Question about Focus and Purple Fringing

[nlk10010]

Advertisement (gone after registration)

I've searched the web pretty well to get a better handle on purple fringing, when it is "too much" and the relationship between it and focus (i.e. if it's potentially greater in parts of the image that are less in focus). I understand low ISO exacerbates it and that, at least in theory, much of it can be removed in Lightroom, but would like to get some info from the lens techies on this forum.

 

I've attached a crop of an image that appears to show a lot of PF, although it was taken at ISO 800 and is woefully out of focus. It was taken with a colleague's Summarit 90mm (2.4) with a B+W XS-Pro Clear filter in place and the lens hood screwed in. Would this be considered an "excessive" degree of PF, say enough to perhaps avoid buying the lens in the first place? Or are the conditions so extreme that this would not be considered a problem?

 

Any info (or pointers to good references) would be appreciated.

 

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!

[jdlaing]

Post the uncropped image.

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[nlk10010]

OK. Didn't realize that would be better. Seems harder to see but I'm eager to learn. 

 

BTW, I stipulate to the fact it's OOF.

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!

Edited March 8, 2018 by nlk10010

[a.noctilux]

Looking at the two images, regarding the 100% factor, for a Summarit-M lens, it's normal for such fringing.

Maybe an Apo lens can do better.

 

Try without filter if it's better.

 

Which is your camera ?

[nlk10010]

a.noctilux:  M10. 

 

I have other test shots I made exhibiting a similar degree of the fringing; If it's normal for a Summarit-M then I would need to weigh the trade-off between the cost and weight of a 90mm Apo (Summicron?) (in which case the Summarit wins) and its (perhaps) better performance in the PF area. 

 

Thanks very much for taking the time to post your opinion.

[jmahto]

With white background (high contrast), you will see some purple fringing at wider apertures in almost all lenses. Only severity will vary.

 

If you don't want them then use either more costly lens (with better correction, such as APO), or use narrower aperture, or use LR to remove them (most of the times it works like a charm), or simply ignore them. In your case, I would ignore them since fringing is very small. You haven't seen severe fringing yet.

Edited March 8, 2018 by jmahto

[nlk10010]

Advertisement (gone after registration)

Yea, you know, I realized after-the-fact (when I read up on PF) that the aperture at which this photo was taken could be an important piece of information (I didn't stop to take note of it when I clicked the shutter) but the M10 body doesn't get it from the lens, hence can't transmit it to the EXIF. I'm really am in the dark here.

 

I just wanted to know what I would be getting myself into if I chose to get this lens:  I guess you're saying that, for at least the copy I tried, this degree of PF is normal for the Summarit. Good to know. And I understand completely your last comment, everything's relative.

 

Thanks much.

[jdlaing]

What was the aperture sett to?

[jdlaing]

What I would do is set the exposure comp down to counteract a little of the bright white or close the aperture a little.

[jdlaing]

And it’s not out of focus that bad either. Again......close the aperture just a little.

[nlk10010]

jdlaing:  Certainly makes sense, I had to return the lens but I can borrow it again and repeat and probably try that same shot at a variety of apertures. At least I know the copy I tried is not atypical so I will have an idea what to expect if I get a Summarit for myself. As for the focus, well, I'm still honing my RF skills and the 90mm is a tough test (both from the point-of-view of the thin DOF and camera shake with lower SS).

 

Thanks for the posts.

[spydrxx]

Put the camera on a tripod to eliminate camera shake as a factor in your evaluation.

[jaapv]

With white background (high contrast), you will see some purple fringing at wider apertures in almost all lenses. Only severity will vary.

 

If you don't want them then use either more costly lens (with better correction, such as APO), or use narrower aperture, or use LR to remove them (most of the times it works like a charm), or simply ignore them. In your case, I would ignore them since fringing is very small. You haven't seen severe fringing yet.

Expensive lenses like the Noctilux and Summilux 50 will fringe quite badly as well Fortunately it is simple to correct in post-processing.

[pgk]

I've searched the web pretty well to get a better handle on purple fringing.....

 

Its a poorly understood 'problem' which has been discussed on the forum many times.

