35/2 Purple fringing

Olrocker · 11-05-2007, 03:18 PM

Is this amount of PF usual for an FA 35/2 (K10D)?

Just wondering...

Ari · 11-05-2007, 04:28 PM

Wow-pretty pronounced, I haven't noticed anything like that with that lens on my *ist d. Gonna have to inspect some shots I took......

ZaphodB · 11-05-2007, 04:41 PM

I do sometimes get a fair bit of PF on white objects in very high-contrast scenes... but not more than most lenses. If I try to avoid blown highlights it isn't really an issue (and I usually try to avoid blown highlights anyway

). Basically I don't think it's a problem that's specific to this lens.

d.bradley · 11-05-2007, 05:50 PM

Are you shooting wide open?

Olrocker · 11-05-2007, 06:42 PM

These were at 2.0 or 2.8, I think.

ZaphodB · 11-05-2007, 06:54 PM

Originally Posted by Olrocker

These were at 2.0 or 2.8, I think.

If I can quote Klaus Schroiff... "Typical for many ultra-large aperture lenses you can experience some purple fringing in extreme contrast conditions at wide-open aperture."

It isn't pretty and can be a P.I.T.A. to correct, but it's completely normal behaviour when shooting wide-ope or nearly wide-open on a lens like this. Unfortunately few lenses are perfect

slick100d · 11-05-2007, 10:08 PM

How much were those cropped? I had never noticed purple fringing with this lens before, probably because I mostly use it indoors. However, while processing a picture the other day I saw some PF while zoomed in. It surprised me. My Promaster (Tamron) 70-300 is the master of PF, so I'm used to it. The picture was wide open (f2.0) and I couldn't see the fringing at normal, full screen viewing.

Olrocker · 11-06-2007, 05:26 AM

These are 10 mp shots, 100% crops from the center. I guess it surprised me too, because I also had mainly used it indoors before and had not noticed any PF to speak of.

roscot · 11-06-2007, 06:06 AM

Purple fringing is really sensor bloom. This is the case in your examples. The DR was just to much for the sensor to handle. It has nothing to do with the lens. CR is associated with a lens where you get red/cyan or blue/green color cast along high contrast edges. I've never noticed any CA from my FA35.

Abbazz · 11-06-2007, 07:00 AM

Originally Posted by roscot

Purple fringing is really sensor bloom. This is the case in your examples. The DR was just to much for the sensor to handle. It has nothing to do with the lens.

I beg to differ. Certain lenses (mostly zooms, but also some prime lenses with many elements) are much more likely to cause purple fringing (call it sensor bloom if you prefer) than others. Lenses with a very contrasty rendering are also more sensitive to this problem.

Originally Posted by roscot

CR is associated with a lens where you get red/cyan or blue/green color cast along high contrast edges.

I think nobody here was talking about chromatic aberrations (I guess this is what "CR" means), which is indeed a different problem.

Cheers!

Olrocker · 11-06-2007, 08:41 AM

Many thanks for all the responses. My first thought was the lens might be defective, but it only does it in very high contrast situations. I'll just need to correct it in PP when it occurs.

Here's a crop of an outdoor shot with less contrast

roscot · 11-06-2007, 09:23 AM

Originally Posted by Abbazz

I beg to differ. Certain lenses (mostly zooms, but also some prime lenses with many elements) are much more likely to cause purple fringing (call it sensor bloom if you prefer) than others. Lenses with a very contrasty rendering are also more sensitive to this problem.

I think nobody here was talking about chromatic aberrations (I guess this is what "CR" means), which is indeed a different problem.

Cheers!

OK I stand corrected. From Wikipedia lots of potential reasons.

Purple fringing - Wikipedia, the free encyclopedia

Other explanations

Purple fringing is usually attributed to chromatic aberration, although it is not clear that all purple fringing can be explained this way. Other attributed causes of purple fringing include:

Sensor effects:
- Chromatic aberration in each CCD cell (microlenses)
- Digital noise in dark areas
- Image processing and interpolation artifacts (almost all CCDs require considerable processing)
- Stray ultraviolet light
- Stray infrared light
- Image bloom from overexposure
- Leaks between cells of the CCD