CheshireCat Posted March 30, 2015 Share #221 Posted March 30, 2015 Advertisement (gone after registration) Here an example of the transmission of an exotic lens.[...] Even this lens is not flat in the visual part. Transmission is not chromatic aberration. You can find CA graphs for that lens are here: http://diglloyd.com/articles/CoastalOptics60f4/images-brochure2/Chromatic_and_Distortion.png@auto.html The LoCA graph clearly shows the 3 zeros of the APO correction. Link to post Share on other sites More sharing options...
Advertisement Posted March 30, 2015 Posted March 30, 2015 Hi CheshireCat, Take a look here Is this CA problem on my 50/1.4 ASPH, it's normal or not?. I'm sure you'll find what you were looking for!
BerndReini Posted March 30, 2015 Share #222 Posted March 30, 2015 Are you sure of that? Yes. Welcome, dear visitor! As registered member you'd see an image here… Simply register for free here – We are always happy to welcome new members! Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! ' data-webShareUrl='https://www.l-camera-forum.com/topic/242349-is-this-ca-problem-on-my-5014-asph-its-normal-or-not/?do=findComment&comment=2789874'>More sharing options...
Rick Posted March 30, 2015 Share #223 Posted March 30, 2015 Hi all. Interesting thread. But, the photos in this thread represent chromatic aberration. But, there is a little more to it. In simple thin lens optics the chromatic aberrations are always in order of the spectrum of light. Red at one edge and violet at the other. But, the images we see in this thread are purple (not a pure color) and green. Doesn't that bother anyone? Once we move to doublets and achromatic lenses the goal of the lens designer is to try and bring two colors separated widely in the spectrum to the same plane of focus (yes at the point of focus). These two colors turn out to be blue (F) and red ©. Both of these colors represent both sides of the visible spectrum and are chosen by optical designers as the two colors to correct in achromatic lens design. Now, we have two widely separated colors in common focus (achromatic) but, all colors of the spectrum are not focused. Those in-between (green) are focused closer and those at the ends of the spectrum (near infrared and near ultra-violet) will focus further. This defocus is called the secondary spectrum. Correcting the secondary spectrum brings the green to the focus plane of the blue and red. When this is achieved the lens is apochromatic assuming that spherical chromatic aberration (spherochromatism) and coma are well corrected. This according to Abbe. But, none of this is perfect. There are many levels of apochromatic correction. All lenses will display some amount of secondary spectrum in front of and behind the plane of focus. That difference in the plane of focus between red and blue vs green, which will focus to a shorter distance. When we focus beyond the trees the blue and red are focused to the outside of the branch and we see the addition of these two colors as purple. When we focus in front of the trees green is focused to the outside of the branch. This is the effect of the secondary spectrum. It is a type of longitudinal chromatic aberration. The effect is most pronounced when focus is behind the trees because the red and blue combined, by design, are focused together and there is more electromagnetic energy than with the much narrower portion of the green spectrum effect we see when focused in front of the tree . So, the effect is more noticeable when when we see the purple fringing than the green type. Also, the blue and red are far apart and refract to different degrees to make a wider fringe. This has little to do with the sensor, but it is more apparent when shot next to bright backgrounds that saturate the entire field or part of it to the right or left of the branch. So far, this is all due to LoCA and I don't see proof, other than internet folklore, that would explain it. As a note: It is important to realize that the fringing is magenta. Blue and red are combined to form strong magenta with blue on the inside and red on the outside of the object (branch)... see the picture taken against the window. Also, it is important to realize that these photos show objects that are not in the plane of focus, so bright backlight can occupy the same space as the edge of the brances which are spread out and non distinct. Rick Link to post Share on other sites More sharing options...
CheshireCat Posted March 30, 2015 Share #224 Posted March 30, 2015 If the purple is coming out of the sunlight and forming a fringe, then there should be somewhere in the image a purple deficient area. I haven't seen one. Or doesn't light work like arithmetic? The "complementary fringe" is easy to see in case of Lateral CA. Each color forms a slightly different size image in the plane of focus, hence the "shift" is noticeable especially at the image periphery. Longitudinal CA is different. With a perfect lens, a white point out of focus would form a white disk (called Circle of Confusion) on the sensor. With a real lens, a white point in front of focus will form the same white disk, but with purple fringing outside due to purple frequencies forming a bigger disk (i.e. purple is more out of focus than the other frequencies). Where is the complementary green disk then ? You cannot see it, as it is smaller, and inside the purple disk. Exactly like arithmetic, the big purple disk and the smaller green disk add their light together to form the white part of the disk. The white disk is very slightly "purple deficient", but you won't be able to notice it in most cases. Some people think the sun is not green, but they are wrong. The sun is also green, and blue, and red, and violet, and UV, and IR, and ... Link to post Share on other sites More sharing options...
jaapv Posted March 30, 2015 Share #225 Posted March 30, 2015 Yep, and as RGB uses colour values to produce luminosity values you will never see the counterfringe, unless you use a colourspace that separates luminosity from colour, like CMYK or LAB. In that case it can easily be shown. Link to post Share on other sites More sharing options...
