Lens Flair, Aperture Issues?

Tim Carroll · March 8, 2009

I shot a number of test with a set of Cooke Kinetal lenses, and I am trying to interpret the results. Any and all help would be appreciated.

The first test was just with the 12.5mm, the 17.5mm and the 25mm.

Unfortunately the frame grabs don't make it show up as well as it did in the telecine suite, but the 12.5 and the 17.5 look more washed out compared to the 25, particularly in the center of the frame, around her face. The 17.5 looked worse than the 12.5 and is even washed out from the 12.5 if you look at the yellow bike frame.

The second test was with the 9mm, 12.5mm, 17.5mm, and the 25mm.

The issue is not as glaring in these shots, but the 12.5 and particularly the 17.5 are again lighter and a bit more washed out.

I then set up a test with a combination of daylight, tungsten and overhead fluorescent light and shot with all five lenses in the set (9mm, 12.5mm, 17.5mm, 25mm, and 37.5mm).

It is more noticeable with these frame grabs but still doesn't stand out as much as it did in the telecine suite, but again the 12.5 and the 17.5 look really burned out in the center part of the frame, with the 17.5 looking the worst. The lighting stayed the same from shot to shot and the aperture was set the same on each lens in all three tests. It was set at f4 in the first test and f2.8 in the second and third test. Is that lens flare, or some kind of iris issue, or do you just need to find a way to light around that characteristic of a lens?

Any and all input is appreciated.

Best,

-Tim

David Mullen ASC · March 8, 2009

It may be that the lens does not have T-stops, just f-stops, and that they aren't matching, hence why one lens is overexposing more than another. I'd shoot a grey scale at the same aperture on each lens and look at that in the telecine.

But yes, older lenses often mismatch in color and contrast a little, requiring some work in post to match them.

Hal Smith · March 8, 2009

Hi Tim,

What really screams out at me is the change in color of the front of the oven. It goes from gray to blue to blue with a hint of lavender. There's enough change there that I would expect it to be noticeable in the viewfinder. If it isn't particularly, then I suspect not all the differences are from the Kinetal's color balance and transmission.

Try setting up the shot again to visually compare the lenses, also have a desklamp or similar fixture off in the corner somewhere lighting a white piece of paper. Keep referring back to that piece of paper to keep your eye neutrally white balanced.

Do you have any record of printer lights or the equivalent? I'll bet David can make some sense of that sort of information.

Tim Carroll · March 8, 2009

David,

The lenses are marked in T stops but I am thinking the scales are just off and I need to do the grey scale test you were referring to.

Hal,

The 25mm is starting to straw and so is the 37.5mm a little bit. Unfortunately in the 1st and 3rd test, when I was at the transfer session, I did not get the numbers from the colorist. Also, as kind of a newby at this DP stuff, I was a little confused by everything he was doing. I kept asking if he could just give me all the lenses at a neutral state, with no color correction, but for some reason that just kept not happening. I was not present at the transfer of the footage from the second test, the footage of the lathe.

Best,

-Tim

Stephen Williams · March 8, 2009

Hi Tim,

30 years ago it was normal to go to a rental house and test 3 or 4 lenses in every focal length to get the best matched set. Some lenses would color match better than others, some would be softer wide open. Depending on the shooting stop you would select different lenses. That all changed with the PV Primos, followed by Ultra Primes Cooke S4's & Master primes.

Stephen

Chris Keth · March 9, 2009

The 17.5 on the oven is so startlingly different that I wonder if you didn't accidentally shoot that one wide open rather than at T2.8.

The others, though, look very close in exposure to me. So close, in fact, that I think some psychological tricks are happening where the enlargement and different composition from one lens' shot to the next makes you think the brightness is changing when it may not be.

Have you tried enlarging one part of each frame so you have matching frames taken with each lens? Our eyes are affected by everything in the picture so maybe having the same frame for each lens, via cropping, would help make it easier to judge. For example, crop the wider lens versions of the shot with the lathe so they all match the 25mm shot. There are small difference there but, if you look only at the top of the the lathe's control housing at that fairly uniform grey, they are very close in exposure. That crop test might help put your judgement of what's really happening in better perspective.

Chris Keth · March 9, 2009

Hi Tim,

30 years ago it was normal to go to a rental house and test 3 or 4 lenses in every focal length to get the best matched set. Some lenses would color match better than others, some would be softer wide open. Depending on the shooting stop you would select different lenses. That all changed with the PV Primos, followed by Ultra Primes Cooke S4's & Master primes.

Stephen

You say 30 years ago but I still check them on a projector and have rejected lenses, on rare occasion, from those very new well-matched sets. I found an S4 not long ago that was shamefully yellow compared to the others. It's not common but it does happen. That S4, BTW, had a serial number very different from all the others so I guess it was just too far in batches of glass from the others.

Stephen Williams · March 9, 2009

You say 30 years ago but I still check them on a projector and have rejected lenses, on rare occasion, from those very new well-matched sets. I found an S4 not long ago that was shamefully yellow compared to the others. It's not common but it does happen. That S4, BTW, had a serial number very different from all the others so I guess it was just too far in batches of glass from the others.

