[Samoht]

Hi,

Have used my K-3 for a some weeks now and have noticed an issue with PF (Purple Fringing) especially with my Tamron 17-50.
I thought that PF and CA was lens property but I have noticed a difference between K-5 and K-3 with Tamron 17-50 where K-3 gives more PF than K-5, I've corrected more PF during the weeks I've had my K-3 than for almost 3 years with K-5 and same lens.

Anyone else seen this behavior when upgrading from K-5 to K-3? Don't have the same issue with for example my DA* 50-135 so it seems to still be lens related but also camera related.

Attached Images

View Picture EXIF

File Date

 

View Picture EXIF

File Date

 

[Adam]

Since the image is bigger technically the fringes would span more pixels, right? But the overall impact should be the same. In this particular photo I think PP could easily correct the fringing

[Sandy Hancock]

It might be worth taking identical images with both cameras for comparison.

[bm75]

Just a fast thought...Could it be due to pixel density? Did you tried different settings for AA filter simulator? Have you tried shooting on tripod with exactly the same parameters with both cameras? Waiting for FF I'm considering to buy a used K3 for shooting my child's sport activity, so, I'm quite interested in your experience. Let us know your findings. Thanks. Matteo

Is your tamron ok? Mine has never shown that much on either camera.

EDIT:


It is common that the front element gets loose on these lenses. I have hade mine fixed.

[Macario]

PF is a charasteristic of a lens. But that said, the sensor and especially the pixel size play a crucial role in it if you see it or not. As a general rule, the smaller the pixels, the quicker you will notice lens faults like CA and PF.

[rawr]

I haven't seen that level of fringing on my K-3 or K-5 with my Tamron 17-50.

Has that lens taken a knock recently? I initially had one sample of the Tamron that exhibited similar PF with my K-5, but it was damaged in-transit (and also was visibly de-centred).

[Markovo]

Never had that much PF with my Tamron 17-50, using k30. Never had PF actually. Maybe the lense is damaged somehow?

[falconeye]

Thanks for this common instance of fringing.

Actually, this isn't the common CA fringing which is corrected by CA profiles. Although, it still is an effect of Lateral CA, Longitudinal CA and Bokeh Fringing (all three are different effects).

The point is that due to the strong contrast between foreground and background (tree and sky) even a very moderate amount of, e.g., Bokeh fringing makes light "spill over" from the bright to the dark areas. This effect can actually become dramatic if one exposes for the foreground.

Bokeh fringing is the effect that not all RGB rays coincide outside the focus plane, even if they do within (i.e., even if the lens has no CA at all). The sky is at a different distance.

All these effects are measured in µm and span more pixels with a K-3.

To really compare though, not only must the image dimensions be equalized, but also the background exposure and tone curve must be made equal. Only then can a comparison be made.

Under such controlled conditions, a few users have reported a similiar phenomenon in images of the Sony A7r vs. A7. Although I still have to see a convincing case myself.

P.S.
I don't think there is anything wrong with the lens. It is common with strong back-lit scenes and a lesser than top-notch lens.

[Na Horuk]

Yeah, with more MP, CA and aberrations will be more noticeable as well. To deal with this you can try underexposing a little (or at least don't overexpose as the examples). Better yet, shoot raw and use PP to reduce this. Its one of those things where the photographer goes crazy dealing with it, but non-photographer viewers rarely notice it.

Edit: Hm, I thought about it a little more. Another thing that could affect how noticeable PF is is the camera calibration in the raw software that you use. This would affect colours in general, but it can affect what shade the "purple" is, and thus make it more visible.
But I doubt the actual sensor of the K-3 is more prone to capture purple fringing

[TER-OR]

When I switched from the K10 to K5 the Tamron 70-300 I used well on the K10 became unusable on the K5 from purple fringing. Granted, that's a budget lens.

[Samoht]

I don't really think the lens is damaged either but it is a few years old so perhaps it's a combination of normal wear 'n tear and the better sensor pulling out both more details and defects, it hasn't been dropped or bumped into something. It doesn't happen in all photos but happens in more photos than with the K-5.
But reading through the answers it's most likely normal behavior. Thanks a lot!

I only shoot in RAW and it's really easy to get rid of the fringe in LR, the eye dropper tool is actually quite effective.

For the shots in my first post I used the Av mode and while the K-5 tends to underexpose in Av the K-3 seems to overexpose instead. Keep forgetting that as you can see in the background.
But it feels like the K-5 handles different lights (dark and bright in the same shot) better than K-3, at least when it comes to sunlit clouds. Perhaps the difference is that K-3 overexposes in Av mode, or I'm doing something wrong.

Btw, after using DA* 50-135 with K-5 and moving it to K-3 it was a step up, almost like upgrading the lens. The same journey with the Tamron had almost the opposite effect, I've always been quite impressed by the quality of the cheap Tamron but I liked it more with the K-5 than with K-3, would have like to had the same feeling with K-3 and Tamron as with DA*. If it's not a placebo effect of course.

If I would to upgrade to another normal zoom which to choose, DA* 16-50, Sigma 17-50 or perhaps a new Tamron 17-50?

---------- Post added 05-07-15 at 05:55 AM ----------

Is your tamron ok? Mine has never shown that much on either camera.

EDIT:


It is common that the front element gets loose on these lenses. I have hade mine fixed.

Haven't even heard about this issue, checked some youtube clips and mine isn't nowhere near being as loose as the one's in the clips.
The big plastic ring which the lens hood attaches to moves about half a millimeter when wiggling it by hand.

[TER-OR]

Generally with these newer sensors it's best to err on the side of underexposure. They record lots of data and I've seen nearly black images recovered in PP from RAW files to perfectly usable images. I don't know whether this would help with the CA issues, that's due to contrast so maybe it would.

[stevebrot]

I thought that PF and CA was lens property

CA (both lateral and axial) is a property of the lens. True PF is more complicated in that it is more prevalent with some lenses, but is only present on digital capture*. It would not surprise me to have different degrees of PF with different sensors. OTOH, I have not noticed that my K-3 is particularly prone to PF even with lenses having a reputation for such (FA 77/1.8 and FA 35/2). Go figure.

Good description/explanation (See section #3 Chromatic Aberrations, part way down the page)


Steve

* I shoot both film and digital and have not been able to generate PF with color film despite making efforts to do so with both my FA 77/1.8 and FA 35/2. I also can confidently report that I have never seen PF in almost 45 years of film photography. Axial CA, yes...PF, no.

[JeffB]

Here is a very good technical explanation of chromatic aberration Chromatic Aberration AKA Color fringing | imatest