Rick Posted March 30, 2015 Share #226 Posted March 30, 2015 I don't believe the magenta is from the sunlight in LoCA off the plane of focus. It is created by the object (tree branch) itself. Rick Link to post Share on other sites More sharing options...
pgk Posted March 30, 2015 Share #227 Posted March 30, 2015 Advertisement (gone after registration) The problem with all this theorising is that it simply doesn't explain the complexity of variability of where purple fringing actually occurs in reality. From my experience it can occur with various lenses, both sophisticated modern designs and substantially poorer and older, far less sophisticated designs. And from my experience the results from various lenses are variable and there is little of the consistency that might be expected, and it occurs on a variety of sensors too. Adjustments for purple fringing in Photoshop show no consistency either and I find that they have to be made on an image by image basis which again is not what might be expected from a simple, single physical cause. Its as if (and I expect it is) there are a (potentially quite large) number of variables at work, sometimes some coinciding and adding together, and at other times some counteracting each other. Its this lack of consistency which makes me think that the problem isn't down to a single, simply explained cause. If it was then I would have expected lens/camera manufacturers to have dealt with it in-camera using software - which makes sense. If on the other hand it is due to a number of variables the additive and counteractive effects of which are too complex to mathematically predict, then this explains why it hasn't been properly analysed (or at least any analysis made by those who have an absolute interest in doing so hasn't been broadcast) and counteracted. The problem with all the examples posted is that they too are from a wide variety of different sensors, lenses, focus distances, contrast conditions, exposures, focus points, etc. You get the idea. Even comparing two similar lenses will have variables such as angle of incidence of the light path to the sensor, absolute precision of focus, absolute aperture and so on. Whatever the causes are I am convinced that there are more variables at play than simply CA...... Link to post Share on other sites More sharing options...
jankap Posted March 30, 2015 Share #228 Posted March 30, 2015 Some people think the sun is not green, but they are wrong. The sun is also green, and blue, and red, and violet, and UV, and IR, and ... The spectrum of the sun is continuous. It is interesting to compare this with the spectrum of fluorescent lights: the cheaper ones. Jan Link to post Share on other sites More sharing options...
Rick Posted March 30, 2015 Share #229 Posted March 30, 2015 The problem with all this theorising is that it simply doesn't explain the complexity of variability of where purple fringing actually occurs in reality. From my experience it can occur with various lenses, both sophisticated modern designs and substantially poorer and older, far less sophisticated designs. And from my experience the results from various lenses are variable and there is little of the consistency that might be expected, and it occurs on a variety of sensors too. Yes, it is expected to be variable with different lenses as they all offer varying degrees of chromatic correction. Sensors are different as well. Adjustments for purple fringing in Photoshop show no consistency either and I find that they have to be made on an image by image basis which again is not what might be expected from a simple, single physical cause. Its as if (and I expect it is) there are a (potentially quite large) number of variables at work, sometimes some coinciding and adding together, and at other times some counteracting each other. Its this lack of consistency which makes me think that the problem isn't down to a single, simply explained cause. If it was then I would have expected lens/camera manufacturers to have dealt with it in-camera using software - which makes sense.. Adjustment in photoshop is always going to be different because lighting changes and the amount of red and blue light will combine to produce varying magenta color. That is why there exists a purple slider to correct for the different color produced. Lack of consistency does not make the effect something that we can't understand. Only, variable in its presentation. If on the other hand it is due to a number of variables the additive and counteractive effects of which are too complex to mathematically predict, then this explains why it hasn't been properly analysed (or at least any analysis made by those who have an absolute interest in doing so hasn't been broadcast) and counteracted. The problem with all the examples posted is that they too are from a wide variety of different sensors, lenses, focus distances, contrast conditions, exposures, focus points, etc. You get the idea. Even comparing two similar lenses will have variables such as angle of incidence of the light path to the sensor, absolute precision of focus, absolute aperture and so on. Whatever the causes are I am convinced that there are more variables at play than simply CA...... It has been studied and it is just the effect of the secondary spectrum produced by compound lens designs utilized to correct for LoCA at the plane of focus. Simply, focus behind an object and you will create a magenta rainbow and it is seen in every one of these examples. It can not be the sensor in anyway other than its ability to take a very narrow plane of rays unlike film that is wider (longitudinally). Rick Link to post Share on other sites More sharing options...
