops are if we lived in a perfect world and T-stop is the reality. T-stops take light loss in consideration. For the old film cameras especially where beam splitters would affect the total light going to the emulsion. F-stops don't account for that...T-stops do.

[Gregg MacPherson]

I don't think Chris or Matthew did any better than me (smiley face ?).

Actually, I'm really curious to find out when or if the entrance pupil is not the aperture. I think I'll google that.

 

Right after I tried to post the above text the forum site became inaccessible. As if Tim or someone had some code running that would "spit the dummy" if these three people were ever in the same place at once.

[Chris Millar]

Heya Gregg,

 

Have a search for the other dozen or so t-stop vs f-stop threads already populating the forum...

 

They tend to follow a pattern:

 

Post 1: The question

Post 2: The answer (that you gave and for all intents and purposes, at least to the OP, is a perfectly adequate answer)

Post 3: The correction re. aperture vs. entrance pupil

Post 4: Same correction by another poster within 5mins of Post 3 (written before post 3 was visible)

 

The threads usually diverge from here but will very likely include some 'what he said' responses, a retraction by poster # 2 after they googled 'f-stop entrance pupil' then discussion regarding all the other threads and how the info is freely available on wikipedia.

[Dom Jaeger]

Is it true that f-stop is actually faster then t-stop? For example T2.8 is same as F2

 

Yes, the T-stop is always slower than the matching f-stop. Rarely as much as a full stop though, unless you're looking at very old zooms. A big, multiple element modern zoom might only lose a third of a stop, a prime even less.

 

For the old film cameras especially where beam splitters would affect the total light going to the emulsion. F-stops don't account for that...T-stops do.

 

T-stops only ever account for light loss in the lens itself, they don't factor in the prism light loss of a camera. Some people get confused about this with Rx lenses for reflex Bolexes, believing that the lens aperture marks compensate for the Bolex prism, but they don't. A T-stop only relates to the lens. The only possible exception might be fixed (ie non-interchangeable) zooms on certain cameras like Canon Scoopics, although to be honest I can't recall if the Scoopic zoom used T-stops or f-stops. Every Super 8 camera with a fixed zoom used f-stops.

 

Actually, I'm really curious to find out when or if the entrance pupil is not the aperture.

 

I recently discussed this in a thread about fast lenses. Quite often the front element or group of elements act as a magnifying glass and the entrance pupil (basically the aperture seen through the front) is considerably larger than the actual aperture. For example an f/1.9 (T2) 180mm Cooke S4 has an "effective" aperture of roughly 95mm but the actual iris opening only measures about 40mm. On other lenses the entrance pupil can be smaller than the actual aperture. On every zoom that maintains a constant f-stop the entrance pupil has to change as the focal length changes, but the iris opening remains fixed. Have a look through the front of a zoom while zooming and you'll see the aperture changing size.

 

What I'm curious about is when T-stops were first introduced. Anyone know?

[Matthew W. Phillips]

I don't think Chris or Matthew did any better than me (smiley face ?).

Actually, I'm really curious to find out when or if the entrance pupil is not the aperture. I think I'll google that.

 

Right after I tried to post the above text the forum site became inaccessible. As if Tim or someone had some code running that would "spit the dummy" if these three people were ever in the same place at once.

 

I seriously did not see your post there, Gregg, when I posted here. I thought Chris was the first to respond?!?

[Matthew W. Phillips]

 

T-stops only ever account for light loss in the lens itself, they don't factor in the prism light loss of a camera. Some people get confused about this with Rx lenses for reflex Bolexes, believing that the lens aperture marks compensate for the Bolex prism, but they don't. A T-stop only relates to the lens. The only possible exception might be fixed (ie non-interchangeable) zooms on certain cameras like Canon Scoopics, although to be honest I can't recall if the Scoopic zoom used T-stops or f-stops. Every Super 8 camera with a fixed zoom used f-stops.

 

Well, this is interesting. When I shot Super 8 (which has accounted for most of the film I have shot), I had always assumed that the markings on those fixed lenses were T-stops and not F-stops. Nonetheless, my footage always came out alright, even using external metering.

