are faster lenses faster across all apertures?

T

thegoldfi

Active member
this may seem like a stupid question (and i apologize for having plenty of them lately):

is a faster lense faster at all apertures? for example, if you compare a f1.2 lens to a f1.8 lens, but are shooting at f2.8, will they be equal, or will the f1.2 lens be superior?

many thanks,
tG
 
Yep. The "faster" lens means it can open up wider than the other. So in your example of 1.2 vs. 1.8, they'd be the same at 1.8, but the 1.2 can open up to 1.4 or 1.2 as well.

Now, also, when we say "they'd be the same at 1.8" -- that's not necessarily meaning the same. They'll be the same sensitivity, but lenses typically perform best when stopped down 1.5 to 2 stops, so the lens A (the 1.2) might be a whole lot sharper and crisper when set to f/1.8, as compared to lens B (which has a maximum of 1.8).
 
Well, maybe, actually. While the "F" stop is the measure of the aperture size, this does not necessarily equate to the light transmitted to the sensor or film. That is why expensive, well made cinema lenses are rated in "T" stops, or "Transmission" stops. It is possible that a higher quality lens, with a higher "F" stop maximum may actually transmit more light then a lower quality lens with a lower "F" stop maximum. Now when comparing lenses (Prime or single focal length Lenses), the amount of light transmitted should be consistent across the aperture range.
 
David G. Smith said:
Well, maybe, actually. While the "F" stop is the measure of the aperture size, this does not necessarily equate to the light transmitted to the sensor or film. That is why expensive, well made cinema lenses are rated in "T" stops, or "Transmission" stops. It is possible that a higher quality lens, with a higher "F" stop maximum may actually transmit more light then a lower quality lens with a lower "F" stop maximum. Now when comparing lenses (Prime or single focal length Lenses), the amount of light transmitted should be consistent across the aperture range.
Click to expand...

Hehe! I'd be pissed if after taking meticulous light meter readings, some lens I paid premium buck$ for over-exposed my footage.

f:4 is f:4.
 
Actually, I believe David is right. F stop measurements may be close-to-equal but it's been shown in Still Glass comparison that SOME lenses are actually faster at certain apertures than others.

I believe the Nikon and Canon zooms show this in comparison to each other.
 
AFAIK

the term 'faster' derives from 'being able to shot at faster shutter speeds'
(because you can let in more light)
 
J Davis said:
AFAIK

the term 'faster' derives from 'being able to shot at faster shutter speeds'
(because you can let in more light)
Click to expand...

that is what i thought. and i was also wondering if an aperture of f1.8 is the same size, regardless of lens.
 
dufloorman said:
f:4 is f:4.
Click to expand...
f/4 is indeed f/4, mathematically. And f/4 on one lens will achieve the same depth of field as f/4 on another lens.

But the f/number does not mean that amount of light is transmitted! David's right, look into T-stops. Cinematographers use the T-stop (which is a measurement of light transmission) for lighting; they use the f-stop for DOF calculations.
 
Barry_Green said:
f/4 is indeed f/4, mathematically. And f/4 on one lens will achieve the same depth of field as f/4 on another lens.

But the f/number does not mean that amount of light is transmitted! David's right, look into T-stops. Cinematographers use the T-stop (which is a measurement of light transmission) for lighting; they use the f-stop for DOF calculations.
Click to expand...

I will look into the T-stop thing. But regarding f-stops, there will certainly be slight variations in light transmission between two lenses, both set at a given F-stop. However, this is only acceptable to a VERY small margin, that margin being of "error". F-stop is, and always has been, a factor used to determine proper exposure ("EV"). F-stop does indeed determine the amount of light transmitted. Consider this. If F-stop were NOT a standardized measure, you would need a different light meter for every lens. Thankfully, we can safely assume that f:4 on lens A will transmit the same amount of light as f:4 on lens B.

25 years ago when I was suffering from Zone System Mania, I would actually test all of my lenses against a defined meter reading for accuracy. Variations certainly existed, but frankly the variety of shutter speeds among my lenses dwarfed any variations in effective aperture (view camera lenses contain both iris (F-stop) and shutter). Granted, these all involved mechanical shutters. The electronic shutters in the 7D and 5D markII are near-perfectly accurate.

Slight variations aside, "F-stop" is a standard measurement of a lens' light transmission. This is why Canon did not make a 50mm f:1.4 and a second "faster" 50mm f:1.4. Rather, they made a 50mm f:1.4 and a 50mm f:1.2.

If I had to set a practical limit on acceptable variation from lens to lens regarding light transmission at a given F-stop, I'd say +/- 1/4 stop. Anything beyond that enters the realm of faulty calibration.
 