pgk Posted March 30, 2015 Share #230 Posted March 30, 2015 Yes, it is expected to be variable with different lenses as they all offer varying degrees of chromatic correction. But Rick, there is no correlation between 'good, modern' and 'poor, older' designs, both can produce substantial purple fringing. This is the bit that fails to make sense to me. In fact I find that better corrected (even 'apo') lenses with better definition of fine detail (ie modern, high resolution lenses) are worse than older, less well corrected designs (an 'edge' effect?). If you are correct, why is this? Link to post Share on other sites More sharing options...
Rick Posted March 30, 2015 Share #231 Posted March 30, 2015 But Rick, there is no correlation between 'good, modern' and 'poor, older' designs, both can produce substantial purple fringing. This is the bit that fails to make sense to me. In fact I find that better corrected (even 'apo') lenses with better definition of fine detail (ie modern, high resolution lenses) are worse than older, less well corrected designs (an 'edge' effect?). If you are correct, why is this? APO only applies to the plane of focus (POF). In other words, your APO can create a point source that is very small (resolving power) and has very little overlap of colors (high contrast) at the plane of focus. Everything in front of and behind what you are focuses on is a blur to some degree. That blur splits up into different colors depending on how much the designer had to bend the different colors of light (red, blue and green) to get them coincident at the POF. Highly Apochromatic lenses may trade off good correction at the POF for more divergent chromatic rays away from the POF. APO only refers to the POF. Don't disparage, it is a trade off. Also, this is one of the reasons people like the OOF (bokeh) of some of the less corrected older lenses that have more spherical chromatic aberration... Mandler designs. Rick Link to post Share on other sites More sharing options...
CheshireCat Posted March 30, 2015 Share #232 Posted March 30, 2015 The problem with all this theorising is that it simply doesn't explain the complexity of variability of where purple fringing actually occurs in reality. Complex art, complex issues. Adjustments for purple fringing in Photoshop show no consistency either and I find that they have to be made on an image by image basis which again is not what might be expected from a simple, single physical cause. Unlike Lateral CA, Longitudinal CA is tridimensional in nature. This is why it is relatively simple to correct the former and not the latter. Even if we have detailed lens profiles, depth information has been lost forever in the photographic process. In the future, we will take 3D photos and there will be more advanced technology to fix this and do other cool things. Whatever the causes are I am convinced that there are more variables at play than simply CA...... You don't know much about CA, so your convinctions are not based on anything solid. Besides, if you are "convinced" about other possible causes, just let us know. Would you pretty please ? It is the 4th time I am asking you to detail your alternative theory, therefore at this point I am [arrogantly] convinced you just love teasing. Link to post Share on other sites More sharing options...
pgk Posted March 30, 2015 Share #233 Posted March 30, 2015 Don't disparage, it is a trade off. Also, this is one of the reasons people like the OOF (bokeh) of some of the less corrected older lenses that have more spherical chromatic aberration... Mandler designs. Rick Some Mandler designs display substantial purple fringing, others don't. Similarly some aspheric designs, both Leica and others, do whilst others don't. I shoot a lot of high contrast scenes so its a problem that I am well aware of. Some I can fully correct others I can't. I am still not convinced that it is as simple as you suggest - I'll bounce the idea of it simply being CA off a friend who is a lens designer next time I see him and see what he says. As for an alternative theory, well mine is that the issue is more complex and involves the interaction of a variety of causes, some of which I outlined as potential causes in another post. Again I will suggest that if it has a 'simple' solution such as being CA, I would have expected software correction to be applied in-camera by manufacturers whose cameras and lenses 'talk' comprehensively to each other; wouldn't this make sense and be in manufacturer's interests? I will add that it seems extremely strange to me that the problem is son complex that it cannot be dealt with in camera but a simple slider in Photoshop can often negate it manually. If a lens manufacturer knows the characteristics of a lens why can this not be achieved in-camera if the cause is one problem? Link to post Share on other sites More sharing options...
