Tokina 10-17mm fisheye overexposed photos

Hello

I have this weird issue with Tokina 10-17mm fisheye and light meter in my Nikon D500. If I set the aperture, shutter speed and ISO so that light meter reads 0.0 (matrix metering), photos are overexposed by about 3 stops. So I need to dial in -3.0EV exposure compensation to have at least approximately reliable light meter. Why approximately? Because it's not always off by 3 stops, sometimes it's off by 2 stops, sometimes by 4 stops, etc. It really depends on the conditions I'm shooting in. But I didn't figure out a pattern yet.

I've checked the aperture blades and they all seem to move normally (I can't see any stuck blades). I also did a test where I've set the optimal exposure on f3.5 and then progressively closed the aperture and took a photo for each 1/3 step. Photo was getting darker as one would expect, which again confirms that aperture blades are not stuck. Contacts on the lens also seem pretty clean to me.

Light meter on the rest of my Nikkor lenses works perfectly fine, so camera can't be an issue.

Is this a normal behavior for a fisheye (Tokina) lens or did I just get a bad copy?

Hi Andrej
I don't have a definitive answer on this, but I'm facing the same issue as yours with mine and my Nikon D90.

A friend of mine told me that the exposure metering was really disturbed when working at extreme fov, which is the case @10mm with fish eye projections. 
 

He suggested to try at the @17mm focal length to see if the issue was still there but I haven't tried yet

Anyway, I agree that's upsetting, any advice to overcome this is accepted ! 

I tested this theory out in my living room right now, but unfortunately it doesn't help. At matrix metering mode exposure meter reads the same value when zoomed out to 10mm, as when zoomed in to 17mm. In case of center weighted metering it showed max 2/3 of a stop difference, but that may be due to the subject filling more of the center of the frame as I've zoomed in. But still the metering was about 2EV off (overexposed). Repeated the test with Nikon 17-55mm under the same conditions and on each end of the zoom range the exposure was perfect.

Is it possible that aperture doesn't close as much as the camera assumes that it will? I did notice on my copy of the lens that the lever which opens the aperture on the lens has some play. I can move mine to about 1/3 and it doesn't affect how the aperture is opened. It only starts opening the aperture from 1/3 on. But even if that was the case, I guess the metering would be correct at the fully opened aperture, but it's not.

Really odd. I use the D500 and Tokina 10-17mm combination a lot and have never come across this issue. 

Do you get the same above and below water?

My inclination would be something with the lens contacts? Maybe worth cleaning them?

Try it at different zoom positions. Some versions have sticky apertures when slightly over zoomed at one end or the other.

Even when zoomed to a different position, there's no difference in metering. This problem is showing on land as well as in water. It's easiest to reproduce on scene with the blue sky taking over approximately 50% of the image.

Though with all this testing and comparisons I did figure out that perhaps the best way to meter with (at least with my copy of) this lens is to use spot meter with highlights warning. Only this way I got somewhat consistent results without severely overexposed highlights.

Wow, that is really annoying. No chance you can send the lens back, I guess?

It certainly seems like it is faulty in some way.  Have you tried placing it in manual and setting shutter speed and aperture that gives the correct exposure with another lens like the 17-55 you mention.  If the exposure is incorrect then it then it is not producing the same f-stop as the camera is instructing it to. 

24 minutes ago, ChrisRoss said:

If the exposure is incorrect then it then it is not producing the same f-stop as the camera is instructing it to. 

Yeah, agreed - hence my suggestion of a dodgy or dirty pin

I've tried this now and with identical settings that give good exposure on 17-55 at 17mm, image is noticeably brighter with 10-17 at 10mm. By my estimation by at least 1EV. Light meter also shows that image will be 1 stop underexposed with Tokina. So if I would set it up so that the light meter reads 0.0, it would be actually about 2EV overexposed.

What I don't understand is how a dirty pin would affect the aperture opening and closing. Isn't that done manually by camera via an aperture pin? And if there is a problem with the aperture closing down insufficiently, wouldn't then the exposure be spot on with the maximum aperture setting? It's not the case with my lens; even f3.5 gives overexposed images.

A wild theory: is it possible that firmware upgrade for D500 would break the metering algorithm for 3rd party lenses? @TimG could you please check which versions you use on your D500? Mine are C 1.30 and LD 2.013

From dpreview of this lens. Definitely not a new problem as this review is from 2012

7 hours ago, Andrej Oblak said:

What I don't understand is how a dirty pin would affect the aperture opening and closing. Isn't that done manually by camera via an aperture pin?

