pgk Posted April 9, 2015 Share #281 Posted April 9, 2015 Advertisement (gone after registration) Can we all accept that purple fringing is actually a form of variegation and the cause is in fact most likely to be viral? Link to post Share on other sites More sharing options...
Advertisement Posted April 9, 2015 Posted April 9, 2015 Hi pgk, 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!
lct Posted April 9, 2015 Share #282 Posted April 9, 2015 Must be the well known « Internetus Psitaccus Virus » i guess . Link to post Share on other sites More sharing options...
Peter Branch Posted April 14, 2015 Share #283 Posted April 14, 2015 The other day I met up with someone who was part of the group responsible for developing films like Portra 160 so I asked his opinion about the “Purple Edge” phenomenon. His reply can be summarised as on the one hand it is easy to explain, on the other hand it is difficult to predict when and where it will occur. Essentially it is a function of which of the three primary colours is most sharply rendered. If the blue image is the “sharpest” then there will be a “halo” of red and green which looks yellow to the human eye. This he suggested is the least desirable outcome as photographers react badly to yellow fringing. With films where the blue sensitive layer is on the top and the blue image tends to be the sharpest this is a problem. The lens used obviously plays a part depending on the degree to which it is “in focus” and the degree to which Chromatic Aberration, all 4 types, are corrected. With digital it is somewhat different but the same principle applies. Camera manufacturers closely guard the secrets of the algorithms they use to write RAW files from the sensor output. But it seems that they attempt to maximise the “Sharpness” of the green image. This makes sense both because the human eye is most sensitive to the green part of the spectrum and also because the Bayer Filter pattern has twice as many green as red & blue sensors. It is this which distinguishes digital sensors from film in that it is much more predictable that the sharpest image will be green giving a blue and red “Halo”, seen as purple. However again the lens and the degree of focus can lead to anomalous results which the camera and post processing software treat is complex ways. Hence it is very difficult in practice to predict exactly what will happen in advance. To me this explains much of the existence and variation in the phenomenon being experienced as “Purple Edge”. It depends in practice on both the sensor, film or digital, and the lens. Link to post Share on other sites More sharing options...
CheshireCat Posted April 14, 2015 Share #284 Posted April 14, 2015 Essentially it is a function of which of the three primary colours is most sharply rendered. If the blue image is the “sharpest” then there will be a “halo” of red and green which looks yellow to the human eye. This he suggested is the least desirable outcome as photographers react badly to yellow fringing. In my opinion, yellow fringing would be much less noticeable than purple fringing. That said, achromatic lenses keep blue and red together, so it is really hard to make yellow fringing happen. My educates guess is that the real difference between digital and film may be due to different film color-sensitive layers being stacked (i.e. different depths, hence different focus). Could you please double check with the "Portra 160 guy" ? Camera manufacturers closely guard the secrets of the algorithms they use to write RAW files from the sensor output. But it seems that they attempt to maximise the “Sharpness” of the green image. Sorry, but I am afraid this is another myth. Can you cite the source ? If they really wanted to maximize the sharpness of a raw image, then they would do that on all three color components. Link to post Share on other sites More sharing options...
Peter Branch Posted April 14, 2015 Share #285 Posted April 14, 2015 In my opinion, yellow fringing would be much less noticeable than purple fringing........ You are, of course, entitled to your opinion but I prefer to pay regard to the opinion of a bunch of seasoned professionals who have been responsible for product lines each of which generates a multi-million dollar annual cash flow. These people are serious, very highly qualified technically, have access to vast facilities and spend a great deal of time analysing customer feedback. Link to post Share on other sites More sharing options...
Peter Branch Posted April 14, 2015 Share #286 Posted April 14, 2015 ..........That said, achromatic lenses keep blue and red together................... Which, of course, is precisely what even modern Apochromatic lenses do not do. Achromatic lenses never have been able to do as you claim. I suggest you look closely at the graphs that E. Puts has published, (I don't have his permission to reproduce them here). Link to post Share on other sites More sharing options...
Peter Branch Posted April 14, 2015 Share #287 Posted April 14, 2015 Advertisement (gone after registration) In my opinion, yellow fringing would be much less noticeable than purple fringing. That said, achromatic lenses keep blue and red together, so it is really hard to make yellow fringing happen. My educates guess is that the real difference between digital and film may be due to different film color-sensitive layers being stacked (i.e. different depths, hence different focus). Could you please double check with the "Portra 160 guy" ? Sorry, but I am afraid this is another myth. Can you cite the source ? If they really wanted to maximize the sharpness of a raw image, then they would do that on all three color components. How can I put this tactfully? .......oh!, why bother....... Link to post Share on other sites More sharing options...
CheshireCat Posted April 15, 2015 Share #288 Posted April 15, 2015 Which, of course, is precisely what even modern Apochromatic lenses do not do. Achromatic lenses never have been able to do as you claim. Check the last graph here (Erwin Puts): http://www.imx.nl/photo/optics/styled-5/ Here's a picture of the light spectrum for your convenience: http://en.m.wikipedia.org/wiki/Visible_spectrum#/media/File:Rendered_Spectrum.png You can easily see the zeros are in the blue and red zone. Link to post Share on other sites More sharing options...