[Gregg MacPherson]

 

I seriously did not see your post there, Gregg, when I posted here. I thought Chris was the first to respond?!?

 

We might have all posted about the same time. Or, Mother Nature might be having a laugh at us.

[Adrian Sierkowski]

Let's see if I can beat Dom, Adrian, Jean-Louis, et al to this :D

 

 

You caught me on one of my rare nights when I was obliged to go out.

[Brian Drysdale]

 

Well, this is interesting. When I shot Super 8 (which has accounted for most of the film I have shot), I had always assumed that the markings on those fixed lenses were T-stops and not F-stops. Nonetheless, my footage always came out alright, even using external metering.

 

I shot some super 8 using an external meter with a camera I had never used before and hadn't a chance to test, it was was underexposed. Although, I've done it with other 8mm cameras without a problem.

[Vadim Joy]

Is there any way to calculate f-stop into t-stop to be on the same wave with other cinema lenses? Why not just change markings on the barrel of the lens?

[Vadim Joy]

I need the T number for the following lenses, please.

 

Tokina 11-16mm F2.8-2.6

Tokina 28-70mm F2.6

Foton 37-140mm F3.5

 

I would appreciate any help. Thanks

[Gregg MacPherson]

Hey Vadim,
For the iris marks to be useful they need to be in standard values....1.4, 2.0, 2.8, 4.0...etc. If you were referencing a light meter then a string of funky numbers on the iris ring would be bad. Better that the standard stop increment numbers are just repositioned slightly and called T stops. The wide open number is often non standard.

 

On the lenses I have had the difference between f and T stop is small. On the Angenieux 12-120 I had it was (I think) wide open 2.2/2.5. The early Zeiss 10-100 was 2.8/3.3. Zeiss S16 superspeed primes are 1.2/1.4

 

If you don't have T stop markings you can just estimate the extra increment open that you need. Maybe someone like Dom could give us an idea of whether the f stop to T stop difference is fairly constant for a given lens, in terms of the required correction, fraction of a stop.

 

Maybe with a digital camera one can set up your own test just with a grey scale.

I googled the Tokina 11-16 and the I think optically identical Tokina cine version has wide open f stop vs T stop of 2.8/3.0. Have you looked around the web and asked on the forums where people may be using these lenses (bmcuser.com or one of the red forums maybe)

[Vadim Joy]

Hey Vadim,

For the iris marks to be useful they need to be in standard values....1.4, 2.0, 2.8, 4.0...etc. If you were referencing a light meter then a string of funky numbers on the iris ring would be bad. Better that the standard stop increment numbers are just repositioned slightly and called T stops. The wide open number is often non standard.

 

On the lenses I have had the difference between f and T stop is small. On the Angenieux 12-120 I had it was (I think) wide open 2.2/2.5. The early Zeiss 10-100 was 2.8/3.3. Zeiss S16 superspeed primes are 1.2/1.4

 

If you don't have T stop markings you can just estimate the extra increment open that you need. Maybe someone like Dom could give us an idea of whether the f stop to T stop difference is fairly constant for a given lens, in terms of the required correction, fraction of a stop.

 

Maybe with a digital camera one can set up your own test just with a grey scale.

I googled the Tokina 11-16 and the I think optically identical Tokina cine version has wide open f stop vs T stop of 2.8/3.0. Have you looked around the web and asked on the forums where people may be using these lenses (bmcuser.com or one of the red forums maybe)

I will do some research.

[Vadim Joy]

So far I found out that Tokina 11-16 is T2.8. How? Well, it's may be not very scientific discovery. Duclos Lenses converts normal Tokina 11-16 into cine version. As far as I know they don't change actual glass elements, only some parts of barrel where markings are and add PL mount to it. Price just immediately from $600 to $3.5K. Maybe I'm missing something and they add some spells to it.

 

Here is a link http://www.ducloslenses.com/products/duclos-11-16mm

Edited April 12, 2014 by Vadim Joy

[Gregg MacPherson]

I read here that Tokina are making their own factory cine version of the 11-16, f2.8/T3.0

https://www.fstopacademy.com/blog/tokina-cinevise-their-popular-11-16mm-zoom/

 

I don't know how Duclose have managed T2.8. Maybe someone else can shed some light on that.