Doesn't the lens have its own motor which is driven and linked to the camera by the the pins on the lens?

7 hours ago, Andrej Oblak said:

A wild theory: is it possible that firmware upgrade for D500 would break the metering algorithm for 3rd party lenses? @TimG could you please check which versions you use on your D500? Mine are C 1.30 and LD 2.013

I would do that with pleasure but am not at home for the next week or so - sorry. I know though I am using whatever the latest upgrades are for the D500. You could check that on the Nikon firmware upgrade website. I was using my 10-17 only last week and it was fine.

7 hours ago, Andrej Oblak said:

 

What I don't understand is how a dirty pin would affect the aperture opening and closing. Isn't that done manually by camera via an aperture pin? And if there is a problem with the aperture closing down insufficiently, wouldn't then the exposure be spot on with the maximum aperture setting? It's not the case with my lens; even f3.5 gives overexposed images.

 

Depends on the lens, newer Nikon lens do this electronically, on older lens it's a lever that stops the lens down.  The Tokina literature indicates this lens needs a Nikon body that has an AF motor so it seems like quite an old design and I would guess it is mechanical - you mention also an aperture lever.  So that mechanism could be sticky.  But you have no aperture ring and the camera sets the aperture and that part might be electronic - so possible it is not receiving that signal properly.  I'm not totally up on all of the Nikon lens series they started adding electronic diaphragm control some years back. 

.

10 hours ago, TimG said:

I would do that with pleasure but am not at home for the next week or so - sorry. I know though I am using whatever the latest upgrades are for the D500. You could check that on the Nikon firmware upgrade website. I was using my 10-17 only last week and it was fine.

No worries, buddy, you gave me all the information I need I also use the latest firmware for D500 (which added support for CFExpress memory cards), so I guess this can't be the reason for metering issues.

Otherwise Tokina doesn't have its own autofocus motor; AF is screw driven by the camera. I'd say that aperture is opened / closed by the camera too, via an aperture pin on the lens. I seriously doubt that such an old tech would have electronically controlled aperture I'm already in contact with the local repair shop. I'll take it for repair next week (planning to do some more shooting this weekend) and then I'll post here what they find out.

However, I regret that I didn't buy Nikon 8-15mm instead.

Good luck on the repair.

I'm using the Nikkor 8-15 as well and just switched the other day from the 10-17 to 8-15.

I read an article (or maybe it was on the Wetpixel Live stream) comparing the 10-17 and the 8-15 where Alex M thought there wasn't much difference in IQ between the two when used on a DX sensor. Certainly not enough to justify the price difference.  I had the 8-15 anyway so I thought I'd give it a go myself and take a look.

Very early days and I've not really had chance to compare the two closely but certainly the 8-15 images with a 8" dome look pretty sharp.

I’ve used 3 different tokina 10-17 lenses over >10 years with a Nikon D90, D7100 and D7200 and have had exactly the same experience. When using an auto exposure mode, they invariably overexpose by 2 to 3 stops. I’ve accepted this as an issue with the Nikon metering system, and have learned to compensate whenever using an auto exposure mode, e.g. shutter priority for whales. When shooting in full manual mode for most diving with strobes, the problem disappears. It doesn’t bother me anymore and has become a normal part of my dive planning. 

48 minutes ago, Matt Sea said:

I’ve used 3 different tokina 10-17 lenses over >10 years with a Nikon D90, D7100 and D7200 and have had exactly the same experience. When using an auto exposure mode, they invariably overexpose by 2 to 3 stops. I’ve accepted this as an issue with the Nikon metering system, and have learned to compensate whenever using an auto exposure mode, e.g. shutter priority for whales. When shooting in full manual mode for most diving with strobes, the problem disappears. It doesn’t bother me anymore and has become a normal part of my dive planning. 

 

Thanks for this message Matt. It kind of reassures me that it's not just my lens that is broken, but looks like Tokina's quality control is pretty bad.

@TimG image quality with Nikon may not be drastically better than with Tokina (haven't compared it yet, so I can't comment on this), but build quality definitely is, as well as better reliablity, works on Z bodies, etc.

3 hours ago, Andrej Oblak said:

but build quality definitely is, as well as better reliability, works on Z bodies, etc.

All  very good points, Andrej

My brand new Tokina 10-17 mounted on a D800 set to A definitely overexposes. Photos taken in a semi-dark room look as if lit by sunlight.  I plan to shoot with it using M, manual settings, as with my other lenses.