Hi Chris,

Idealy S4 serial no's should match accross the set.

Stephen

Satsuki Murashige · March 9, 2009

30 years ago it was normal to go to a rental house and test 3 or 4 lenses in every focal length to get the best matched set... That all changed with the PV Primos, followed by Ultra Primes Cooke S4's & Master primes.

Hi Stephen,

I've heard from several different people I trust (not including Chris K. ;)) that the modern lens sets aren't always as perfectly matched as they are advertised to be. I think one reason why lens matching doesn't seem to be an issue these days is because the heavier "looks" commonly applied to the footage in post tend to mask inconsistencies in the glass.

Olex Kalynychenko · March 9, 2009

It may be that the lens does not have T-stops, just f-stops, and that they aren't matching, hence why one lens is overexposing more than another.

I can add:.

The design of lens included module with aperture mechanism.

This is disk with aperture blades inside and slit with moving ring.

The outside ring with aperture scale connect with moving ring of aperture mechanism by long screw.

And the real value of aperture ( on aperture mechanism) can be adjust by turn of body of aperture mechanism inside of main body of lens.

That's why, i can suppose, the lenses have different value of real ( T ) aperture with similar position of aperture ring with scale.

Need use of special devices or similar, hand made, devices for test and adjust of aperture control of lenses.

Serge Teulon · March 9, 2009

....I think one reason why lens matching doesn't seem to be an issue these days is because the heavier "looks" commonly applied to the footage in post tend to mask inconsistencies in the glass.

The way that post these days allow us to fix irregularities, means that we can take those chances and cut certain corners far more than in years gone by!

Chris Keth · March 10, 2009

I've heard from several different people I trust (not including Chris K. ;))

Ooh, wicked burn. :(

Jokes aside, I figure that it only takes me a few minutes to check the lenses on a projector and, when I find ones that don't fit that the off color one I mentioned above, it makes those few minutes worth it.

Satsuki Murashige · March 10, 2009

Ooh, wicked burn. :(

Well, what I meant to say was, "in addition to Chris."

Sam Wells · March 10, 2009

I think your expectations for these lenses may be a bit too high. They are not going to have the resistance to flare / veiling glare as more modern glass. The Kitchen shot with the 17.5 does look a bit overexposed with respect to the others, but it might be lens internals as well - shooting into the light source you are flashing the negative in a sense....

-Sam

Tim Carroll · March 10, 2009

They are not going to have the resistance to flare / veiling glare as more modern glass. The Kitchen shot with the 17.5 does look a bit overexposed with respect to the others, but it might be lens internals as well - shooting into the light source you are flashing the negative in a sense....

-Sam

Sam,

That is one thing I was thinking about. I am doing the grayscale test David talked about, and that should resolve the T-stop issue, but I am wondering if that lens is just flaring badly. The "hot" areas of the other tests, the model's forehead and face, and the chuck on the lathe, all seemed to cause the same issues with that lens.

Best,

-Tim

Tim Tyler · March 10, 2009

The lighting stayed the same from shot to shot...

In the kitchen, looking specifically at the window light reflection on the right, camera-facing cupboard door (above the paper towels), I would suggest that the daylight level was actually changing.

Tim Carroll · March 10, 2009

In the kitchen, looking specifically at the window light reflection on the right, camera-facing cupboard door (above the paper towels), I would suggest that the daylight level was actually changing.

Good pickup Tim, I hadn't noticed that before. It was a typical diffused light Portland afternoon (meaning lots of cloud cover), but it does look like the sun broke through the clouds on that shot.

Best,

-Tim

Tim Carroll · March 10, 2009

I think your expectations for these lenses may be a bit too high. They are not going to have the resistance to flare / veiling glare as more modern glass. The Kitchen shot with the 17.5 does look a bit overexposed with respect to the others, but it might be lens internals as well - shooting into the light source you are flashing the negative in a sense....

-Sam

I did find one other test shot with these lenses, at least the 12.5mm, 17.5mm, and 25mm. The light was typical Portland overcast coming through a big bay of windows, with the mini-mag light and another flashlight shining on the white coffee mug. This was shot with a camera that turned out to be out of adjustment (FFD was long and ground glass was not set properly) so everything is soft. But it does not show flaring too badly in the 17.5mm shot, and all these were shot wide open at T2.

Best,

-Tim

Chris Keth · March 10, 2009

Well, what I meant to say was, "in addition to Chris."

I was kidding. I know nothing was meant by it :P

Tony Brown · March 11, 2009

But yes, older lenses often mismatch in color and contrast a little, requiring some work in post to match them.

....and the nomination for understatement of the year goes to..... :)

PV 'C' series anyone?

Tim Carroll · March 21, 2009

It may be that the lens does not have T-stops, just f-stops, and that they aren't matching, hence why one lens is overexposing more than another. I'd shoot a grey scale at the same aperture on each lens and look at that in the telecine.

But yes, older lenses often mismatch in color and contrast a little, requiring some work in post to match them.

Well I did the gray scale test and sure enough, the 12.5mm lens was off by about 1/3 to 1/2 of a stop and the 17.5mm lens was off by 2/3 to a full stop. So now I know.

Thanks for everyone's help.

Best,

-Tim