[lessisme]

hello,



no, i am not always happy, when my expected new tests don´t shine up -



BUT: thank you so much for those, i could find - over the years. I discovered f.i. samsung nx-lenses, and the very qualities of Canon fx....



okay?



there is one rather new question or idea when looking at the pictures of the latest test of sony´s 135mm-lens.



You showed in detail the kind of flare or fringes (purple) at highlights as part of "bokeh".

Did you ever try this effect with diffrent sensors (for the same lens of course). After whatching that "orbs-issue" with Fuji F10 - there could be a question about the sensor reacting to highlights.



As i use some diffrent lenses on Canon 5d (II), it should be interesting and simple, whether you will reproduce the same "picture" or effect on other sensors.



thank you again!



guest 1

[Klaus]

The origins of purple fringing are not really clear - at least to me. The best guess is a blooming effect by the micro lenses on the photodiodes - and all sensors have them.

[wim]

[quote name='Klaus' timestamp='1337362560' post='18286']

The origins of purple fringing are not really clear - at least to me. The best guess is a blooming effect by the micro lenses on the photodiodes - and all sensors have them.

[/quote]

If I may expand on this:



PF also occurred in the film era, although generally not as pronounced and certainly in B&W not as obvious, but always as either purple fringes/haloes at high contrast transitions in colour, or intermediate grey tone haloes/fringes in B&W.



It only occurs at large apertures, and not with all lenses, but mostty large aperture lenses. Because it is stronger in digital, one can indeed not help but think some blooming effect aiding in the creation of this deect, but as this also occurred with film, something else must contribute to its creation as well. Since PF can occur anywhere in a picture, just as strong in back and foreground or centre and edges, in combination with the fact that it may surround an object fully provided its edge is a high contrast edge, it can not be CA, as there is no specific tangential or sagittal relation. Neither does it get worse towards the edges, thereby excluding general optical flaws. It does get less when stopping down, to the point of disappearing at smaller apertures, so therefore it does seem lens related in some way.



IOW, there appears to be both a lens related component and a sensor or medium related component at play.



There has always been a fair amount of speculation on the cause of PF, but so far no one has offered a proper scientific explanation, not to my knowledge anyway.



Kind regards, Wim

[Guest]

Hi Wim



I'm not sure that the fact that PF diminishes with stopping down necessarily indicates there is a lens involvement directly.



As you stop down, I think the angle of the light rays hitting the sensor area changes.



In which case PF could be related to the angle the light hits the emulsion (film) or micro lenses (digital)



So yes, lens settings have an impact on the result, but maybe not due to any flaw.



I wonder if there is any correlation between telecentricity and reduced PF?

[Klaus]

Are you sure that this was PF ?

Bokeh fringing can look like PF in certain situations.

[Guest]

[quote name='Klaus' timestamp='1337412749' post='18297']

Are you sure that this was PF ?

Bokeh fringing can look like PF in certain situations.

[/quote]





Just to throw another ingredient into the soup, 'older' lens designs seem worse for PF, in addition to the other factors such as wide aperture.



An example is the Nikkor 180mm F/2.8.



So you could argue that yes, telecentricity is a factor in PF. However, the example I quote here is a tele lens, and a redesign would be unlikely to make incident light rays much more perpendicular to the sensor.



And yes, before anybody says it, some modern designs show the effect too. For example Nikkor 35mm F/1.8 DX



Regarding the effect of sensor, has anyone seen an improving trend with newly-introduced sensors? I'm still back in the dark ages with D40/D200.

[Brightcolours]

A lot of different things get attributed to purple fringing here in this thread already.



But not all fringes are "purple fringing", even if the fringe is magenta, purple, blue-ish or reddish.



First we have CA or lateral CA. This kind of fringing never occurs in the center, and stronger the more you go to the edge of the frame. This is caused by different wavelengths of light get bend more or less through the optics, projecting some colours bigger than others. This kind of fringing always comes in pairs or colours, although at times you only see the 2nd colour on closer inspection. Blue-yellow, red-blueish green, magenta-green.



Then there is LoCA or longitudinal CA. This appears on OOF areas, mostly with wide aperture lenses. It shows up as green and magenta fringing around bright areas and spots. It occurs when different wave length light does not get focussed on the same plane.



Then there is a normal transition in blur, which often gets mistaken with LoCA. It also shows up around edges between light and dark, and is just a normal darkening of the blue sky for instance, mixing with the dark subject colours like leaves and branches. It is no kind of aberration, just something our minds have trouble in identifying as we do not see it when we look at things with our eyes (because we only look at what is in focus with eyes, but on photos we can actually look at things that are blurry in the background).



