I came across this almost by chance, but it's worth sharing. At this point, I'm convinced that the appearance of this phenomenon with the OM-D is not limited to the Panasonic 7-14mm f/4. It is definitely most obvious with the 7-14mm, but it's clearly there even with the 9-18mm. I was working on processing this shot of a bicycle taken with the 9-18mm. And I wanted to see if I could saturate the colours of the bike a little more. {} By chance, I pushed the purple slider to +100. I wasn't expect that this would be a useable setting, but I like to test extremes before I work my way to the middle. This is what I got: lots of purple in places that I wouldn't have expected: {} So I started looking a little closer at the original image. I've found that the purple "flare" definitely exists even with the 9-18mm. It's a lot harder to see, but its clearly there. You can see the side by side here: {} It isn't overwhelmingly visible on the left side, but there is still very much a faint purple colouring. The defringe local adjustment brush doesn't do anything to it. Anyway, looks like 9-18mm OM-D users will need to watch their lighting as well, though perhaps not as much as 7-14mm users.
The coating on the EM5's sensor / filter is purple-ish. Doesn't' anyone think this is the likely source is the purple flaring? Or is this just too simplistic? Brian
I think others have suggested the same. I'd say its probably the best explanation that I've read--particularly since it doesn't seem to happen on other cameras...at least not to the same degree.
Heh, sure enough there it is... I wonder if this is common to a number of Sony sensors? Someone posted a while ago an article talking about people seeing purple spots like this with the iPhone 5 camera, which also has a Sony sensor. I am really hoping this isn't present on the GH3 too now that it seems almost certain it's using a tweaked version of the same sensor from the OM-D EM-5. Yet another example of why I hate to see everyone homogenizing the market to one sensor manufacturer.
I don't know. In the thread for the OM-D with 7-14mm on this issue, someone, I think, posted some shots from the E-PL5 that (though still having a little purple) were far, far better than the OM-D. So, if it is the sense, then there is, at least, room for improvement.
This has been assigned to the OM-D in particular because the same lens does not apparently cause the same problem on older Olympus bodies or on Panasonic bodies (or at least not to the same degree)
I shot the OMD against another camera -- same lens -- and the purple "fringing" was far worse on the OMD than the other cameras. It's one reason I moved the OMD along. Some of the worst pf was not correctable. Some were fine with it, but I could see it everywhere with the OMD. Some similar criticisms have been leveled against the NEX7. The NEX7 (but not the NEX 5) and the OMD have a similar pixel pitch. I wonder if there isn't an issue with the Sony sensor at this pixel density. This is a reason I'm stalling on getting the EPL5 or EPM2. I want to see if this continues to be an issue or not.
You may want to look at this thread: https://www.mu-43.com/f38/panasonic-7-14mm-flare-case-study-solution-33183/index6.html#post356096 The 7-14mm is far, far worse than the 9-18mm (obviously), and the purple still does show slightly in the E-PL5 images, but even the 7-14mm is much better with the E-PL5. I would expect that it moves from being a bit of an issue with the 9-18mm and the OM-D to not being an issue at all with the 9-18mm and the E-PL5, considering how minimal the 7-14mm is on that body. Since I don't have the 7-14mm, I'm pretty satisfied with the OM-D. I don't often enough directly at the sun or bright lights with the 9-18mm for it to matter. My recommendations would be: OM-D w/9-18mm This is a safe combination as long as you avoid extreme post processing. And I do mean "extreme." Even the "before" image in my OP above where the purple is very, very minimal (and probably safe for a 16x20) has dramatic shifts in highlights and shadows in LR4 as well as a somewhat extreme tone curve applied to the highlight's roll off. OM-D w/7-14mm Either be really careful, be willing to do some intensive post processing and editing, or get a different body. I would anticipate, also, considering the huge difference with the 7-14mm between the OM-D and the E-PL5, that the E-PL5 w/9-18mm would solve the problem entirely. Also, I would like to see someone come out with a 7mm f/2.8 prime. I love the range of the 9-18mm, but I'm sometimes a little jealous of that extra bit of wide angle.
Yeah my EM1 is just as bad as my EM5 with the 7-14mm. I don't think it has anything to do with the sensor itself or who makes it, but rather the filter over the sensor, it seems like Panasonic has a strong uv filter on their sensors, and Olympus doesn't.
Yes - there were several discussions a while back on dpreview that suggested a sharp UV cut would help. I use a Hoya UV+IR cut filter on the 9-18mm and 12-50mm, and although it does not completely eliminate the purple, it perhaps halves the strength. I do not think that the IR cut is necessary. Note that most filters will increase the quantity of flare slightly (coatings...) - so you are to some extent trading a slight increase in flare pattern against the flare colour. I usually prefer to use the PL25mm with a hood and no filter at all, and I do not think that I have had a single non-test image where flare was a major problem as a result.
Yeah, I have a love-hate relationship with my 7-14. It performs well when you are not shooting with bright lights in the frame, as the lens is very sharp across the frame, but when you have very bright lights in the frame the purple blobs will rear its ugly head. And I do think too it's worse on the E-M1 than it is on the E-M5. Here's an old shot with the E-M5 where I was able to avoid the purple flare. Thank goodness Olympus is coming out with their 7-14/2.8 lens. I am willing to bet this is going to be a superb uwa zoom lens.
Another option (instead of mounting a filter on the front of the lens) is to mount a filter at the rear of the lens as described in this thread.