Again, look up why all professional cinematography lenses are calibrated with T-stops. The difference is not usually very large, nor is it necessarily "marginal"; for example, an f/2.8 lens might be marked as T3.1. Which is about 1/4 stop.

F-stop is a measurement of the diameter of the iris, and of course it's a guide to the amount of light transmitted, but only in a perfect world is it a truly accurate representation of the amount of light transmitted. Which is, again, why the T-stop was invented.
 
dufloorman said:
I will look into the T-stop thing. But regarding f-stops, there will certainly be slight variations in light transmission between two lenses, both set at a given F-stop. However, this is only acceptable to a VERY small margin, that margin being of "error". F-stop is, and always has been, a factor used to determine proper exposure ("EV"). F-stop does indeed determine the amount of light transmitted. Consider this. If F-stop were NOT a standardized measure, you would need a different light meter for every lens. Thankfully, we can safely assume that f:4 on lens A will transmit the same amount of light as f:4 on lens B.

25 years ago when I was suffering from Zone System Mania, I would actually test all of my lenses against a defined meter reading for accuracy. Variations certainly existed, but frankly the variety of shutter speeds among my lenses dwarfed any variations in effective aperture (view camera lenses contain both iris (F-stop) and shutter). Granted, these all involved mechanical shutters. The electronic shutters in the 7D and 5D markII are near-perfectly accurate.

Slight variations aside, "F-stop" is a standard measurement of a lens' light transmission. This is why Canon did not make a 50mm f:1.4 and a second "faster" 50mm f:1.4. Rather, they made a 50mm f:1.4 and a 50mm f:1.2.

If I had to set a practical limit on acceptable variation from lens to lens regarding light transmission at a given F-stop, I'd say +/- 1/4 stop. Anything beyond that enters the realm of faulty calibration.
Click to expand...


Now we see the differences that these new convergence cameras bring to the fore. I have absolutely no experience in still photography. However, I have seen variations in exposure, at similar rated "F" stops, of up to a full stop, when trying to mix and match cheap lenses shooting film. Now this is more pronounced with zoom lenses, than with primes. I have had some embarrassing incidents because of this actually (shooting slow film stocks at WFO and then finding out that F1.4, or what ever, ain't so) and have a taste to test lenses to make sure of their performance. I have been considering getting a HD-DSLR and have been looking around at lens options, and it is dizzying. The best way to go, it seems to me, is with primes. Can some one answer me this, does anyone make still zoom lenses that are faster than F 3.4/4.0? And if they do, are there any that can be purchased with mortgaging the farm?
 
Barry_Green said:
Again, look up why all professional cinematography lenses are calibrated with T-stops. The difference is not usually very large, nor is it necessarily "marginal"; for example, an f/2.8 lens might be marked as T3.1. Which is about 1/4 stop.

F-stop is a measurement of the diameter of the iris, and of course it's a guide to the amount of light transmitted, but only in a perfect world is it a truly accurate representation of the amount of light transmitted. Which is, again, why the T-stop was invented.
Click to expand...
I checked it out. Interesting stuff. Cinematographers a few decades ago would bench test their lenses (much as I described my testing my view lenses) and calibrated to their lens which was presumably the most efficient at a given F-stop. So they would maybe have one lens for which T:5.6 was also f:5.6. Another lens' f:5.6 stop might be marked "T:6.3" after testing the actual luminance at f:5.6. A good idea, especially for cinematographers. Us still shooters could just bracket 1/2-stop. Those guys don't have that luxury.

In short, the "F" and "T" mean roughly the same thing. Though I kinda laugh that lens manufacturers would switch to the "T" designation, given that the whole idea was to calibrate the lenses in a given DP's specific collection of glass. But the topic is very interesting indeed.

As to the OP, the question was about still lenses where the T-stop doesn't really apply (though it actually could I guess). Rather, the question was if the transmission at any given F-stop between two lenses of differing maximum aperture design is different. The best answer is probably "no." If tested, might one find as much as a 25% (1/4-stop) variation? Perhaps, but given that those faster more expensive lenses usually incorporate more elements in more groups, I'd bet that they would prove to be the less efficient at a strictly defined and measured aperture.

But the T-stop system makes perfectly good sense.

