Chromatic aberration
Jan 14, 2018 22:24:34 #
rehess wrote:
I encounter fringing the most with long old lenses, like AdaptAll and Takumar lenses.
I still have a Pentax Spotmatic, sold all My Other film cameras, Minolta 7000 and 9000, when I entered the digital realm, but have not shot film since early to Mid 90’s. All lens on my digital stuff is pretty much current glass.
Jan 15, 2018 11:48:32 #
The fringing is a defect that some lenses have (for reasons mentioned above) If you process in Lightroom, it has a correction for that.
Jan 15, 2018 13:19:06 #
NP Hound wrote:
New to posting, but I’ve been reading on here for awhile. I have a weird situation, and I am at a loss. My wife has a Canon SL2 with a Tamron 16-300, and all of the sudden, she is getting green and red edges. She switched to a Canon 18-135, and all is well. I even let her use my Tamron 16-300, and the problem returned. She loaded her photos on the computer, opened them with Corel Paintshop, and the green and red edges were gone. Any thoughts or advice would be greatly appreciated.
Here's what DPReview said about your lens, as it relates to CA: "A generous helping of chromatic aberration is to be expected with a superzoom. There's strong magenta/green towards the edges at wide-angle, reducing to low levels through 50-100mm, before returning with a vengeance from 200-300mm. Canon DSLRs won't correct for it in JPEGs (Nikon DSLRs will try to) but it can be neutralized very effectively in your average Raw conversion software, provided you don't mind putting in a little time on a computer."
Jan 15, 2018 15:42:32 #
I don't have the lens and am generally inclined to use Canon lenses on Canon gear (and really didn't like non-Nikon lenses when I had Nikon gear). However, is it possible that you have some lens correction checked in the menus on the sl2? If so, turn it all off. I have the camera, but I am not near it, and can't remember whether or not it has it. I do know that on better Canons (5 d mk4, 5dsr), you need to turn off all lens correction to get 3rd party lenses to work right with them. It's possible that this would be why the lens will work on your 80D and not her sl2--you have the cameras set up differently. I am sorry that I can't be more specific, but I generally turn all the lens correction off in my cameras because I only shoot in raw, so I can't remember which cameras have what.
Jan 15, 2018 15:54:23 #
amfoto1
Loc: San Jose, Calif. USA
Actually most lenses exhibit some chromatic aberration at certain settings and under the most extreme conditions.
CA occurs when a portion of the color spectrum focuses differently from the rest of the spectrum.
There are two types of CA... axial (longitudinal) and transverse (lateral). Axial occurs when colors are focused at different distances behind a lens, not all exactly on the film plane, front to back. Transverse occurs when the colors all focused on the same plane, but some are shifted to the side, making for sort of a ghosting effect (similar to a "3D image").
Both can occur at the same time in an image, but generally lens designers work harder to reduce axial because it's more difficult to correct later with digital post-processing. Both kinds also can be present in all types of lenses, but are addressed differently depending upon the lens design. For example, the highest quality wide angle lenses use aspherical and acrhomatic elements to correct for CA, while high end telephotos might use low dispersion, ultra low dispersion elements or fluorite. More affordable lenses don't have these types of elements or don't use as many of them, so are more prone to issues.
Zoom lenses are more difficult to correct due to their much greater optical complexity and the wider range the zoom, the more difficult it is to correct. For example, a 16-300mm is an "18.75X" zoom (300 divided by 16), while an 18-135mm is only "7.5X". Most of the "best" and most pro-quality zooms tend to be 2X, 3X and maybe 4X at most (10-22mm, 17-55mm, 16-35mm, 24-70mm, 70-200mm, 100-400mm, 200-400mm). A few manage to give very high image quality in spite of a wider range (15-85mm, 28-300mm L-series for example), but they tend to have very complex optical formulas to achieve that).
To counteract CA and optimize other image quality factors, Canon has used fluorite in many of their "better" telephotos for decades. In order to do so they pioneered growing their own fluorite in, as well as making significant advancements working fluorite into the necessary shapes for use as lens elements, because it's rare to find naturally occurring and pretty fragile to work with. Right now many Canon lenses longer than 200mm use fluorite. Even 3 out of 4 of their 70-200mm do, too. Canon's commitment to using fluorite may have worked against them in at least one case... NASA bumped them from consideration because of concerns about fluorite elements failing under the rigors, vibration and G-forces of a rocket launch. They chose to use Nikon and Hasselblad instead, for the large part, neither of which were using much fluorite. Nikon has just recently revamped their 70-200mm f/2.8 and nearly all their modern primes longer than 300mm to use fluorite, too (look for the "FL" designation). Most other manufacturers are not using it presently, though most have at least dabbled a bit, using it occasionally in the past. Sigma does currently make telephotos using what they call "FLD" elements that they describe as "fluorite like", whatever that means.
