Posted by Dan Roberts on March 23, 2011 at 9:03 AM
One thing that has always puzzled me about photography is the obsessive discussion over low f/stop lenses but no one ever even mentioins t/stop. In any other discussion of optics other than photography, the two primary issues are typically magnification and light transmission (t/stop). But us goofy photographers are stuck in this archaic terminology of focal length and f/stop. After all, focal length implies a magnification factor so why not just say it like they do for video or P/S cameras. And f/stop is a measure of the geometric relationship between the aperture size relative to focal length, not an indication of what percentage of light reaches the focal plane. Use of f/stop in comparing one lens to another implies/assumes that the light transmission capabilities of the lenses are equivalent. Which is a gross (and innaccurate) assumption. Yet in reading gear reviews on lenses, I can't recall that I've ever seen anyone actually quantify the light transmission characteristics of a lens.
Does anyone know of t/stop data available anywhere? Or am I just missing something?
Dan,
That is a very good question and one that I have never considered before you raised it. My guess is that since exposure is based on the assumption that for a given shutter speed at a given ISO, each increase in f stop will produce twice the light falling on the sensor or film, that the lenses are built to accomodate this assumption. In other words, if we expect f/8 to produce twice the light and twice the exposure of f/11, the transmissivness of the glass has to be consistent or the iris has to be adjusted in such a way as to open (or close) enough to allow an increase (or decrease) in the amount of light passing through the iris and reaching the sensor. Thus, the increase (or decrease) in the size of the iris opening may not be geometrically consistent, but it has been adjusted to permit the expected passage of light.
Now granted, this is just a theory. I may be off my miles, but that's my $0.02 for what it's worth. (Oops! The dollar has just been devalued again. It's now worth $0.01)
Posted by Garry Revesz on March 23, 2011 at 1:44 PM
Dan you raise a good question but I have to admit my thinking falls somewhere along the line of Nick's comments. The f/stop of a given lens is in relation only to the other apertures of that particular lens, ie; a 1/2 stop down allows 50% more light, and so on.
I have never heard of the t/stop but can easily see how that stat would be extremely useful in comparing different lenses like a 70-200 Canon to a 70-200 Nikon or Tamron. Another piece of useful information with which to make decisions on lenses.
I don't think lenses are typically more than 1/3rd stop off in actual light transmission from their same aperture cousins, so in the grand scheme of things it's not all that much to worry about. It's nice to know that regardless of whether I put on a 50mm at f/4 or a 500mm at f/4 I will get the same shutter speed / iso settings to get the same exposure.. and it generally works to within 1/3rd a stop or so.
As far as focal length and magnification factor, focal length is a little more instructive because magnification factor depends on the distance to the subject ( M=f/(f-d) ). It's pretty easy to find information about the magnification factor - bhphoto lists it on their specs. Just to point out how confusing it can be, a canon 50mm is 1:6.6 and a 500mm is 1:8.3, a true macro is going to be 1:1. None of those numbers will give you any indication of what the field of view or depth of field are like, and all those numbers are at the minimum focusing distance (so not generally useful knowledge). If you mean the '10x' designation on most point and shoots and binoculars and such - if I'm not mistaken that number is a rather arbitrary value which is obtained by dividing the focal length (in 35mm equivalent) by about 50mm (since 50mm is considered the eye's natural field of view). It comes from the formula for magnification used for 2-lens systems like telescopes and microscopes where M = fo / fe, where fo = focal length of the objective lens and fe = focal length of the eye piece.
So.. focal length is a much more useful specification to know since it will give you a precise idea for all of the following: 'magnification', the field of view, and depth of field. You can always calculate the magnification factor easily enough given the focal length if that's important to you.
Posted by Dan Roberts on March 23, 2011 at 2:33 PM
I don't think lenses are typically more than 1/3rd stop off in actual light transmission from their same aperture cousins, so in the grand scheme of things it's not all that much to worry about. It's nice to know that regardless of whether I put on a 50mm at f/4 or a 500mm at f/4 I will get the same shutter speed / iso settings to get the same exposure.. and it generally works to within 1/3rd a stop or so.
With regards to determining proper exposure I agree with you. Particularly with TTL metering the difference in light transmission from one lens to the next is totally irrelevant. But when evaluating lenses against one another the light transmission capabilities could make a huge difference in "usability" of the lens in low light situations. Also with AF, the better light transmission the quicker and more accurate AF will be. The two lenses that I use for wildlife photography are both zoom lenses with fixed f/4 max aperture. So theoretically equivalent when wide open. But I can tell you that looking through the viewfinder the difference is as dramatic as that of looking through a pair of $20 10x binos from Walmart versus a pair of mil grade Fujinons. And the difference in the specs of those two examples fits right in with your example of 1/3 to 1/2 stop difference. So that's my reason for studying this topic. Better light transmission doesn't necessarily mean better exposure capability. But it does equate to better visibility for the photographer, potentially better AF performance, and potentially 1/3 to 1/2 stop lower ISO or faster shutter speed for a given aperture setting under a given set of lighting conditions. I think most wildlife photographers would agree that these things are important and yet curiously this subject doesn't get any attention.
Regarding the focal lengths, I was simply making observations. Clearly an interested party with basic math skills can figure it all out. But when someone says they shot a photo with a 105mm lens, that is totally irrelevant information unless one knows what size the capture medium was. Wouldn't it be nice if there was a universally applicable standard?
Clearly I have too much time on my hands....
Posted by David Bostedo on March 23, 2011 at 6:12 PM
My guess is that since exposure is based on the assumption that for a given shutter speed at a given ISO, each increase in f stop will produce twice the light falling on the sensor or film, that the lenses are built to accomodate this assumption.
