CA made worse by CPL?

[Davejl]

I have the Canon 17-85mm lens which has notoriously bad CA at the wide end. A while back Canon added "Digital Lens Optimizer" to DPP, which I have found to be pretty good at improving this. However, today I took some shots using a CPL (a supposedly good quality one) and this shot seems to have pretty bad fringing on the tree branches against the sky despite using the DLO. Is it likely to be due to using the filter?


Canon 70D; Canon 17-85mm @ 22mm

Dave

[Shadowman]

Hard to visualize without being there but are you positioning the camera at the right angle in relation to the position of the sun?

[Davejl]

Hi John,

I think so. The sun was pretty low in the sky and about to my right and I think I aligned the arrow on the CPL with it.

Dave

[Manfred M]

The polarizer should not have any affect on the CA. CA is dependent on the lens design and is related to the different wavelengths of light not falling on exactly the same spot on your camera's sensor.

BUT what a polarizer does is that it reduces reflections off non-metallic objects. This reduces glare so the CA that is already there is sharper and hence more noticeable.

[Mike Buckley]

The sun was pretty low in the sky and about to my right and I think I aligned the arrow on the CPL with it.

The length of shadows indicates that the sun was indeed fairly low. The direction of the shadows indicates that the lens was at about a 60-degree angle to the sun. With that combination of factors, I'm confident the polarizer had the potential to dramatically effect the color of the plain blue sky and to eliminate or minimize glare in the scene. Keep in mind, though, that it's very possible to affect the sky too intensely from purely an artistic viewpoint. My point is that you would be better off positioning the polarizer's ring according to the image's look you want, not according to the relationship between the marking on the ring and the position of the sun.

[Shadowman]

Hi John,

I think so. The sun was pretty low in the sky and about to my right and I think I aligned the arrow on the CPL with it.

Dave

Dave,

I wonder with the lens you used if just a tad bit more light hit the lens? What was your aperture? 22mm is just short of wide for a crop sensor but enough to increase CA effects in your image, allowing more light to reach the sensor will also create the look. I those aspects, wide angle, position of camera in relation to the sun, and the added effect of a cpl might be the cause of CA in your image.

[Manfred M]

+1 to what Mike has written. The 90 degree "rule" only impacts darkening the sky, where the amount of darkening is maximum when the camera with polarizer is pointing in a direction that is perpendicular to the position of the sun and minimal when the sun is directly in front of or behind the shooter.

Reducing reflections (leaves tend to be excellent little reflectors) is something that needs to be done visually without regard to the position of the sun or any markings on the filter.

[Mike Buckley]

Reducing reflections (leaves tend to be excellent little reflectors) is something that needs to be done visually without regard to the position of the sun or any markings on the filter.

When viewing a scene through the small viewfinder, I sometimes have a difficult time determining which position eliminates the glare on the leaves, grass, etc. When that happens, I raise the camera to temporarily include the blue part of the sky but being sure not to move the lens to the right or left. I then adjust the ring on the polarizer to render maximum effect on the sky. Having done that, I also know that position will reduce the most amount of glare once I have lowered the camera to capture the scene.

[Manfred M]

, I raise the camera to temporarily include the blue part of the sky but being sure not to move the lens to the right or left.

That trick certainly won't work on a cloudy day (and I shoot landscapes with a polarizer on cloudy days too) and won't necessarily give you the optimal results on a sunny one either as the direction that the leaves are reflecting the light does not necessarily match the position of the sun.

I know what you mean about it being difficult to figure this out, but found (like many other things in photography), the more I shot this way, the more my eyes were attuned to what is happening in the scene.

[Davejl]

The polarizer should not have any affect on the CA. CA is dependent on the lens design and is related to the different wavelengths of light not falling on exactly the same spot on your camera's sensor.

BUT what a polarizer does is that it reduces reflections off non-metallic objects. This reduces glare so the CA that is already there is sharper and hence more noticeable.

Thanks Manfred.

The reason I thought the CPL might be the culprit was because the shots I took with similar settings but without a CPL (in this thread: Autumn colours) seem to be free of CA.

