Cyan Edges

Discussion in 'Nikon' started by kenneth_smith|7, Feb 9, 2014.

  1. This is not a Nikon optic, but the people here know a thing or two about lense, so I thought I'd ask here. This is the Tamron 12-24 which is supposed to be as good as the twice the price Nikon. I'm not so sure, and in fact this may be one of the reasons for the Nikon. I have cyan edges. Is this chromatic or vignetting?
    Any way to correct it? Does the Nikon do this? Sigh. I've been chasing my tail lately.
  2. can't see it ... maybe a crop of the image would show it?
  3. Could be an automatic vignette from your post processing software. If there's an option to turn it off, try it.
  4. Sounds like chromatic aberration (purple fringing), but the image is too small for us to see anything of importance. If it is, it is easily corrected in PS, NX2, et al.
  5. Sorry I thought this would be large enough to see. I see it clearly. It might be something chromatic, but it's not fringing. This is the entire edge/ sides of the photograph, not along a contrast edge. There appears in the center of the entire photo a slight redishness, albeit slight red cast to the white snow. Then to the edges left and right, top and bottom, it begins to go cyan. I tried NX2 chromatic aberration corrections to no effect. I also looked at many snow scenes with Nikon lenses, and saw nothing like this.
  6. [[I tried NX2 chromatic aberration corrections to no effect.]]

    [[not along a contrast edge.]]

