Purple Smearing w/K-5IIs

I've noticed distinct purple smearing in some shots taken with bright light either in or just off of the frame. Please look at the following.

These were taken with the DA 16-45, which I've owned for 10 years, and the DA 10-17, which I've had for 8 years. Neither lens has done this before. I've previously used the *ist D, K10D, K20D, K-5, and K-5II, and did not get this kind of purple smearing with those bodies.

For comparison, here is a photo taken with the K-5II in similar contrasty light. There's no purple smearing.

So, the lenses have performed fine on previous bodies. All previous bodies have performed fine. I have to conclude, therefore, that one of the following is causing the problem.

1. I have a defective K-5IIs.

2. Removing the AA filter increases the susceptibility of the sensor to this kind of smearing.

Since the camera seems fine otherwise, I am inclined to think that removing the AA filter makes the sensor susceptible to purple smearing in the right light situations. This is a problem for me, since these are light situations in which I often shoot. I may have to either get rid of the K-5IIs, or use it only for landscape work. If I am right, I may not be able to buy any future Pentax cameras, since Pentax seems to have permanently removed the AA filter.

I would appreciate any throughts.

Can anything be done to get rid of the purple smearing?

Thanks,

Joe

No filter on the front Joe and you have a lens hood attached?

On the first pic, I would expect purple fringing in that situation whether I was using the k10, k20, k5 or k5iis.

The second pic - is see contrast lost on the left side of the frame, but I'm not seeing the purple smearing. If you're talking about the goldish structure to the top of the frame, are you sure that's not blue and there's a white balance problem or some other color shift? It looks too perfect to be smearing or fringing... but then I'm only looking at the size you uploaded so I can't zoom too close.  This second pic it looks like light is coming in more to the photographer than the 3rd pic where you can see the highlight and reflection on the far wall - the light is clearly going away from the photographer so you wouldn't expect flaring like you would in the second pic.  And I think a hood would have resolved much of the problem.

I'm not seeing clear evidence of a defect...

From old habit I have a skylight filter on the DA 16-45. None, of course, on the DA 10-17.

I don't use lens hoods. Too bulky in the bag.

But these have been my habits for many years, and this problem has only cropped up with the K-5IIs. So filter and lens hood are not the problem.

Interestingly, I *always* use a lens hood and *never* use a filter, other than the occasional (rare) polarizer. That's an old habit that works for me.

Joe I really think any camera would do this but suspect a lens hood would have helped the extraneous top and side lighting flaring like it has, mind the top daylight through the dome is a tough one to control. I am pretty certain the lack of the AA has any great effect IMO.

Hi Joe,

I notice in your gear list you have a standard K5 (if you still own it) so you could try a test in similar lighting situations using both cameras with and without a hood, the would clear up any debate!

The first shot where the entire circular room seems to be lit by this dome is asking a hell of a lot from a lens given that the interior of the room is perfectly exposed and it looks blueish rather than purple, there are extremes of contrast, this looks more like flare within the lens.

The second shot could have been helped by a hood just to stop the glare falling on the front of the lens, it does not always have to be sunshine, I've had flare even in moonlit situations, extreme contrast is enough.

Joseph Tainter wrote:

MightyMike wrote:

Have you taken the same photos with previous cameras, your example didn't show anything clear about the same conditions (bright light sources in or near the frame)

Mike, look at 2 and 3. They are the same light conditions--darkish indoor light with bright light at the left side of the frame.

Joe

image 2 = iso 2200

image 3 = iso1600

image 2 sensor has considerably more gain than image 3 and is correspondingly more able to display flare from off axis light sources.

also note a few degrees variation of light angle will effect whether flare occur or not and in your two image the angle is considerably different

It also looks like image 2 bright light was streaming through stain glass window (hence tourist taking front on picture) and image 3 an overcast day.

My thoughts are that if you think there is an issue with the K5 sensitivity to Purple Blooming, why don't you do a controlled comparison using the same lens, with the same scene and different cameras that you have. That would be better than speculating. I for one would like to see that.

I could imagine that the situation your K5iis captured is just the way it was inside the church.

You vae a very specially designed dome with a glass element in the center. This is not common for the design of domes - dome light usually is indirect light comeing from the sides of the dome and reflecting within the dome - the light becomes very soft. Your dome is designed in a different way - it works like a flood light resulting in very harsh light.

There is a glass element in the center of the dome - do you know in which way the light from outide breaks within this element? it could lead to this bluish smear you see in the pitures - our eye will not recognize this as it is not able to resolve contrast situations like the one we have in your photo.

The last photo of your series has completely different light conditions - so you cannot compare them.

My conclusion would be: If you do not like the light in the church - blame the architect of the church - nor your K5iis

Best regards

Holger

Joseph Tainter wrote:

Thanks for your comments, Holger.

Actually, photo one is the Pantheon in Rome. The hole in the dome (the oculus) is a hole--no glass.

I've just never had this kind of problem before.

Joe

Hello Joseph,

thank you for your information. It seems that I do not know enough about architecture. The Pantheon and the Occulus - I just knew it by Name. I should travel more to got better education and cultivation.

I just googled for your church and for some picutres:

https://www.google.de/search?q=pantheon+oculus&source=lnms&tbm=isch&sa=x&ei=n3pfu-cac-je0qxr0ohoag&ved=0cayq_auoaq&biw=1099&bih=479&dpr=1.35&safe=vss

There are many photos showing this effect - and I do not think that it is because of Problems with the cameras but an effect of the light that Comes through the occulus. Maybe the angle of the sun has an effect on the extension of the blue smear of light.

I could imagine that the entire building works a Little like a very big camera obscura with the occulus as the hole - at least the bulding is designed in a very special way. The occulus is lens and aperture at the same time - and these elements may cause braking of light or spliting it up into ist wavelenghts.

I do believe that your photos show an effect that is really there in this building under certain light conditions.

Best regards

Holger

Joe, the purple smearing is just from badly over exposed highlights. It has nothing to do with the presence of an AA filter or not. I have seen it many cameras when the contrast range is beyond what the sensor can capture.

Here are two shots on the K-5 IIs. In the first, you can clearly see the purple smearing on the left hand side of the window. The second has some heavy negative exposure compensation dialed in and the purple smearing is gone.

-0.3 EV

-2.7 EV

These are not quite the same shot, but I think they illustrate the point. I noticed you were using quite a high ISO which would have reduced your DR some.

Just to prove the K-5 IIs can take pictures of cupolas. I missed on DOF for this one, but the exposure is good.