Love the discussion!! : )
 
David G. Smith said:
...I have seen variations in exposure, at similar rated "F" stops, of up to a full stop, when trying to mix and match cheap lenses shooting film. Now this is more pronounced with zoom lenses, than with primes. I have had some embarrassing incidents because of this actually (shooting slow film stocks at WFO and then finding out that F1.4, or what ever, ain't so) and have a taste to test lenses to make sure of their performance. I have been considering getting a HD-DSLR and have been looking around at lens options, and it is dizzying. The best way to go, it seems to me, is with primes. Can some one answer me this, does anyone make still zoom lenses that are faster than F 3.4/4.0? And if they do, are there any that can be purchased with mortgaging the farm?
Click to expand...

David,

That 1-stop variation sounds VERY frustrating! I once had a lens that developed a sticky iris, and would not close down during exposure. Dampened my elan!

I think an exposure error of a full stop probably indicated a malfunction.

And yes, it seems the price of lenses roughly doubles with every increasing aperture diameter. But I'm told that the difference in design and build between mid-range lenses and "fast" lenses more than explains the cost difference. Lens design becomes vastly more challenging when talking about fast glass.

I too am looking at getting my first HD DSLR. I'll be putting my beloved Beseler 57MB cold light enlarger w/ three Schneider lenses up for sale this weekend, along with my Hasselblad w/ three lenses (50mm, 80mm, 150mm) and Nikon FA w/ 80-200 f:2.8. Blows my mind that all that will probably just cover a 7D, fast wide-angle, fast 35mm and an inexpensive but usable tripod/fluid head.

But I'm keeping my Deardorff 5x7, dammit! I can always make contact prints : )

Good luck with whatever camera you get. Are you looking at the 7D? GH1?
 
The back element tells it all. Say you have two 50mm lenses. One is 1.2 and one is 2.8.
If both set to 2.8, the one with the bigger back element diameter will be faster.

We tested that several times. Works always, thou the differences are minimal.

Frank
 
As stated before, check reviews of your lenses. At their maximum aperture (minimum f-number), that's where most lenses are least sharp, distort the most, and are darkest around the edges (vignette).

So an f1.4 lens at f2.8 is often (usually) sharper than an f2.8 lens at f2.8.

Those little f-stops are great for the d.o.f.-crunching look that the 7d cinematographers are often going for. That's why the 50 mm f1.8 or f1.4 was recommended to you, because you're probably going to want those nice, defocussed backgrounds with good bokeh. Those faster prime lenses at stopped down to f2.8 will look nicer than an f2.8 zoom wide open.
 
steventimothy said:
As stated before, check reviews of your lenses. At their maximum aperture (minimum f-number), that's where most lenses are least sharp, distort the most, and are darkest around the edges (vignette).
Click to expand...
Absolutely true.

However, for some of those elements, when used on a 7D, they become of less concern. Corner sharpness and vignetting, for example -- because the 7D doesn't use a "full frame" sensor, it won't see all the way into the corners, so a lens that vignettes or gets soft in the corners on a "full frame" camera, may still look just fine on a 7D or GH1.

However, one thing that doesn't change is chromatic aberration and softness. Lenses usually get aberrational and soft with flat contrast when used at wide open, even my premium ZF set gets nasty chromatic aberration and flat contrast at wide open, and that shows up on the whole frame, not just the corners.

So an f1.4 lens at f2.8 is often (usually) sharper than an f2.8 lens at f2.8.
Click to expand...
Usually, and I'd dare say almost invariably (all other things being equal, of course).

But, then again, it depends on what purpose you're using the lens for. If it's for still photography, this can matter a lot. If it's for video on a DSLR, it doesn't matter all that much. Even a mediocre lens wide open is probably passing enough detail to give a DSLR all that it can handle, when it's in video mode.

But for stills, yes, better glass being used a stop or two down from wide open will definitely give you much better results than cheap glass used wide open.
 
Thanks, Barry, for being more blunt; I was trying to be tactful. But you're right... It's 2 Megapixels! It's 1.6 Megapixels actually!

Just look at Barry's tree still comparing a video Hd still to a downsampled photo still, and you'll notice that a little aberration or distortion *doesn't matter!*

Still, better to be future-proof, I guess, for when Canon releases the hoped for super-codec, aliasing-eliminating, deluxe cinema DSLR (as a firmware update) : )
 
steventimothy said:
Still, better to be future-proof, I guess, for when Canon releases the hoped for super-codec, aliasing-eliminating, deluxe cinema DSLR (as a firmware update) : )
Click to expand...
Exactly why I torched $7k (so far) on ZF glass. Complete massive ridiculous overkill for a DSLR, but should be well positioned for an S35 Scarlet or a proper large-chip video camera when such ever arrives.

For those who can't afford it, don't sweat it -- if you're using lenses for video mode on a DSLR, you do NOT need super-sharp lenses.
 
You must log in or register to reply here.
Back
Top