When does ƒ/2.8 not equal ƒ/2.8?

I know that a fast prime that's stopped down to 2.8 reduces the effects of some lens aberrations, so that the image will generally be better than a lens with a max f-stop of 2.8 shot at that f-stop.

But I've also read that two lenses of identical focal lengths, but having different max apertures, will not transmit the same amount of light. Specifically, the lens with the larger max aperture will transmit more. For example, the 85L at f/1.8 will transmit more light than the 85mm f/1.8 at f/1.8. Is there any truth to this and, if so, can you explain?

The aperture to focal length ratio doesn't change but different lenses do transmit different amounts of light. For example, with the exposure set manually and constant for both shots, my 50mm f/4 will result an a brighter image (maybe 1/3 stop) than my 24-105. There's a lot more glass in the zoom lens to absorb or reflect the light before it makes it to the sensor.

Im no expert, but I would say "No".

The only thing I could think they might be referring to is a brighter viewfinder since a lens "rests" at its maximum aperture and only stops down for a split second (or however long the shutter speed is) the moment the shutter is pressed. So, even though the 85 1.8 @ 1.8 will transmit the same amt of light during the shot as the 85L @ 1.8, the "resting" state of the 85L is at 1.2 and the viewfinder could be brighter as a result.

In principle, the f-stop setting should be an absolute measure of the light transmission through the lens. f2.8 is f2.8 no matter what. Now, the calibration of the aperture may mean that f2.8 differs somewhat between lenses, but that's not the same thing and in general it won't vary too much.

When looking through the viewfinder of course, the lens is always wide-open, so in that sense the 85 1.2 lets more light in than the 85 1.8 - but only for the viewfinder. If you set f1.8 on both lenses, then the sensor should see the same amount of light since the 85 1.2 will be stopped down when the shot is taken.

The f number is a ratio: focal length divided by aperture diameter. The light transmission will vary between lenses of different design when set to the same f number.

If you want/need a setting for constant light transmission between different lenses, you want the T number. Do you remember seeing old movie cameras with a lens turret at the front? Those lenses were calibrated to T-stops so that the light transmission would not change when the turret was rotated.

frugivore wrote in post #13109533
But I've also read that two lenses of identical focal lengths, but having different max apertures, will not transmit the same amount of light. Specifically, the lens with the larger max aperture will transmit more. For example, the 85L at f/1.8 will transmit more light than the 85mm f/1.8 at f/1.8. Is there any truth to this and, if so, can you explain?

There is some truth in what you've posted, but also some, let's say "unfounded assumptions."

The first thing to understand is that the focal length and max aperture that are printed on the outside of the lens have more to do with marketing than engineering because they are often rounded off. The Sigma 50mm f/1.4 is a example, because comparing it side by side with other 50mm primes the Sigma is noticeably wider, I would say that the Sigma is more like 45mm or so.

The next thing to understand is that focal length measurements are always based on having the lens focused at infinity. Many lenses have a much shorter focal length as the focus distance approaches minimum. Many Nikon shooters have been upset about the current version of the 70-200, which is -much- shorter than 200mm when focusing on a close subject.

Given the above variations in focal length, and since aperture is tied to focal length, there is plenty of room for focal lengths to be inaccurate even before the marketing department gets their hands on them.

So having said all that, it is possible that the 85/1.2 actually has a slightly larger aperture at "f/1.8" than the 85/1.8 has when set to "f/1.8". It is equally possible that it is the other way around. The Canon "85/1.8" might actually be an "82/1.7" or an "81/1.9" or an "88/1.7" or an "87/1.7" or who knows what.

Basically, you really cannot make any sort of universal statements about this. And any differences in aperture are likely to be so small that you would never be likely to notice them in real world use of the lenses.

RTPVid wrote in post #13109591
The f number is a ratio: focal length divided by aperture diameter. The light transmission will vary between lenses of different design when set to the same f number.

Yes, it's a ratio - but of the 'width' of the lens to its 'length' (I'm simplifying here - don;'t take it too literally). The result is essentially a measure of the light transmission of the lens and for different lenses when set to the same f number it will be the same. If not, then photographing a grey card with a 50mm lens and f2.8 would give a different exposure to doing it with a 100mm lens at f2.8. It doesn't! The T number takes into account the fact that lenses 'lose' some light transmission since they're not 100% efficient. As you say, this is used in cinematography, but I believe (prepared to be proved wrong!) that the SLR lenses we use have these effects compensated for by the calibration of the f-stops.

frugivore wrote in post #13109533
But I've also read that two lenses of identical focal lengths, but having different max apertures, will not transmit the same amount of light. Specifically, the lens with the larger max aperture will transmit more.

That *can* happen, but when it does it's because of one of two things (that I know of):

* The "selected" f-stop and the *actual* f-stop can vary due to inaccuracies in the aperture mechanism. If a wide-open zoom is compared to a stopped-down prime, it may be that the prime is not stopping down completely, but is actually open a little wider (e.g. 1/3 stop) than the selected f-number.
* Lens transmissivity can vary between lenses due to the number of elements and lens coatings. A large aperture prime lens may have fewer lens surfaces or better coatings, giving it as much as 1/3 stop brighter light transmission than a zoom.

"f/stops" vs "T" stops

The light transmission of lenses calculated in f/stops is measured mathematically. It is the focal length divided by the effective diameter (aperture) of the opening. The light transmission of say an 90mm lens at f/4 and a 50mm lens at f/4 is pretty close but,is not exactly the same. Differences in the build of the lens impact the transmission. Usually, the results are close enough for still photography to be considered equal.

However, this was not always the case with professional cinematography. Exposures from one lens to another had to be spot on and exactly the same because shots from one lens needed to be cut in with shots from other lenses to form a sequence. The exposures had to match exactly. Our emulsions were often very slow (often ISO-16) and the latitude was very narrow.

The solution was a set of lenses whose transmission was delineated in "T" stops. A "T" stop was determined my actually measuring the light transmission before designating the "T" stop. Therefore the "T" stop from one of the lenses in the set had EXACTLY the same light transmission as the same "T" stop on another lens in the same set. The T-stop is the true speed of the lens, calculated by compensating for its light absorbance and reflectance, and will result in accurate exposure.

Another reason that T-stops were necessary was that older lens coatings were not as efficient as the coatings of top modern lenses and therefor the transmission of different lenses varied considerably...

Note: an f/stop is written as a fraction while the T-stop is written with a dash. In other words: f/8 and T-8 are the ways the f/ and T stops are written.

Needless to say, a set of "T" stop lenses was exceptionally expensive. The increased use of zoom lenses for varying focal lengths of shots was the demise of "T" stop lens sets...

RPCrowe - fantastic post, thanks. Learned a lot.

Great information in this thread! So it seems that there are some factors that affect the amount of light transmitted through the lens, aside from just the effective aperture.

The review I read was at Cameralabs.com discussing the 35mm lenses. Granted, the fact that the f/1.8G is designed for DX cameras might also play a part in the transmission discrepancy, but that's what got me wondering.