[Aapo Lettinen]

I need the T number for the following lenses, please.

 

Tokina 11-16mm F2.8-2.6

Tokina 28-70mm F2.6

Foton 37-140mm F3.5

 

I would appreciate any help. Thanks

The Foton zooms are either T4.3 or T4.4 depending on the version (different coatings??)

The Fotons are not really usable wide open, I'd say stopped down T5.6 2/3 or more they are usable although the ca may be an issue and the long end at 130 - 140mm is useless due to mushy image and huge amount of ca

Edited April 12, 2014 by aapo lettinen

[Vadim Joy]

I read here that Tokina are making their own factory cine version of the 11-16, f2.8/T3.0

https://www.fstopacademy.com/blog/tokina-cinevise-their-popular-11-16mm-zoom/

I didn't know that. But they changed only housing, not optics. What's the point of spending $1K for a lens if you can buy normal still lens and some follow focus rings and throw on it for less than $600.

[Vadim Joy]

So far we know that according to Tokina themselves, Tokina 11-16 f2.8 is T3. Foton is T4.2/4.3. What about Tokina 28-70mm f2.8? Can it be T3?

[Brian Drysdale]

The odds are you won't be far off using T3 on the Tokina 28-70mm f2.8. The difference with modern lenses isn't large, however, on old zoom lenses like the Angenieux 9.5 to 95 f2.2 it;s T2.8.

[Gregg MacPherson]

 

On the lenses I have had the difference between f and T stop is small. On the Angenieux 12-120 I had it was (I think) wide open 2.2/2.5. The early Zeiss 10-100 was 2.8/3.3. Zeiss S16 superspeed primes are 1.2/1.4

 

Not that it matters much but the S16 SSs are T1.3. I know them well and I have one sitting beside the keyboard so I don't know I did that.

[Dom Jaeger]

As a very general rule you can assume a modern zoom to lose a third of a stop, an older zoom to lose a half and a prime to lose a quarter. If you want to be more accurate, there are ways to measure transmission loss but it can get a bit complicated, especially with still photography zooms that can vary depending on the focal length.

 

There are probably others but the DxO Mark site has pretty comprehensive lens tests that cover transmission loss in many still lenses. The Tokina f/2.8 11-16 is there:

http://www.dxomark.com/Lenses/Tokina/Tokina-AT-X-Pro-SD-11-16-F28-IF-DX-II-Canon---Measurements

 

If you click on the transmission data it shows the lens is generally T3.1 with a dip in the middle to T3 at 13mm.

 

It's interesting to compare other zooms, the Tokina f/2.8 16-28 for example is T3 at the wide end but T3.3 at 28mm, so the variation is a fifth to a third of a stop. The Nikkor f/4 70-200 is a very stable T4.4 through the range, conforming to the third of a stop loss generalisation.

 

There is no test data for the Tokina 28-70 but you could perhaps assume a degree of similarity with Canon's EF f/2.8 24-70:

http://www.dxomark.com/Lenses/Canon/Canon-EF-24-70mm-F28L-II-USM---Measurements

 

That one is T2.9 at 24mm, T3 from 35 to 50mm and T3.2 at 70mm.

 

Lenses designed specifically for cinema tend to be more stable in their transmission loss.

[Dom Jaeger]

There is no test data for the Tokina 28-70 but you could perhaps assume a degree of similarity with Canon's EF f/2.8 24-70:

http://www.dxomark.com/Lenses/Canon/Canon-EF-24-70mm-F28L-II-USM---Measurements

 

Heh, actually probably not..

 

Looking at a few other brand versions of f/2.8 24-70 zooms, Tamron's measured a uniform T2.8 (which has to be either a measurement error or Tamron's f-stop marks are off) while Sigma's is rated at mostly T3.3, a half a stop loss.

 

Which probably brings up another issue, the accuracy of consumer grade lens f-stop marks.