I actually doubt this is true. I think it's much more likely that for a given f-stop, you actually aren't getting the amount of light that you'd think. I think Dan is probably correct that they have different t-stops for the same aperture. Since the f-stop number is a physical number, lenses probably match it very precisely. It's more likely that you simply aren't getting the light you might think you are.
However, Garry's contention would still apply. So pretend you have 2 lenses, both capable of f/2.8. If you put one on your camera, in aperture priority mode, and check the auto shutter speed of a scene for the same exposure, I'd guess you could see something like this (assuming the shutter speeds could actually fall on these values) :
Lens 1, f/2.8 : 1/500 s
Lens 2, f/2.8 : 1/460 s
Lens 1, f/4 : 1/250 s
Lens 2, f/4 : 1/230 s
So while lens 2 doesn't permit as much light through, for just that lens the relationship between f-stops holds. That's a bit of a guess on my part, so if someone can confirm, that would be great.
Posted by Dan Roberts on March 23, 2011 at 6:45 PM
David is right. The f stop number is purely indicative of physical geometry. It is not adjusted for the light transmission capability of the lens (which is my whole point of starting this thread).
Discussion
6 repliesDoes anyone know of t/stop data available anywhere? Or am I just missing something?
That is a very good question and one that I have never considered before you raised it. My guess is that since exposure is based on the assumption that for a given shutter speed at a given ISO, each increase in f stop will produce twice the light falling on the sensor or film, that the lenses are built to accomodate this assumption. In other words, if we expect f/8 to produce twice the light and twice the exposure of f/11, the transmissivness of the glass has to be consistent or the iris has to be adjusted in such a way as to open (or close) enough to allow an increase (or decrease) in the amount of light passing through the iris and reaching the sensor. Thus, the increase (or decrease) in the size of the iris opening may not be geometrically consistent, but it has been adjusted to permit the expected passage of light.
Now granted, this is just a theory. I may be off my miles, but that's my $0.02 for what it's worth. (Oops! The dollar has just been devalued again. It's now worth $0.01)
I have never heard of the t/stop but can easily see how that stat would be extremely useful in comparing different lenses like a 70-200 Canon to a 70-200 Nikon or Tamron. Another piece of useful information with which to make decisions on lenses.
Good one.
As far as focal length and magnification factor, focal length is a little more instructive because magnification factor depends on the distance to the subject ( M=f/(f-d) ). It's pretty easy to find information about the magnification factor - bhphoto lists it on their specs. Just to point out how confusing it can be, a canon 50mm is 1:6.6 and a 500mm is 1:8.3, a true macro is going to be 1:1. None of those numbers will give you any indication of what the field of view or depth of field are like, and all those numbers are at the minimum focusing distance (so not generally useful knowledge). If you mean the '10x' designation on most point and shoots and binoculars and such - if I'm not mistaken that number is a rather arbitrary value which is obtained by dividing the focal length (in 35mm equivalent) by about 50mm (since 50mm is considered the eye's natural field of view). It comes from the formula for magnification used for 2-lens systems like telescopes and microscopes where M = fo / fe, where fo = focal length of the objective lens and fe = focal length of the eye piece.
So.. focal length is a much more useful specification to know since it will give you a precise idea for all of the following: 'magnification', the field of view, and depth of field. You can always calculate the magnification factor easily enough given the focal length if that's important to you.
With regards to determining proper exposure I agree with you. Particularly with TTL metering the difference in light transmission from one lens to the next is totally irrelevant. But when evaluating lenses against one another the light transmission capabilities could make a huge difference in "usability" of the lens in low light situations. Also with AF, the better light transmission the quicker and more accurate AF will be. The two lenses that I use for wildlife photography are both zoom lenses with fixed f/4 max aperture. So theoretically equivalent when wide open. But I can tell you that looking through the viewfinder the difference is as dramatic as that of looking through a pair of $20 10x binos from Walmart versus a pair of mil grade Fujinons. And the difference in the specs of those two examples fits right in with your example of 1/3 to 1/2 stop difference. So that's my reason for studying this topic. Better light transmission doesn't necessarily mean better exposure capability. But it does equate to better visibility for the photographer, potentially better AF performance, and potentially 1/3 to 1/2 stop lower ISO or faster shutter speed for a given aperture setting under a given set of lighting conditions. I think most wildlife photographers would agree that these things are important and yet curiously this subject doesn't get any attention.
Regarding the focal lengths, I was simply making observations. Clearly an interested party with basic math skills can figure it all out. But when someone says they shot a photo with a 105mm lens, that is totally irrelevant information unless one knows what size the capture medium was. Wouldn't it be nice if there was a universally applicable standard?
Clearly I have too much time on my hands....
I actually doubt this is true. I think it's much more likely that for a given f-stop, you actually aren't getting the amount of light that you'd think. I think Dan is probably correct that they have different t-stops for the same aperture. Since the f-stop number is a physical number, lenses probably match it very precisely. It's more likely that you simply aren't getting the light you might think you are.
However, Garry's contention would still apply. So pretend you have 2 lenses, both capable of f/2.8. If you put one on your camera, in aperture priority mode, and check the auto shutter speed of a scene for the same exposure, I'd guess you could see something like this (assuming the shutter speeds could actually fall on these values) :
Lens 1, f/2.8 : 1/500 s
Lens 2, f/2.8 : 1/460 s
Lens 1, f/4 : 1/250 s
Lens 2, f/4 : 1/230 s
So while lens 2 doesn't permit as much light through, for just that lens the relationship between f-stops holds. That's a bit of a guess on my part, so if someone can confirm, that would be great.