@elvishefer
Try the same shot, but wrap the camera with a black velvet cloth or other good shading device to make certain no extraneous light is entering the camera anywhere. Tape it on well and see if this corrects the problem. Oh, was your eyepiece covered during this exposure? If not... most likely this is the cause of the problem.

I echo DN's comments. I have experienced similar issues using a CPL on a wide angle (20-24mm) lens. The polarizing effect is greatest at 90 degrees to the sun, and this would emphasize any banding resulting. And, the polarizing effect and the rotation of the CPL is situation specific in the sense that one needs to test or trim the filter on each photo. But a very nice image anyhow.

This discussion reminds me that it could also be a light leak. 2 minutes in daylight is not the same as 2 minutes at night (the normal environment for long exposure). I think it might not be a coincidence that there was a thread about light leaks with the 24-70. IIRC, it was coming in through the focal distance window, which could be covered with gaffers tape to fix the problem. I'll go see if I can find that thread.

Edit: http://nikonrumors.com/forum/topic.php?id=2227

Correlli said:
That is true. But if the banding line is always parallel to the sensor side and does not move with any rotation (camera or filter) then I would say it is sensor related, otherwise it can be an effect that is caused by the filter / filter combination.

What makes me think this is not really related to the CPOL is, that I have the impression the dark band is not only in the sky but also on the pavement and the grass.

Mmm if it is not related to the filters, then the OP would have banding on all of her/his pic.s.. , or at least on the long exposure ones.. ( using just the ND filter? ) .
Since I do not read anything about that in this post, i asume that its the combo causing this efect under this circumstances.....

Nice tree.

I think the vertical band is only in the sky. The green-yellow grass isn't directly below the sky band and might just be a bit of parched grass.

Correlli said:

What makes me think this is not really related to the CPOL is, that I have the impression the dark band is not only in the sky but also on the pavement and the grass.

Yea, this is rather interesting. How does one get this variation across the image by a polarizing process, when the light from the grass is not polarized? It may be some strange reflection through the various surfaces of the filters but to get the orientation which seems to correlate with the way the shutter curtains would be moving remains to be seen. I wonder if a several minute exposure at night without the filters might be done to rule out any shutter anomalies.

DutchNikon said:
Chalenge here is : The banding might not have been there just minutes later, because the angle of the sun changes, the humidity of the air ( causing almost invisible local reflections ) changes, a cloud drifts in or out of the picture etc., so in theory it's possible that this effect is different in a series of shots within minutes, while not changing any setting and not moving the camera 1 inch..

That is true. But if the banding line is always parallel to the sensor side and does not move with any rotation (camera or filter) then I would say it is sensor related, otherwise it can be an effect that is caused by the filter / filter combination.

What makes me think this is not really related to the CPOL is, that I have the impression the dark band is not only in the sky but also on the pavement and the grass.

Correlli said:
Although I agree that the strength of the polarization effect does change with the angle to the sun I have never seen such a quick transition resulting in that type of banding. I would try to get some more shots at different angles (first like you did before, then having the horizont at app. 45° and then landscape format). Same thing with the different filters (rotating them in 45° steps). That way you should be able to see if the banding moves with the sensor of with the filters.

Chalenge here is : The banding might not have been there just minutes later, because the angle of the sun changes, the humidity of the air ( causing almost invisible local reflections ) changes, a cloud drifts in or out of the picture etc., so in theory it's possible that this effect is different in a series of shots within minutes, while not changing any setting and not moving the camera 1 inch..

Although I agree that the strength of the polarization effect does change with the angle to the sun I have never seen such a quick transition resulting in that type of banding. I would try to get some more shots at different angles (first like you did before, then having the horizont at app. 45° and then landscape format). Same thing with the different filters (rotating them in 45° steps). That way you should be able to see if the banding moves with the sensor of with the filters.

Thanks all, you've given me a bit to think about, the least of which is that I may have been using the CPOL wrong all this time. :)

I knew facing perpendicular to the sun yielded the greatest CPOL effect, sometimes resulting in a color variation in the sky if the facing angle was off too much. But what I didn't understand was the effect of rotating the CPOL. I always thought that it was - for lack of better words - a pre-determined rotational cycle of least effect to greatest effect and back. That's why I proceeded to mark on the filter where the polarization effect was greatest, and then in the field I'd try to use that and my facing angle to bring the CPOL effect to light. No pun intended. I will have to go out shooting and do more experimentation.

Thanks for the kind words on the image itself; I'll get it right and post something in the PAD in a few days.

And who knew Tao was a budding sketch artist? ;)

Seriously, thanks everyone for your insight. It's what I like about this forum, people of broad and knowledgeable backgrounds willing to help n00bs like me.