 

FWIW these are my thoughts on it. I think that its an 'edge' effect - a problem which occurs under a variety of conditions on imaged edges of great contrast. There are probably a number of factors which come into play to both produce and exacerbate it. One of these is the edge transition of the image projected by the lens and in well designed, fast aperture lenses this transition is very abrupt at wide apertures and as a result fringing is marked. Personally I doubt that the fringing is inherent in the lens produced image, but it is probably added to be longitudinal chromatic aberration as transitions become less abrupt as they become out of focus - so its not a simple and easily resolved problem.

 

There are probably other factors such as; interaction of the imaged fine detail, their contrast and colour, and the sensor array, light ray incidence angle onto the sensor, offset of micro lenses (its usually worse in the corners), specific sensor design, light 'colour' and so on. If it had one single, easily solvable cause it would have been sorted out long ago but just looking at the potential interaction of some of the possible 'causes' or 'exacerbating factors' above suggests that its not an easily calculable and solvable problem.

[wda]

I presume your camera was set to recognize the lens when attached?

[LichtUndDunkelheit]

I have seen color fringing in many of my images, but so far it never bothered me in any image to the degree I wanted it removed.

 

If the image quality is otherwise excellent, who cares about a bit violet or green in places ?

 

Heck there are people who intentionally add glare and ghosts to their images just for the flair.

 

Theres the special case when people wear black and white stripes when it probably would be desireable to remove them, but thats a really rare kind of cloth for people to ever wear.

[TomB_tx]

Sean Reid's reviews run tests set up to compare such fringing, and finds some amount in nearly all lenses. Because he sets up the same test targets and lighting conditions you can compare his reviews at different apertures. I don't recall his results on the 90, but generally the Summarit series do quite well. His site is a paid subscription, but if you are concerned about such things it is a good resource.

Edited March 9, 2018 by TomB_tx

[lct]

Some lenses are more prone to fringing than others and some apo lenses can fringe as well. I would avoid overexposure, as far as possible, in the first place.  

[nlk10010]

Thanks to all the posters, really, my purpose in posting (which I probably did not articulate very well) was to determine if the PF I was seeing with this copy of the lens was "typical" for the Summarit.

 

I realize the degree to which PF (and other lens characteristics) bother people is personal:  I recall reading a thread on these boards (when I searched) about a "famous" PF problem with a 50mm lens (don't recall if it was a Noctilux or Summicron or what). Yet many people love that lens regardless. I also use Olympus and in that world there is a famous question of PF with Pana-Leica lenses on Olympus bodies, the prescription often given is to use a particular type of UV filter. Now the Summarit sample(s) I posted were exported from Lightroom after the 90mm Summarit Lens Profile was applied (with Chromatic Aberration checked) so this is about as good as it's going to get, unless there's an additional way to correct for PF in LR that I'm not seeing.

 

Again, I realize how you react to this issue is personal, so instead of asking "is this acceptable" (although that's maybe how it came out) I just wanted to know if this degree of PF was what I normally would expect with the Summarit. I could then make my decision as to whether the pluses outweighed the minuses.

[Advertisement]

Hallo,

Look here: Leica M

[UliWer]

A recent example I saw were results from the new 1:1.25/75mm Noctilux-M and the 1:2/75mm Summcron-M:

 

The Noctilux showed rather strong purple fringing even with smaller f-stops; for the Summicron it was negligible.

 

At the same time the Noctilux showed more contrast - microcontrast as well as brightness contrast.

 

I once tried to produce purple fringing with a fully opened 1:1.5/85mm Summarex - without any success; with an 1:2/90mm Summicron Apo. Asph. it is very rare that you see purple fringing, but can't rule it out completely.

 

As a thumb's rule - always to be used with a grain of salt - one  may say:

1. A lens with high contrast, very well corrected for high opening will show more purple fringing than a low contrast lens with less correction even if the latter  is wide opened (Noctilux./.Summarex). 

2. If the high contrast is achieved notwithstanding a wide opening, the risk of purple fringing will increase compared to a less wide opened lens (Noctilux./.Summicron). 

3. A lens with high contrast and longer focal length which has no special correction against chromatic aberration (i.e the cheaper lens) will probably show more purple fringing than a similar focal length with apo-correction (Sunmmarit./.Summicron AA).