Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
07-09-2013, 09:16 PM   #16
Forum Member




Join Date: Jul 2013
Location: Atlanta, GA
Posts: 58
QuoteOriginally posted by adpo Quote
Looking at various lens reviews of the Tamron, it seems to be a common issue with this this lens, with the worst purple fringing at it's widest setting. It seems to simply be something about the optical design lens itself that's causing the effect. With that being said, I've typically found this rather simple to fix in post (one click with lightroom) and you also have the added benefit of in-camera correction as well.

Yeah, I have an older Tamron zoom and I'm getting some purple fringes too.
Do the Pentax brand lenses do better on this?

07-09-2013, 10:55 PM   #17
Pentaxian
carrrlangas's Avatar

Join Date: Mar 2012
Location: Cottbus (Berlin)
Photos: Gallery
Posts: 1,738
QuoteOriginally posted by rawr Quote
Do these fringes show up when opening a PEF or DNG in PDCU4, or just JPG's out of camera, by the way?
Purple fringe is a phenomenom that depends on the sensor + lens combination, not the recording format. So yes it will show up on RAW files. However, PP software has gotten really really good at getting reed of it (almost all of it) with very little loss of IQ. Stopping the lens down a little helps reduce its effect.
07-10-2013, 01:25 AM   #18
Pentaxian




Join Date: Jun 2009
Location: GMT +10
Photos: Albums
Posts: 10,600
QuoteOriginally posted by carrrlangas Quote
Purple fringe is a phenomenom that depends on the sensor + lens combination, not the recording format
Not necessarily true. As you no doubt are aware, there are some in-camera options that will only work (as in produce an effect) on JPEG's in the K-5 and other cameras. If the purple fringing is being generated or influenced by these JPEG settings, then the RAWs may not have the issue.
07-10-2013, 04:35 AM - 1 Like   #19
Veteran Member




Join Date: Apr 2008
Location: Pittsburgh, PA
Photos: Gallery | Albums
Posts: 666
While the purple fringing in the first image is extreme, I am not really surprised. The image is taken in very bright sunlight with a very wide angle lens. With the sun that you had, if I was shooting the same building with the 10-17 fisheye I would expect to see a lot of PF in the same places. I don't think it is your sensor, with the exception that PF is a digital photography problem.

Each lens is going to behave differently, but PF is a bigger issue with wide angles. The coatings and elements of the lenses do affect the image too, so you will find some lenses that are more prone to PF and green CA than others. For my lenses, I expect PF from the 10-17, a little less from the 12-24, and only in extreme circumstances from the 16-50. With my primes, the 77 loves to give me a little green CA in the right conditions but the dfa 100 wr never gives me any in the same conditions. Still, all of this has a two click solution in Lightroom, so I don't let it bother me.

07-10-2013, 05:21 AM   #20
New Member
Calmsea's Avatar

Join Date: Jun 2013
Posts: 15
Original Poster
Thanks everyone - I'm digging away to follow up on all suggestions

Wow. Great and helpful responses, which I am working through, and on.

Today's to do list now includes:

- responding to individual replies. Apologies - new to forums - not sure of the protocol re 'likes' etc., but I have found all replies helpful, and several very educational - keep them coming, and do not worry, I am not easily offended, very keen to learn and open re my limited level of understanding - I am reasonably technically competent in other areas, (IT), and Google is my friend, and I will research and learn, so don't be afraid to hit me with the detail, and/or your own observations and experiences.
- doing the liveview test that JinDesu suggested (it's light again now :-)
- installing lightroom 4 (I have it, but got sidetracked with libraries and stuff. Wanted to be clear in my head where I was going to save stuff prior to starting to use it in anger. Yes, I know - analysis paralysis..)
- reading up on LoCA and PF (and differences between - apologies for using wrong term(s))
- taking some purple shots of trees, then resetting custom functions and retaking with identical settings (I save as RAW+ and have no filters resident on any lenses, though do like to use Cokin P series graduated when taking landscapes with bright skies)
- making contact with Pentax Service Centre UK (wherever I get to with the analysis and understanding personally, I think that I'm going to ask them to take a look at the camera, and particularly sensor - they must have some analysis/debug tools to assess the health/performance of the components etc. Thus far I am leaning toward the scenario whereby my camera has some degree of censor/related software issue, which may be intermittent, and may be exacerbated by settings, and/or lens characteristics, but to get such weird and prominent results consistently, something has to be 'wrong' somewhere - consensus does appear to be that the examples that I am getting routinely are exceptionally bad/extreme. I really don't want to be taunted by, and have to do battle in the Lightroom with, the dreaded PF after every landscape/tree/sky shot)

Oh, and then I also need to fit the more mundane things that wont do themselves into what remains of the day:

- mow the grass
- do the shopping
- walk the dog
- vacuum the house
- congratulate my son for not doing any of the above, spending several days in the same shirt and turning nocturnal to avoid the unusual heat (probably the first time in his 16 years that the sun has been out here (Macclesfield UK, 20 miles south of Manchester) for more than 2 days running..)

Puzzles me how I ever found time for 'work'

Don (aka Calmsea)
07-10-2013, 06:19 AM   #21
New Member
Calmsea's Avatar

Join Date: Jun 2013
Posts: 15
Original Poster
QuoteOriginally posted by stevebrot Quote
Welcome to digital photography. As you noted, PF (not the same as LoCA) was not and is not a concern when shooting film*. Some lenses are more prone than others as are some subjects and sensors. Conventional wisdom is that you can expect better results with digitally optimized lenses, though I have to admit to not having a huge problem with most of my vintage glass.


Steve

* Field curvature in combination with LoCA can produce something that looks a lot like PF, but close examination will show that the false color is localized to a radial band around the center of the frame (usually towards the margins). The Zuiko 35/2.8 on my Olympus XA has this fault.
Thanks Steve. After a little digging, my improved understanding is:

- LoCA (Longitudinal Chromatic Aberration, (or 'bokeh fringing'?)) occurs in out of focus areas, whilst PF occurs at contrast transitions, which can be in focus.

I think that I am getting both (PF and LoCA)! The Buxton pavilion clearly being very 'PF', whilst some of my distant purple trees, perhaps being LoCA?

I think that this is all starting to add up and make more sense for me. Using a wide angle lens, wide open does appear to be a good recipe for generating LoCA. And despite Tamron's protestations to the contrary, taking a picture of black and white detail, in bright sunlight, at 10mm is also likely to summon PF, particularly if a lens is prone to, erm, CA.. With regards to the part the sensor may play, in my head, I keep reverting to Adam's excellent description of demosaicing in his recent thread on sensor size and image quality:

QuoteQuote:
The big issue at hand is that color digital image sensors only record a fraction of the color data (light) that they are trying to portray. Each pixel on a color sensor is only sensitive to one of three color channels: red, green, or blue. Most sensors today are designed using a Bayer pattern, which uniformly lays out the pixels in a checker pattern consisting of 50% green pixels, 25% red pixels, and 25% blue pixel (because green light is the most abundant). Thus, after the sensor captures an image, at each individual pixel location, a computer algorithm has to try to calculate (a.k.a. guess with high accuracy) the intensity of the two missing color channels (this is known as demosaicing). Demosaicing techniques generally work by analyzing the neighbors of each pixel and calculating missing color intensity values based on the intensities found at neighboring locations. But what if the neighboring intensities are not accurate due to noise? The short answer is that the effectiveness of demosaicing drops dramatically, as the guesses about the missing two color channels could be bad or even flat-out wrong. This is what leads to significant loss of edge detail and color noise.
If noise can play a part in confusing the demosaicing algorithms with incorrect neighbouring colour intensities, and hence result in bad colour 'guesses', couldn't a similar issue be caused in areas of very high contrast? Perhaps exacerbated by overexposed (and hence inaccurate colour intensity) highlights, and maybe even LoCA from a lens prone to it.

Question: Do lenses that are known for LoCA always generate more PF? Does this reduce as censor size increases?

I suspect I may be going a little of piste here, so I will desist, and seek to understand more :-)
07-10-2013, 06:26 AM   #22
New Member
Calmsea's Avatar

Join Date: Jun 2013
Posts: 15
Original Poster
QuoteOriginally posted by rawr Quote
I'm with BirdDude007. Purple fringing (PF) happens, and is easily eliminated, especially with Lightroom, but also with DxO Optics Pro and other tools.

My worst PF (and green fringing) lenses are my Tamron 70-300 and Pentax K 300. I also had a bad copy of the Tamron 17-50 that was a un-correctable PF monster, but the 2nd copy I got has no PF at all. So I think PF is 99% a lens issue, not a camera issue.

I also have to ask OP: were you using any filters on your lens(es)? If yes, take the filters off and just use a hood. I wouldn't be surprised at all to see a bad UV filter introduce PF into a lens.
No filters or hood.
07-10-2013, 06:35 AM   #23
New Member
Calmsea's Avatar

Join Date: Jun 2013
Posts: 15
Original Poster
QuoteOriginally posted by Lloyd_Christmas Quote
Yeah, I have an older Tamron zoom and I'm getting some purple fringes too.
Do the Pentax brand lenses do better on this?
At risk of looking even more stupid, my understanding now is that my kit DA lenses do indeed display LoCA (purple tree phenomena, usually a bit blurry and often when shooting on wide aperture. I am using single point of focus, hence cannot guarantee that trees are actually in focus) but not really displaying the PF that my Tamron does. The latter excels at PF, largely at widest angles, and most open apertures.

I suspect, as has been pointed out by several, (thanks), the focal length, aperture, and contrast levels of subject are key. The more prone the lens is generally to CA, the more likely it is to paint purple fringes on things when pushing these parameters (my interpretation :-))

07-10-2013, 06:52 AM   #24
New Member
Calmsea's Avatar

Join Date: Jun 2013
Posts: 15
Original Poster
QuoteOriginally posted by Allison Quote
While the purple fringing in the first image is extreme, I am not really surprised. The image is taken in very bright sunlight with a very wide angle lens. With the sun that you had, if I was shooting the same building with the 10-17 fisheye I would expect to see a lot of PF in the same places. I don't think it is your sensor, with the exception that PF is a digital photography problem.

Each lens is going to behave differently, but PF is a bigger issue with wide angles. The coatings and elements of the lenses do affect the image too, so you will find some lenses that are more prone to PF and green CA than others. For my lenses, I expect PF from the 10-17, a little less from the 12-24, and only in extreme circumstances from the 16-50. With my primes, the 77 loves to give me a little green CA in the right conditions but the dfa 100 wr never gives me any in the same conditions. Still, all of this has a two click solution in Lightroom, so I don't let it bother me.
Thanks Allison. Helps my understanding. Also understand the Lightroom angle, and will get it up and running and see if I can't come to make friends with my PF a little more, or at least accept them.

Something that does not quite tally in all of this being the insistence from the Tamron UK distributor that they have tested the rebuilt lens and not seen any CA or PF.. I did ask on which camera body they tried it, but did not receive a response to that. May chase on that one.

From the Tamron UK horses mouth:

QuoteQuote:
Refrenece our phonecalll conversation as discussed as a precautionary measure we replaced all element group's within your lens and re aligned the light axis to suit, After a Full check and test of your lenss we found it to be within Manufacturers specification. After seeing the images you attached and took post receiving the repaired lens back from us i would advise getting the Camera body inspected as we are to believe there is an issue this end rather then with the lens.
He (the Service Manager) is a full size sensor Nikon user apparently, but was adamant that he had never seen anything like these pictures throughout his time in the industry (didn't establish how long that was).
07-10-2013, 06:57 AM - 1 Like   #25
Site Supporter




Join Date: Jun 2011
Location: Texas
Photos: Gallery
Posts: 1,116
Good analysis. The fringing is most certainly primarily a lens based issue due to CA that is most obvious with high-contrast edges. Sean Reid (Reid Reviews) evaluates the tendency with controlled setups when he tests and compares lenses. Fringing isn't always purple, depending on which color of light is least corrected in the lens design. Some lenses have more reddish fringing, with is often less obvious than purple. A true APO lens shows much less fringing, again showing that it is CA based. On my Leica M9 I find Zeiss lenses tend to have low levels of fringing. Leica recently introduced a lens designed to eliminate the issue (as did Zeiss), but the cost to do so puts the lenses out of reach of most of us. The normal Leica Summicron 50mm f2.0 lens is only about $2K, and does show significant PF, while their new APO ASPH Summicron 50mm f2.0 is $7K.
As noted, fringing can also be corrected in processing, and several cameras correct it with in-camera firmware, which is more cost-effective than trying to make a perfect lens. I believe Nikon also takes this approach.
07-10-2013, 01:46 PM   #26
New Member
Calmsea's Avatar

Join Date: Jun 2013
Posts: 15
Original Poster
Liveview and reset of camera results

QuoteOriginally posted by rawr Quote
Do these fringes show up when opening a PEF or DNG in PDCU4, or just JPG's out of camera, by the way?
PEF and JPG's both exhibit.

QuoteOriginally posted by JinDesu Quote
Can you take your Tamron or your DA 18-55 and put the camera on a tripod, set to F8, and manually focus in Liveview - until distant objects look sharp, and then take a few backlit shots and show us what happens?
I have now done this.

QuoteOriginally posted by rawr Quote
This got me to thinking about whether there are any settings that the OP may have made to their custom image or JPG settings that may be playing a role too. It's almost like a camera filter is active on high-contrast edges. Maybe a reset of camera custom functions and menu settings (pp 335-336 of the K-5 manual) to their defaults is worth a try.
I've done this too.

Four pictures below, shot on a tripod, with my DA 18-55, at F8, and manually focussed using Liveview, as requested by JinDesu. What I did was shoot two shots of exactly same subject, one focussed on foreground, and one on background. I then did the same after the camera was reset, but being a bozo I had moved the camera by then hence the subject for the latter is different. Actually I took a lot more shots before the reset, and have selected those that are best. That's my excuse for not leaving the camera in situ - I didn't know which shots were working, and which not. Just to add more to the mix, the first pair are at 55mm focal length, the second 18mm. As anticipated, the PF is far more noticeable at the shorter focal length.

My observations/conclusions:

- I could see the purple come and go on the trees as I changed focus in Liveview (which I think proves its a lens issue, as the majority have said all along)
- I think that the majority are more prominent when in focus(ish - actually they seem at their worse when just coming in, or just going out, of focus) so in the first shot, they are hardly evident anywhere, but in the second, with the focus moved to the trees in the background, they were lurking again in the areas of high contrast (gaps in the foliage). Third and fourth both have distinct fringes, and its hard (for me) to be sure on what was in focus completely on the latter - intended to be the trees in the background, but I struggled using manual focus via the back screen in daylight. The tree on the left looks pretty sharp, and very purple in places.. On the third picture, the edges of the bamboo lantern are reasonably in focus, and fringing.
- Pretty sure that I had something going on with the custom settings also. The custom image icon on the status screen certainly changed after reset. Very possible that whatever I had done with this sometime previously was not helping.

So, I'm going to park the trip to the Pentax Service Centre and start afresh, armed with a reset camera, a rebuilt lens, a lot more knowledge re what sort of stuff antagonises PF, (and maybe LoCA), and my Lightroom 4 disc poised for installation, to be used to defringe just as soon as I get some images I want to keep for posterity, or that merit sharing.

Thanks to all for assisting. Please don't hesitate to provide more comments, observations or advice. I'm beginning to realise the road to competence digitally is looking a lot longer than I had anticipated - thought my previous film experience would bear me in good stead, but no longer think that is the case - new rules, and lots more to learn it would seem.

I really want to persevere with the K5 - I like everything about it, apart from my peculiarly purple results thus far :-)


Don (Calmsea)
Attached Images
View Picture EXIF
PENTAX K-5  Photo 
View Picture EXIF
PENTAX K-5  Photo 
View Picture EXIF
PENTAX K-5  Photo 
View Picture EXIF
PENTAX K-5  Photo 
07-10-2013, 01:52 PM   #27
Loyal Site Supporter
TER-OR's Avatar

Join Date: Apr 2009
Location: Dundee, IL
Photos: Gallery | Albums
Posts: 5,417
I used to sue a Tamron 70-300. There was fringing on the K10, but it was much worse on the K5 once I upgraded. I bought a DA55-300 and it's much less prone to the fringing.