Cheers,
Jeremy

Dutchnikon is right about orienting your polarizer in the situation you're going to use it. In this case because you won't be able to see much through the ND filter, you probably need to take it off and look through it while you get the orientation you want, then put it back on.

The sky's polarization varies. The area with the max polarization will be a band that you can find by pointing one hand at the sun and always keeping the other arm at a right angle to that one. The band of the sky that the second one can point to will look the darkest when the polarizer is rotated the right way. This is what DN was referring to about wide angles. You capture lots of different parts of the sky.

If I understand what you're asking about, it looks like clouds or a contrail passed through the frame. It could have been too thin to notice without the polarizer, but any light scattered from a cloud would be Mie scattered (white and unpolarized), while the light that gets the effect you want is Rayleigh scattered (mostly blue and partially polarized), so it would stand out in the photo.

Cool shot, elvis.

Like others have said, maybe you just had too many filters on your lens?

Having too many barriers to light is a bad thing- it scatters light just a tiny bit because you're changing the medium light is traveling in more.

Not that it should cause a significant change, but I would try shooting the same scene again without the UV filter just to see what happens.

>> CPOL is set to max, as far as I know. I did a test a while back and aimed the camera at my computer monitor and adjusted it until I couldn't see the screen. I then marked the rotation on the CPOL with a permanent mark <<

THis part i do not fully understand.. trimming a CPL filter on a monitor does not give anby indication of the optimal rotation in any other circumstances, the CPL needs to get trimmed depending on the actual direction of the light in shooting circumstances and direction.

Also Pol filters do not work well on wide angle lenses, they cause "Banding"because of the different directions of light within the shot, so what zoom setting was your lens at ? at 24mm its queit possible that it results in uneven exposure, at 70mm much less so..

in adition to that polarized light travels different to glass elements, so both the ND filter and the UV filter can contribute to unexpected effects...

ALl in all, to many variables to predict the effect of the combo of filters and lens i think..

That is a great Picture idea! I love it.

I have done some storm shooting (painfully unsuccessful - but that's a different matter) But I have seen something similar on my photos. What I noticed with a CPL and a 2xND filter was that darker sky crept in and at longer exposures it was really pronounced. I found it hard to describe so I drew a crude picture. Basically what I figured was that there are parts of the sky brighter than others and I was at the angle to catch both the light and dark parts and the filters really made it pronounced.

I probably didn't draw it the best - basically the sun is rising/setting and part of the sky closest to the horizon opposite of the sun is becoming darker. Then the FOV is at an orientation that it captures part of the light and dark sky.

That's what I figured what I did - of course at sunset you can visually see it. I'm guessing your ND filter actually brought out the darker sky.

OR... Occam's razor - your filter isn't actually uniformly colored. Maybe the variable has something to do with that where part of it could be darker. I would send the photo to B+W and see what they say. It would be interesting if there is a drawback to the variable filters.

I do love your photo though! That is cool!

What I see in the image from right to left is a dark band, about 20% of image, lighter band about 5%, and normal exposure the rest of the image. This would somehow seem to be a shutter related exposure problem. First curtain opens from right to left, hesitating at the point of the left edge of the lighter band. Second curtain then comes across early and hesitates at the right side of the light band.

Now, I do not think this is what is happening, but it offers another idea as to how to view the image and possibly think about what may actually be happening. Maybe try again with different times, different apertures... I like to think outside the box when attempting to solve problems as sometimes this is where the answers come from.

What happens if you point it up so you can only see sky then turn yourself in 90 degree increments and take four pictures - at the same inclination but therefore with the sun in a different position? Does the disruption stay the same or move?

It looks like it could be a kind of flare due to the stacked filters to me.

Hi,

I've been thinking about this kind of image for a while now and tried shooting it today, but I got a weird result and I was wondering if anyone more experienced could shed some light on what is happening...

The scenario is this:

- Daytime long exposure - 2 min @ f/22, ISO 200
- Filter stack - 24-70mm f/2.8 -> UV filter (too lazy to remove) -> 10 stop ND filter - Thin CPOL
- All filters are B+W and have no visible defects
- CPOL is set to max, as far as I know. I did a test a while back and aimed the camera at my computer monitor and adjusted it until I couldn't see the screen. I then marked the rotation on the CPOL with a permanent mark. I line up the marks, I should, technically have the 'maximum effect' of the CPOL.
- Composed and took the shot to see what effect I got and whether it matched what I thought I would get.

So the 'problem' is the vertical band of light about 1/4 in from the right side of the image, where there is a glow and clear delineation of the CPOL effect (grass is more less reflective, sky is more blue, etc.)

Did I hit the CPOL and knock it out of alignment (I don't think I did, but I didn't check thoroughly at the time because I didn't see the problem on the camera view screen)? Should a CPOL not be stacked in this fashion? Should a CPOL not be used for long exposures? I've not run across anything saying one way or the other.

Any insights?

The pic: