How much IRE diff is acceptable for an fixed aperture zoom lens?

Igelkott Film

Igelkott Film

Well-known member
Hi,

I'm am very stressed over my sigma 17-50 f2,8.
The iris blades got stuck somehow and it did not act as a constant 2.8, the colors got really bright and ugly zooming out.

Now i got rent lens while mine in the shop, its the same sigma 17-50.

But now when i test the lens against a white reflector and with the cameras spot % function, its telling me that i have 6-8 IRE difference zooming from 17-50

I have set it to 70 IRE at 17mm and at 50mm its 64 IRE.

I'm wondering whether that's normal or just a shitty crappy product from Sigma?

If you do this test with your fixed aperture zooms and post what its performance is, set it to 70 IRE and zoom all the way in.
I'll be very interested in the results.:thumbup:
 
i've noticed this same thing with canon zoom lenses and my 7D. you can see an example of the problem here:

http://vimeo.com/18032392

after much discussion on a number of boards the best we could come up with is that longer focal lengths require more light for the same exposure. so the physical size (diameter) of f5.6 at 300mm needs to be much larger than f5.6 at 24mm in order for both aperture values to let in the same amount of "exposure" at a given focal length.

on canon lenses at least it appears the light value is constant when the lens is wide open (as it would be when framing stills - the camera is always wide open until the shutter is released at which point the correct aperture is engaged) but as soon as you stop down you see the effect. unfortunately this doesn't translate to zooming during live video recording and you can see the results above.

this is a subject that always seems to prompt message board users to reply with "make sure you have auto exposure turned off" and it seems that many people just don't notice it. best bet is to stick to lenses that are dedicated to video production or use primes.

so far i haven't found any canon zooms (L series or otherwise) that are not affected by this issue. i've repeated the problem with multiple bodies, including 5d mk ii, 7D and a 60D. i've also seen it to varying degrees with every canon zoom i've tried (even multiple versions of the same zoom at my local pro camera shop).

hope that helps - you're not crazy! it does happen!

karl
 
Last edited:
Oh your lens is not doing a very good job :/


"In a constant aperture lens for instance the new Tamron 28-75 f2.8 DI.

As you zoom the lens the actual physical diameter of the lens opening changes to maintain the same fstop based. The mechanisim inside the lens that accomplishes this is the priamry reason constant aperture zooms are so expensive and so much larger than the variable aperture cousins."

After more testing, having it at 49mm then pushing it to 50mm it constantly drops 6 IRE.
And zooming from 49mm to 17mm is a journey from 67-71 IRE...

Its impracticable to set it to 70 ire and zoom in/out and com back to that number...
 
Last edited:
It's the transmission of the lens.
If the lens was f-2 all the way through then that's the diameter of the entrance pupal divided by the focal length. So it's a theoretical number based on that information, it doesn't take into account the light absorbed by the lens itself. That's what makes the difference between f-stops and t-stops.

Cinema glass is often measured in t-stops as the lens is put on a bench and accurately measured and calibrated taking the "transmission" or light absorption into account. So although your lens might be f-2 all the way through it's not t-2 all the way through.

Try out a cinema grade zoom that's measured in t-stops and you will see a difference.

Hope this helps.
 
Last edited:
Exactly -- it is not at all unusual to see a still camera zoom lens listed as "F 3.5/4.5" It is a 3.5 wide, a 4.5 pushed out.
 
Did you check and confirm that the white reflector was lit evenly? ie. with an incident or spotmeter?
 
Shooter said:
Did you check and confirm that the white reflector was lit evenly? ie. with an incident or spotmeter?
Click to expand...

No i didn't.. i just tryed to stay in the same place with a tripod and the spot meter in the center.

But the light loss is more apparant then just that, if I aim at the wall with some 70% zebra, I zoom in and the rand of zebra disappears.

I suspect that the iris blades do not close and open completely friction free.
 
Did you understand what I posted earlier? You will not get the same amount of light in a zoom lens at each end of the focal range even at the same f-stop. Its not how a lens works.
The difference in light is the "transmission" or the light that is getting absorbed as it travels down the lens.
What you are experiencing is normal.

F-stops are N=F/D

Where N is the f-stop number and F is the focal length and D is the Diameter of the entry pupil. Beyond the entry pupil nothing is taken into account.
So as the light travels down the lens some is absorbed by the lens itself.

Shooter, you know your stuff, am I missing what we are talking about??
 
@endcredits

Igelkott has a lens claimed to be constant aperture and he is perturbed that it isnt. I would be too!

Must admit that I have never ever tested a lens for f-stop constancy thru its range. I assumed if the lens is constant...it is constant and I have never seen one perform otherwise noticeably. At the end of the day... still lenses are still lenses and we should not forget this compromise.

It would be super disturbing to find this in a cine zoom.

I do know that it takes little lighting control to light perfectly flat and constant especially when one is going from wide to close. It would be the first thing I would check. If is was unevenly lit...I would expect the IRE (levels) to change. The flatness of the lighting should be visible in the wfm. (Horizontal and thin line)
 
A lens that is consistent aperture through out the focal range measured in F-stops is only a measure of the light passing through the entry pupil.
If you are measuring the light getting to the film plane then you are referring to T-stops. Cine glass is usually measured in t-stops.
f4 & t4 are not the same, they are close but not the same. f4 is the same on all lenses but t4 is different for every lens.

I don't mean to be rude to any one, I just thought the answer was rather clear. I also started to doubt my self. But the difference in ire that you are getting sounds about correct. Lenses absorb some light. The longer the focal length the more light it absorbs. T-stops take that into account.

That's the fundamental difference between T-stops and F-stops.
Have a google at how aperture is measured. (again not trying to be rude or abrasive)

http://en.wikipedia.org/wiki/F-number this might help.
The bit on T-stops explains the situation clearly, and better than I have.

Igelkott I think your lens is fine.
 
Last edited:
FWIW

I ran tests on my 2 x Nikkors this morning. Both f2.8 constant.

Flat lit grey scale card. I exposed it right in the middle of the wfm (50%)

The level on the IRE scale did not change at either end of respective ranges. 17-35mm or 28 - 70mm

To be noted though.

The line on the wfm is very thin when at long end of focal length. The line predictably thickens (expands vertically) at the wide end. This makes determination a little less precise. For the test... I centred on the wider line and when zoomed in...the thin line was on the wfm line at 50% where I exposed it. Takes a bit of fiddling to achieve but my lenses are declicked which helps.

Conclussion: Ffor all practical purposes...I trust my constant aperture zoom lenses remain constantly exposed at each end of their ranges. Practically means I would not have to adjust exposure but n real life I probably would dos so anyway. I am usually of the thought that I expose with more contrast and darker for wide shots and less so for close ups (eg. a face)
 
You must log in or register to reply here.
Back
Top