The difference you are seeing with the lens on a fairly entry-level SL2 versus an 80D might be because the latter has built-in "lens profiles" that automatically correct for CA, including one for that specific lens. It is possible that there's a newer firmware version for the SL2 that updates it with additional profiles that would address the problem. (Though Canon is not really obliged to load profiles in their cameras for other manufacturer's lenses).
You might also be seeing this happening with the Corel software you mentioned. Some image editing programs also have built-in lens profiles to automatically correct CA and other things (distortion, vignetting, etc.) If you view the image in other software that lacks the profile, you see the CA... but when you view or work with it in the more advanced program, you don't see it because it's being corrected. A third party software such as Corel or Adobe products, etc., are more likely to have CA corrections (etc.) for a third party lens like the Tamron... than Canon's own software would.
It is possible to manually correct CA in some software. There may be "sliders" used to dial it out (Lightroom). Or, a slower manual process, can be corrected using a color replacement tool such as Photoshop provides.
CA also effects image sharpness, causing edges to appear more "blurred". In fact, it often occurs more strongly in slightly out of focus areas, too. Because it reduces apparent sharpness, CA will even have some effect on black & white images (whether done in-camera or a conversion in later post-processing).
CA occurs when a portion of the color spectrum focuses differently from the rest of the spectrum.
There are two types of CA... axial (longitudinal) and transverse (lateral). Axial occurs when colors are focused at different distances behind a lens, not all exactly on the film plane, front to back. Transverse occurs when the colors all focused on the same plane, but some are shifted to the side, making for sort of a ghosting effect (similar to a "3D image").
Both can occur at the same time in an image, but generally lens designers work harder to reduce axial because it's more difficult to correct later with digital post-processing. Both kinds also can be present in all types of lenses, but are addressed differently depending upon the lens design. For example, the highest quality wide angle lenses use aspherical and acrhomatic elements to correct for CA, while high end telephotos might use low dispersion, ultra low dispersion elements or fluorite. More affordable lenses don't have these types of elements or don't use as many of them, so are more prone to issues.
Zoom lenses are more difficult to correct due to their much greater optical complexity and the wider range the zoom, the more difficult it is to correct. For example, a 16-300mm is an "18.75X" zoom (300 divided by 16), while an 18-135mm is only "7.5X". Most of the "best" and most pro-quality zooms tend to be 2X, 3X and maybe 4X at most (10-22mm, 17-55mm, 16-35mm, 24-70mm, 70-200mm, 100-400mm, 200-400mm). A few manage to give very high image quality in spite of a wider range (15-85mm, 28-300mm L-series for example), but they tend to have very complex optical formulas to achieve that).
To counteract CA and optimize other image quality factors, Canon has used fluorite in many of their "better" telephotos for decades. In order to do so they pioneered growing their own fluorite in, as well as making significant advancements working fluorite into the necessary shapes for use as lens elements, because it's rare to find naturally occurring and pretty fragile to work with. Right now many Canon lenses longer than 200mm use fluorite. Even 3 out of 4 of their 70-200mm do, too. Canon's commitment to using fluorite may have worked against them in at least one case... NASA bumped them from consideration because of concerns about fluorite elements failing under the rigors, vibration and G-forces of a rocket launch. They chose to use Nikon and Hasselblad instead, for the large part, neither of which were using much fluorite. Nikon has just recently revamped their 70-200mm f/2.8 and nearly all their modern primes longer than 300mm to use fluorite, too (look for the "FL" designation). Most other manufacturers are not using it presently, though most have at least dabbled a bit, using it occasionally in the past. Sigma does currently make telephotos using what they call "FLD" elements that they describe as "fluorite like", whatever that means.
The difference you are seeing with the lens on a fairly entry-level SL2 versus an 80D might be because the latter has built-in "lens profiles" that automatically correct for CA, including one for that specific lens. It is possible that there's a newer firmware version for the SL2 that updates it with additional profiles that would address the problem. (Though Canon is not really obliged to load profiles in their cameras for other manufacturer's lenses).
You might also be seeing this happening with the Corel software you mentioned. Some image editing programs also have built-in lens profiles to automatically correct CA and other things (distortion, vignetting, etc.) If you view the image in other software that lacks the profile, you see the CA... but when you view or work with it in the more advanced program, you don't see it because it's being corrected. A third party software such as Corel or Adobe products, etc., are more likely to have CA corrections (etc.) for a third party lens like the Tamron... than Canon's own software would.
It is possible to manually correct CA in some software. There may be "sliders" used to dial it out (Lightroom). Or, a slower manual process, can be corrected using a color replacement tool such as Photoshop provides.
CA also effects image sharpness, causing edges to appear more "blurred". In fact, it often occurs more strongly in slightly out of focus areas, too. Because it reduces apparent sharpness, CA will even have some effect on black & white images (whether done in-camera or a conversion in later post-processing).
Jan 15, 2018 16:07:15 #
Thanks for this explanation of lens construction. I learned a lot from it that I didn't know.
If you want to reply, then register here. Registration is free and your account is created instantly, so you can post right away.