I will have to look into the controls in Lightroom. I do tend to use the lens profiles in there, just out of habit.

I do always use a lens hood, always have. OK, there are some old M lenses I don't have hoods for...
07-10-2013, 04:08 PM   #28
Veteran Member
crewl1's Avatar

Join Date: Jul 2010
Location: Los Angeles, CA
Photos: Gallery | Albums
Posts: 9,795
QuoteOriginally posted by Calmsea Quote
Four pictures below, shot on a tripod, with my DA 18-55, at F8, and manually focussed using Liveview,
I don't think I have seen that much PF from a kit lens before, but it has been a while since I owned one.
My purple monster is a Tamron LD Di 70-300 but it focuses quick, is sharp and I can easily fix the PF in Lightroom.
07-10-2013, 04:13 PM   #29
Pentaxian
carrrlangas's Avatar

Join Date: Mar 2012
Location: Cottbus (Berlin)
Photos: Gallery
Posts: 1,738
QuoteOriginally posted by rawr Quote
Not necessarily true. As you no doubt are aware, there are some in-camera options that will only work (as in produce an effect) on JPEG's in the K-5 and other cameras. If the purple fringing is being generated or influenced by these JPEG settings, then the RAWs may not have the issue.
I donīt understand what you are traying to say but the PF shown on the RAW file canīt be either generated or influenced by the JPEG settings. At most, the jpeg engine may have a function to deal with PF but thatīs it. (probably LR does it better anyway)
07-10-2013, 05:11 PM   #30
Pentaxian




Join Date: Jun 2009
Location: GMT +10
Photos: Albums
Posts: 10,600
QuoteOriginally posted by carrrlangas Quote
I donīt understand what you are traying to say
The in-camera JPG settings include:

- white balance,
- CA correction (with supported lenses)
- optical corrections (with supported lenses)
- highlight and shadow correction
- noise reduction
- image size and quality
- via the Custom Image settings, image sharpness, saturation, contrast etc
- via the Digital Filters menu, a variety of effects, including High Contrast
- via the Cross-Processing menu, a variety of effects.
- via the HDR menu, a variety of in-camera effects.

Hence what I was trying to say is that if, for example, when shooting JPG's the user was operating with some Custom Image adjustments that over-boosted certain colours or contrast levels, or had his white balance set very strangely, or had his camera stuck on a Digital Filter mode (like High Contrast or Extract Colour), you certainly might expect weird things to happen, purple fringing included.
Reply

Bookmarks
  • Submit Thread to Facebook Facebook
  • Submit Thread to Twitter Twitter
  • Submit Thread to Digg Digg
Tags - Make this thread easier to find by adding keywords to it!
camera, dslr, k-5, k-5 ii, k-5 iis, k5, lens, pentax, pentax k-5, re, shots, tamron, trees
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
Pentax: long ago and now again Vektor Welcomes and Introductions 2 03-08-2012 09:18 AM
Pentax and tethered shooting, again. SveinG Pentax DSLR Discussion 12 12-26-2011 04:53 AM
Hi again...and kudos emulsify Welcomes and Introductions 3 12-04-2011 12:09 AM
Cold fusion (again) and politics jeffkrol General Talk 4 07-21-2011 11:52 PM
What do you shoot again and again? betsypdx Post Your Photos! 45 03-04-2008 05:37 PM



All times are GMT -7. The time now is 08:37 AM. | See also: NikonForums.com, part of our network of photo forums!
  • Red (Default)
  • Green
  • Gray
  • Dark
  • Dark Yellow
  • Dark Blue
  • Old Red
  • Old Green
  • Old Gray
  • Dial-Up Style
Hello! It's great to see you back on the forum! Have you considered joining the community?
register
Creating a FREE ACCOUNT takes under a minute, removes ads, and lets you post! [Dismiss]
Top