    These two statements are obviously related.
  7. The two statements are related but not to the photograph sample. The sample does not have chromatic aberration as we normally encounter it, however I tried the NX2 correction anyway, just to see if there was an effect. I also am wondering if there is another way for chromatic problems to manifest. This is an extreme wide angle lens, so possibly the center is one color and the edges another.
  8. Oh well, if its an extreme wide angle, its natural light falloff, corrected by a center filter commonly used in super wide fixed lenses. Filter threads vary, but if its important enough try digging up Schneider optic product catalogues, they have an array of center filters. It's cool to see what their doing anyway.
  9. I think it's vignetting causing the corners, and to a lesser extent, the edges to be darker. This has an effect of increasing the saturation of colour in these areas.
    As the WB could be a tad 'Cold' (!) and the sky is pale blue the result is bluer corners.
    I'm not so sure I really see the red. Maybe that's more of a human complementary-colour response? An area that seems less blue/cyan is thus more red?
    Which camera body are you using? and is it a RAW conversion?
  10. Thanks Don. The filter cost would probably be more than the lens . I wonder if the Nikon 12-24 does this?
    The camera used is a D7000, with a raw conversion. I like the idea of color response for the red. Although all whites, just like all blacks have some color tone in them. I know from making C-Prints for years that there is always a vestige of some color in white. It's when it varies like this that it becomes a girl dog to print. If I printed one of these myself I'd have to wave a little taste of red over all corners, all at the same time.
    Is there a plug in for selective color dodging? I don't have PhotoShop CS. I have Capture NX2, and use an older Photoshop 7 for touch ups.
  11. I'm afraid I can't see the red centre/cyan corner effect either Kenneth. All I'm seeing is the expected vignetting that you're always going to get from using an ultrawide lens.
    Are you sure the colour variation isn't an artefact of your monitor? Nearly all LCD monitors show a colour and/or brightness change away from the point where your eyes are centred. You need a CRT, IPS or S-IPS type monitor to overcome the effect, and the latter two don't come cheap.
    Chromatic aberrations, BTW, don't manifest as large areas of different colour. They're limited to small fringes around high contrast detail in the image, and can usually only be seen at close to 100% pixel-peeping magnifications.
  12. If you do excessive and painful adjustments to the saturation and contrast, you get this. It clearly shows the vignette of the circular lens coverage and the effect of trying to force an overall colour balance, producing reddish..pinkish/cyan colour nastiness!
    Interestingly, it's not entirely central, symmetrically speaking??
  13. Thanks for the vindication Mike. I have a refined taste in finding things gone wrong, and often it's unconvincing to others. This was a great idea. The unevenness also discourages the filter idea.
    Now what? Just live with it I guess. I'll use the lens as best I can, and maybe avoid the snow scenes. I shoot too many of them anyway.
    Still wondering if the Nikon lense has this problem. Anybody use the Nikkor 12-24?
  14. Throw a little more subject at it and it's not really noticeable. If the Nikon build quality were higher I might have held out for it, but when you read it's a bit plasticky and kinda cheap feeling, you don't exactly get too excited. I for one don't mind a little heft in a lens. Go ahead with the metal next time Nikon.
  15. Ken,
    Don't forget the Vignetting adjustment under Camera & Lens correction. Also add some Highlight protection under Quick Fix, then adjust for exposure, curves, etc
  16. Thanks Joe. I will give that a try.
  17. Just for the record, this was the Tokina. I said Tamron throughout, but it's the highly rated Tokina. As for the Nikon, I managed some feedback elsewhere from a user who went out in the snow with his and reported good even color. That's why it's more expensive. I kinda wish I'd stuck it out and paid the extra, but lesson learned. In reading a dozen tests over a period of a week, no one mentioned this. They warned of more chromatic aberration, but nary a word about uneven color overall. Point Nikon.
  18. Looks like ordinary light falloff, common to most wide angle lenses, combined with white balance problems.
    Winter, snow, and early morning/late afternoon-evening lighting all conspire to confound most efforts at one-shot white balance or color correction. Ultra-wide angle lenses and polarized light in early morning and late afternoon exacerbate the challenges.
    You may be seeing some slight reinforcement of spectral sensitivity due to differences in lens coatings. But it doesn't look like anything that can't be solved with editing.
    Lightroom and DxO offer quick fixes to light falloff/vignetting, and custom white balances. And Lightroom can be customized for lenses not already in the existing database. Sometimes a different camera profile can resolve a problem that appears to be due to a lens.
  19. This is simple vignetting combined with poor color balance. It is easily corrected in PS CS6 and other software (?). Forget any filters. The lens you are using is excellent, but it is noted for vignetting (if not corrected with software).
  20. It is an excellent lens and I'll keep it. No sense running to the Nikon lens when the issue is vignetting, and the Nikon might do that as well. Although my other tester reported evenness, no pictures were sent. I'd like to see it for myself.
    But as you say, software is the fix for this lens. The two color problem becomes one. Unless there remains a lens coating issue. There was a second version of this lens that received a better coating, but that was probably for CA, and maybe flare. I've had no unexpected flare problems with mine though. Quite impressive there.
    Thanks to you both for your response.
  21. It IS going cyan at the edges. You may not notice the shift in 98% of the photos you shoot. I had the Nikon 12-24 on a D200 and D2 and never had a color shift at the edges.
    Interestingly, my Zeiss ZF 21/2.8 has the cyan shift when used on a Nikon D3, but not(thankfully) on my D800.
    There is unfortunately no substitute for doing your own testing if/when you get a chance to shoot with a Nikon 12-24 or 10-24.
  22. Yes it is, but I think the consensus was that the darkening of vignetting itself was the culprit, and that the coating of the lens was not causing the cyan color. It appears the Nikon doesn't have the vignetting, therefor doesn't have the color shift. I think.
  23. Coatings may be a factor. Some of my older third party lenses show slight color shifts compared with my Nikkors. The Tamron Adaptall 24/2.5 is slightly warm (using daylight white balance as a reference point), while the Vivitar Series 1 70-210/2.8-4 is slightly cool. Both are easily corrected in editing.
    But edges and corners that are relatively cooler compared with a center that is relatively warmer would be more of a challenge to one-shot fixes, including custom WB. It would be interesting to see whether a custom lens correction in Lightroom or other program would fix the color shift as well as the light falloff. Lightroom also offers custom color balance sliders in addition to the WB presets, but I've seldom used them.
  24. I decided against LightRoom recently because I thought very highly of Capture, especially it's handling of NEF color. I also don't work in the high volume that LightRoom was specifically designed for.
    However Capture has sat on the sidelines now for a good while and this very kind of lens corrective that you speak of has been developed by LR and Dx0. That and many other subtle and advanced picture editing tools that I discovered using their 30 day trials. Actually LR is very reasonably priced. I suppose I could throw it in the mix.
    Even Captures vignetting correction is mild at best. But I'm very much in love with the U-Point technology of Capture. As far as this lens goes, I will just have to modify it the best I can with other tools. A color burn and dodging tool would certainly help. Maybe Layers. I'm not very well versed in post processing. I'd almost rather just avoid the problem, and shoot what it can shoot without showing it's weakness. In my particular case this lens was just a luxury item. I'm not generally a wide angle shooter, although I'm liking it more than I expected.

Share This Page