Peter Branch Posted April 15, 2015 Share #289 Posted April 15, 2015 ...........You can easily see the zeros are in the blue and red zone. One last time - I'm certain other Forum members are getting very bored by this - the sensor, be it film or digital, is not just sensitive to the precise wavelength where the graph crosses the y-axis. Can I respectfully refer you to: "An Essay on Criticism" by Alexander Pope. May I wish you good fortune and many grandchildren - goodbye. Link to post Share on other sites More sharing options...
lct Posted April 15, 2015 Share #290 Posted April 15, 2015 [...] At high intensity light levels, the information becomes non-linear and depending on its intensity, blooming, whatever, will be increasingly spurious. Even when there is underlying CA there will be additional distortions and sometimes even when there is no CA there will still be distortions in the data. So, whilst you may be dealing with CA (or sometimes not) there are other elements causing problems in images. [...] The pic below seems to illustrate what you wrote above if i understood you well. Foveon sensor (Sigma DP2 Merril), 30/2.8, f/2.8, 200 iso. No "green behind focus, purple in front" fringes here, both are in front of the plane of focus (background trees) by far. I would have called this sensor blooming ten years ago and the CA theory can hardly convince me that i was mistaken so far. Full size: http://lctphot.smugmug.com/Other/samples/i-qdXXq6F/0/O/DP2M0112_sips.jpg 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=2798653'>More sharing options...
CheshireCat Posted April 17, 2015 Share #291 Posted April 17, 2015 One last time - I'm certain other Forum members are getting very bored by this - the sensor, be it film or digital, is not just sensitive to the precise wavelength where the graph crosses the y-axis. Actually, it is the x-axis, and the x-axis is where the sensor is. The graph represents the focus error, and has nothing to do with wavelength sensitivity. But thanks for your "last time". You and "certain other forum members" are free to ignore this thread, and avoid posting sarcastic nonsense. Link to post Share on other sites More sharing options...
CheshireCat Posted April 17, 2015 Share #292 Posted April 17, 2015 No "green behind focus, purple in front" fringes here, both are in front of the plane of focus (background trees) by far. I would have called this sensor blooming ten years ago and the CA theory can hardly convince me that i was mistaken so far. Full size: http://lctphot.smugmug.com/Other/samples/i-qdXXq6F/0/O/DP2M0112_sips.jpg It is lens flare. The different colors are caused by different coatings in lens elements. Check this interesting thread, for better examples clearly showing that it is an optical phenomenon: http://www.dpreview.com/forums/thread/3248008 Note that in your picture, examining areas not contaminated by the flare, you can see good old longitudinal and lateral CA. Link to post Share on other sites More sharing options...
lct Posted April 17, 2015 Share #293 Posted April 17, 2015 This lens has very little flare (pic) the same way as my 50/2 apo has very little CA. Both can show a lot of purple and otherwise colored fringing having little to do with CA. You give me the feeling to see the forest for the trees with respect. The CA theory you seem to endorse is at loss to explain why the best apochromatic lenses can cause the worst purple fringing, nor why purple fringing is almost invisible on film. 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=2799488'>More sharing options...
jaapv Posted April 17, 2015 Share #294 Posted April 17, 2015 It is lens flare. The different colors are caused by different coatings in lens elements. Check this interesting thread, for better examples clearly showing that it is an optical phenomenon: http://www.dpreview.com/forums/thread/3248008 Note that in your picture, examining areas not contaminated by the flare, you can see good old longitudinal and lateral CA. Highly coincidental that the "coating" colours appear exactly in the same shade and sequence as the alleged CA colours.... Link to post Share on other sites More sharing options...
CheshireCat Posted April 17, 2015 Share #295 Posted April 17, 2015 Highly coincidental that the "coating" colours appear exactly in the same shade and sequence as the alleged CA colours.... If you care to check the examples in the link I posted, you will see: the same colors in what is obviously a beautiful caleidoscopic lens flare (or whatever other optical phenomenon is happening inside the camera). Thinking that a digital sensor could create something like a caleidoscopic lens flare is a bit naive Link to post Share on other sites More sharing options...
CheshireCat Posted April 17, 2015 Share #296 Posted April 17, 2015 This lens has very little flare (pic) the same way as my 50/2 apo has very little CA. Both can show a lot of purple and otherwise colored fringing having little to do with CA. You give me the feeling to see the forest for the trees with respect. The CA theory you seem to endorse is at loss to explain why the best apochromatic lenses can cause the worst purple fringing, nor why purple fringing is almost invisible on film. Your DP2 picture of the lamp does not show flare and does not show purple fringing either (or what you incorrectly call sensor blooming). Make the flare happen and your "sensor blooming" will magically appear We have already discussed in this thread why the best APO lenses can cause purple fringing: - The APO correction is on the focus plane only. - Some "best APO lenses" are not actually "the best". E.g. the APO Summicron 90. Link to post Share on other sites More sharing options...
lct Posted April 17, 2015 Share #297 Posted April 17, 2015 OK 500w... Green above and purple below... Must be CA then, there's no doubt about it... 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=2800020'>More sharing options...
CheshireCat Posted April 18, 2015 Share #298 Posted April 18, 2015 OK 500w... Green above and purple below... Must be CA then, there's no doubt about it... DP2M0186_web.jpg It surely is an optical phenomenon. Call it lens flare, internal reflections or whatever. A digital sensor does not choose to render green above and red below... Link to post Share on other sites More sharing options...
jaapv Posted April 18, 2015 Share #299 Posted April 18, 2015 - Some "best APO lenses" are not actually "the best". E.g. the APO Summicron 90. With all due respect, this really sounds like grasping at straws... Link to post Share on other sites More sharing options...
Peter Branch Posted April 18, 2015 Share #300 Posted April 18, 2015 ............- the sensor, be it film or digital, is not just sensitive to the precise wavelength where the graph crosses the y-axis............ I should have written "....where the graph crosses the zero value of the y-axis......." My apologies for being less than precise. Link to post Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.