I must say that I hadn't appreciated that the CPL can be used in the way you suggest, and will experiment more.

Dave

[Manfred M]

Thanks Manfred.

The reason I thought the CPL might be the culprit was because the shots I took with similar settings but without a CPL (in this thread: Autumn colours) seem to be free of CA.

I must say that I hadn't appreciated that the CPL can be used in the way you suggest, and will experiment more.

Dave

Your most recent shot was taken at f/6.3 while the older shots were taken at f/11 (all three of them). CA impact can be reduced by stopping down the lens as depth of field will increase. As the depth of field increases, the CA should come down. I suspect that this might be the main difference in the images.

[Davejl]

Dave,

I wonder with the lens you used if just a tad bit more light hit the lens? What was your aperture? 22mm is just short of wide for a crop sensor but enough to increase CA effects in your image, allowing more light to reach the sensor will also create the look. I those aspects, wide angle, position of camera in relation to the sun, and the added effect of a cpl might be the cause of CA in your image.

I can't access the EXIF data at present, but I know that the aperture I used was certainly larger than I used on some recent shots that didn't seem to suffer from CA - probably f8 as opposed to f11, because with a fairly stiff wind blowing today I didn't want to have too slow a shutter speed.

Dave

[Shadowman]

I can't access the EXIF data at present, but I know that the aperture I used was certainly larger than I used on some recent shots that didn't seem to suffer from CA - probably f8 as opposed to f11, because with a fairly stiff wind blowing today I didn't want to have too slow a shutter speed.

Dave

f/6.3 for this particular image, not wide open but enough to let in some stray light at the angle you positioned yourself to the sun.

[proseak]

Try shooting test subjects with and without the CPL.

[Davejl]

Thanks Manfred and John.

I hadn't appreciated that CA is dependant on focal length AND aperture, that is very useful to know. I should clearly have used a higher ISO and smaller aperture. I had used f6.3 for a close-up leaf shot and obviously left the setting on that. I must get into the habit of taking my glasses out with me: I don't need them except for reading (e.g. camera controls). Ideally I could do with a magnifier on a neck cord.

It seems therefore that the CA correction feature in DPP is good within limits.

Dave

[Davejl]

Try shooting test subjects with and without the CPL.

Thanks Peter,

That makes sense, though it seems that choosing too wide an aperture was the problem here. I wouldn't normally have used f6.3 for this sort of scene. It is also possible that I left the lens hood off for this one. Normally I use it, but to turn the filter I did remove it at times.

Dave

[Manfred M]

Thanks Manfred and John.

I hadn't appreciated that CA is dependant on focal length AND aperture, that is very useful to know. I should clearly have used a higher ISO and smaller aperture. I had used f6.3 for a close-up leaf shot and obviously left the setting on that. I must get into the habit of taking my glasses out with me: I don't need them except for reading (e.g. camera controls). Ideally I could do with a magnifier on a neck cord.

It seems therefore that the CA correction feature in DPP is good within limits.

Dave

I don't know how well DPP works as I am not a Canon shooter. I do know that the various Adobe products do a decent job of automatically reducing CA, as do some of the other raw converters that I use. No software is perfect and certainly the Adobe products have a manual feature that lets me do more aggressive CA correction than is possible through the purely automated routines. It would be worthwhile checking DPP's capabilities.

[GrahamS]

Dave, if you are using the AF-S 17-85mm f2.8 USM, in my experience, nothing completely eliminates the CA, which is bad even around the f8 to f11 range. I had one and tried the in-camera jpg correction, Photoshop CS, Lightroom, DXO. With RAW capture I tried Adobe Camera Raw, PS CS, LR5 (all the same engine) and the Canon proprietary software. I could never get a clean image. Bear in mind that there will be a variation of the severity of the problem from lens example to example, but eventually I sold the lens and called it a day. The CPL filter will have no effect on the severity of the CA other than an apparent reduction where the filter reduces the presence of spectral highlights and primary reflections that may enhance the CA.