In the example of the Sony 135mm STF we do see PF. And since we do not see that with every lens on the same Sony camera, it is not caused by the micro lenses only, the lens does need to have influence on it occurring.

Then there is purple fringing. It does not occur with film, it is a digital age thing. Like Klaus suggests above (micro lenses) is not correct, else we would see it with every lens with the same light intensity and focal length (and we do not see that). So it has nothing to do with the sensor all on its own. It is lens dependent. The odd thing about purple fringing is that it is well understood by camera and lens makers, but on internet almost no one seems to have a clue... kind of like how exactly lens and camera work together with phase detect AF.

Lens manufacturers have changed lens coatings to combat PF (big sign also that it is not micro lens related). Tamron and Sigma have actually put that change in their nomenclature:

Lenses with changed coatings from Tamron started to get the Di addition in their name (and then APS-C only lenses got Di II).

Lenses with changed coatings from Sigma started to get DG in their name (and APS-C only lenses got DC).



The reason why PF shows up has to do with light intensity. Very bright areas will show a magenta, blue-ish or even reddish glow around them. Stopping down helps, because stopping down lowers light intensity.

]PF is caused by the sensor surfaces being reflective. Light gets reflected back to the lens, and somehow gets again reflected back to the sensor, creating blooming effects. Exactly how the light behaves I do not know (I am not in the camera or lens industry). But the added coatings in modern lens formulations do get rid of the PF. Usually we only see PF with certain older lens designs or certain cheap-ish compact cameras.

[wim]

[quote name='DavidBM' timestamp='1337394188' post='18292']

Hi Wim



I'm not sure that the fact that PF diminishes with stopping down necessarily indicates there is a lens involvement directly.



As you stop down, I think the angle of the light rays hitting the sensor area changes.



In which case PF could be related to the angle the light hits the emulsion (film) or micro lenses (digital)



So yes, lens settings have an impact on the result, but maybe not due to any flaw.



I wonder if there is any correlation between telecentricity and reduced PF?

[/quote]

I was only saying it is lens related, not necessarily (purely) optics related. FWIW, it could be caused by/related to/influenced by internal lens reflectins and/or reflectins of the last element(s). Furthermore, IMO it is likely that if it is also sensor related, it could well also be emulsion related to some degree. I can think of a few scenarios that could make this possible, and IMO it is telling that it is less pronounced with emulsion than with sensors.



Anyway, as mentioned, nobody has come up yet with a proper explanation by what it is really caused by.



Kind regards, Wim

[wim]

[quote name='Klaus' timestamp='1337412749' post='18297']

Are you sure that this was PF ?

Bokeh fringing can look like PF in certain situations.

[/quote]

Even lateral CA can look like PF in certain conditions, especially when high contrast items are involved, e.g., when chrome parts are part of the objects in the pictures, etc.



This is why I like your pictures of reflections off water and/or droplets, which show very well the complete haloeing/encircling effect of PF, which true optical only defects don't show. Relatively small objects which are high contrast items show PF best.



Kind regards, Wim

[wim]

[quote name='AAC7man' timestamp='1337418561' post='18299']

Just to throw another ingredient into the soup, 'older' lens designs seem worse for PF, in addition to the other factors such as wide aperture.



An example is the Nikkor 180mm F/2.8.



So you could argue that yes, telecentricity is a factor in PF. However, the example I quote here is a tele lens, and a redesign would be unlikely to make incident light rays much more perpendicular to the sensor.[/quote]

I would think specific internal reflections, which are more and more eliminated by stopping down, aid and abet to this effect. Maybe the 180 has more of these flaws than another lens. I am thinking coating, and possibly reflections of metal parts in the lens.

And yes, before anybody says it, some modern designs show the effect too. For example Nikkor 35mm F/1.8 DX



Regarding the effect of sensor, has anyone seen an improving trend with newly-introduced sensors? I'm still back in the dark ages with D40/D200.

I would expect, everything else staying the same, that removal of AA-filters, making AA-filters thinner, and using the best possible coatings on AA-filters would diminish the effect. I expect the Leica M-Monochrom to show very little PF... <img src='http://forum.photozone.de/public/style_emoticons/<#EMO_DIR#>/biggrin.gif' class='bbc_emoticon' alt='' />



Kind regards, Wim