Exodies Posted March 30, 2015 Share #234 Posted March 30, 2015 One way it could simply be too complex to correct in camera is if it depended on the distance of the source to the focal plane. This distance is not recorded. Link to post Share on other sites More sharing options...
Rick Posted March 30, 2015 Share #235 Posted March 30, 2015 Some Mandler designs display substantial purple fringing, others don't. Similarly some aspheric designs, both Leica and others, do whilst others don't. I shoot a lot of high contrast scenes so its a problem that I am well aware of. Some I can fully correct others I can't. I am still not convinced that it is as simple as you suggest - I'll bounce the idea of it simply being CA off a friend who is a lens designer next time I see him and see what he says. As for an alternative theory, well mine is that the issue is more complex and involves the interaction of a variety of causes, some of which I outlined as potential causes in another post. Again I will suggest that if it has a 'simple' solution such as being CA, I would have expected software correction to be applied in-camera by manufacturers whose cameras and lenses 'talk' comprehensively to each other; wouldn't this make sense and be in manufacturer's interests? I will add that it seems extremely strange to me that the problem is son complex that it cannot be dealt with in camera but a simple slider in Photoshop can often negate it manually. If a lens manufacturer knows the characteristics of a lens why can this not be achieved in-camera if the cause is one problem? Sorry, fringes posted here are simply caused by LoCA corrections made in the lens design but not corrected at a point that is off the POF. Rick Link to post Share on other sites More sharing options...
adan Posted March 31, 2015 Share #236 Posted March 31, 2015 If you have an alternative, more precise term for CA outside of the "focus plane", just let me know and I will comply. How about OOFCA (Out-of-focus CA)? Also a tip of the hat to the old Leica catalog deisgnations (BEOOY, FOOXB, SLOOZ, VIDOM etc.) Link to post Share on other sites More sharing options...
Rick Posted March 31, 2015 Share #237 Posted March 31, 2015 How about OOFCA (Out-of-focus CA)? Also a tip of the hat to the old Leica catalog deisgnations (BEOOY, FOOXB, SLOOZ, VIDOM etc.) OOFCA: I've read that term before in 2012 when Adobe added their defringing tool to PS 4 and ACR7. OOF CA is not a classical aberration because it is always zero in the focal plane. It is more like a defocus blur aberration. It has also been called Bokeh-CA. But, I do like your persistent point that it is technically not LoCA. Rick Link to post Share on other sites More sharing options...
pgk Posted March 31, 2015 Share #238 Posted March 31, 2015 Sorry, fringes posted here are simply caused by LoCA corrections made in the lens design but not corrected at a point that is off the POF. Rick Sorry Rick but having looked through a few of my photos which exhibit fringing I am still of the opinion that whilst there is some CA and even some LoCA in some there is way too much inconsistency (even in a series of images of the same subject at the same settings) to attribute it solely to this 'single' issue. We'll simply have to agree to disagree. And FWIW Adobe seem to refer to fringing as 'chromatic artifacts' which is a somewhat puzzling, catch all term if ever there was one! Link to post Share on other sites More sharing options...
Rick Posted March 31, 2015 Share #239 Posted March 31, 2015 Sorry Rick but having looked through a few of my photos which exhibit fringing I am still of the opinion that whilst there is some CA and even some LoCA in some there is way too much inconsistency (even in a series of images of the same subject at the same settings) to attribute it solely to this 'single' issue. We'll simply have to agree to disagree. And FWIW Adobe seem to refer to fringing as 'chromatic artifacts' which is a somewhat puzzling, catch all term if ever there was one! If, you read Adobe from 2012 they report 3 types of chromatic artifacts. Lateral CA. LoCA. And, Sensor Bloom. Adobe lumps them all together as Chromatic Artifacts (color problems that need a slider). Adobe' slider corrections are for all three. Although, there are examples posted here representing all three, the one we are talking about from the OP is simply OOF-CA. New Color Fringe Correction Controls Rick Link to post Share on other sites More sharing options...
pgk Posted March 31, 2015 Share #240 Posted March 31, 2015 Point taken on OP. My problem is that when I look at some images I find 'fringing' variability across the frame, in intensity, spread and so on. As far as I can see, sometimes its clearly a mix of problems. What I still can't get my head around is the incoherent variability which seems to have no correlation with any common factor. I can correct most and often all of it, but the settings are varied too. And sometimes its simply not correctable. I wonder if on occasion there are other localised impacts (such as reflections within micro-lenses or the sensor cover glass due to very high light intensity in some areas) which add in to cause yet more variability? Link to post Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.