Different sensitivity in one side of the sensor

uncristobal

Hi guys.
I don't know if this is a common problem, it may be... I have noticed that the sensor from at least two of my nikon cameras has more or less sensibility depending on which side of the sensor is. This is kind of difficult to explain for me because my english is terrible.
The thing is, when I shoot a white background, perfectly illuminated by two flashes or strobes or whatever, always one side of the image get one or two stops darker than the other side of the image, and if I turn the camera to the other side and take the same shoot, the darker side of the image also change, so is very clear that is a sensor problem. I want to clarify that I know how to illuminate a white background. I always use a photometer to get sure that the light is just how I want. I have seen this problem with a D200 and a D800.
Is anybody else with the same issue?, Is this normal?. For how Nikon charge as, it shouldn't.
My lens are all pro line, (24-70 2.8 G ED/N etc. etc...).

If someone has the same issue or know something about this I would really appreciate the info.
thanks a lot y saludos a todos!

C.

Ironheart

Oh boy! Where to start. First off, we need to see pictures, full size, with EXIF data. This will answer many questions such as ISO, speed, aperture, PSAM, etc... Secondly does this show up in normal photos or just pics of white walls with flash guns? Since this happens on more than one camera, we need to know about your setup. Which flashes, how are they arranged, how far from the wall? What is the position of the camera.

Sorry to say, but this sounds like something in your methodology rather than a camera issue.

Ade

Using too fast of a shutter speed (exceeding the shutter sync speed) might give the appearance of this problem. See if using a slower shutter speed (such as 1/160s) solves the issue. Problems with post-processing might also cause something similar.

As Ironheart says, it's unlikely to be a sensor issue.

Msmoto

The fact both a D200 and D800 show the same phenomenon, strongly suggests it is a methodology issue. Two f/ stops difference is not possible to explain by a sensor defect. My suggestion would be to use a long lens, shoot a wall of a building illuminated by very open shade. Repeat with a sun illuminated wall where the sun hits at about 45 degrees from directly behind the camera

uncristobal

Hi!,
thanks for your interest.
I completely understand what you're saying but the light setup is correct, two profoto in 45 degrees against the background at the same power and distance. Both measured separetly with a Sekonic and one profoto Three f/ stop lower than the background. up in front (a little to the left) of the model with an octa. Classic studio "lookbook" setup.
And is not a shutter speed synchro problem. Maybe is not as much as two F/ stop of difference but is very noticeable when you are doing the postpo in photoshop. One side give you the 255/255/255 values and the other give you 250/252/254 etc...
And this is what's concern me... When I rotate the camera 180 degrees, the dark side of the image change... Conclusion, is the sensor, not the setup. How can this happen with two different cameras in different situations??.
Im not sure if im explaining the situaction properly, idiomatics problems...
Next thursday I have a shooting with white background. I will take some images from that session to show you better and I will take the two cameras just for test

Thank you very much,

C.

TaoTeJared

Three f/ stop lower than the background. up in front (a little to the left) of the model with an octa. Classic studio "lookbook" setup.

It is not the sensor if you are showing it on both the D200 & D800. An issue like this would be posted everywhere and it is not. Octa will drop exposure 1-3 stops. Some lenses will actually meter higher/lower on either side (maybe 1/3 stop at most).

If you post some photos, also post what your set-up looks like. Also post the camera settings and filter (brand/type) you use. My guess is that something is adding light rather than dropping it. Strange how turning it 180 deg it moves with it.

Golf007sd

@uncristobal: Remove any filters you may have on your lenses. Use a custom WB. Test each body with the same lens and let us know which lenses are having the most noticeable issue with. I have to agree that it is very unlikely that this is a sensor issue than your setup/configuration.

Ade

The difference between 250 and 255 is very slight, not anywhere close to 2-stops. It's more like 1/5th of a stop.

That level of variance might come from your lenses instead of from the sensor. It's not unusual for lenses to show 1/5-stop variances (or more) from edge to center to edge, and between opposite edges.

So maybe you can try exposing the sensor without a lens. To help ensure the lighting is evenly illuminated, you can place a glass diffuser in front of the mount.

Also make sure your sensor is perfectly clean (maybe get it professionally cleaned) before testing.

If you want to get scientific, you can do a lens/sensor uniformity test using packages like Imatest, which can precisely map f-stop contours of your setup.

Benji2505

Take a picture of a white piece of paper with exposure time longer than second (paper poorly but evenly lit). Move camera during the shot. Take the pic out of focus. No other light source. How does the picture look?

TaoTeJared

And is not a shutter speed synchro problem. Maybe is not as much as two F/ stop of difference but is very noticeable when you are doing the postpo in photoshop. One side give you the 255/255/255 values and the other give you 250/252/254 etc...

I'm re-reading this and how you are testing this. It looks like you are taking color readings. It is fairly usual to have different flashes pump out different color especially if they are not OEM/Nikon. That may account for some of the discrepancy. Ade is correct on his approximate 1/5-stop variances and Golf007sd brings up a good point in the filters you are using. If you are not using a high end UV filter, you can see a difference in color as well. That could account for the 180deg shift as well.

Without sample photos it is really hard to tell.

uncristobal

Hi,
I'm using a B+W filter MRC nano XS - PRO. I think is a good filter, but i will do the test without it. At this point, for what i'm reading here Is probably a lens issue.

Ade is right, I make a mistake saying that is a 2 f/ stop, is much less than that.

And yes, i'm taking RGB reading in order to get a 255/255/255 background, I think is the only way to measure that?, but yes, you're probably right to. Could be a little difference between the color of the flash units but im using two Profotos from the exact model at the same distance, power and everything. Both units buy together so they have more or less the same usage (shoots).

I understand that is a very little difference between for example 250/253/255 and 255/255/255, but when you look at this in a screen is quite noticeable.

Next thursday i be able to upload some images after the photoshoot and we will see.

I really appreciate the interest and good comments on this. Is being very clarifying.
Thanks a lot!!

FlowtographyBerlin

Could be a little difference between the color of the flash units but im using two Profotos from the exact model at the same distance, power and everything. Both units buy together so they have more or less the same usage (shoots).

As long as you go close enough to the background with the camera (say something below 1 m with a portrait lens on), and a little further away with the monolights, you will have an image area that is so small that there should be no separation of the individual light cones and there should be no variation from that. I reckon you're using the D1's without a reflector for this one, right?

Also, don't aim for the full-blown (=255 white) exposure, you might wanna aim at one f-stop below, or even just the measured EV giving you a neutral gray.