Clive Murray-White Posted January 9, 2011 Share #1 Posted January 9, 2011 Advertisement (gone after registration) I think I can see a pinkness to some parts of pictures I take with my Nokton 35/1.2, this is an example. Skin tones mainly. Am I imagining it? Lighting in the room was from halogen spots and to me the pink exactly matches the B+W IRcut/Uv filter colour. Auto WB What can be done to fix it, I have Lr and Photoshop? Thanks - Clive Welcome, dear visitor! As registered member you'd see an image here… Simply register for free here – We are always happy to welcome new members! Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! ' data-webShareUrl='https://www.l-camera-forum.com/topic/141314-pink-nokton-3512-iruv-filter/?do=findComment&comment=1553336'>More sharing options...
Advertisement Posted January 9, 2011 Posted January 9, 2011 Hi Clive Murray-White, Take a look here pink? Nokton 35/1.2 + IR/UV filter. I'm sure you'll find what you were looking for!
farnz Posted January 10, 2011 Share #2 Posted January 10, 2011 Clive, That looks to me like a little bit of IR 'bleed through'. There is a school of thought that the B+W, Heliopan and other brand IR cut filter have slightly different performance and may let some IR through. I note that the picture frame appears to be a good white so it's not a general magenta cast and it mostly affects skin, which is notoriously IR-prone owing to its reflection off the blood vessels just below the surface. I would fix it by pushing the A curve slightly towards green in LAB colour in Photoshop (unfortunately RGB or CMYK won't produce a similar result). If you're familiar with working in LAB it would be reasonably straightforward but if not I recommend reading The Canyon Conundrum by Dan Margulis although it could be a long and challenging road. Pete. Link to post Share on other sites More sharing options...
Clive Murray-White Posted January 10, 2011 Author Share #3 Posted January 10, 2011 Thanks Pete, No LAB colour hasn't been something that I've needed before - so I'd better go off and do some studying - Thanks, Clive Link to post Share on other sites More sharing options...
RichardM8 Posted January 10, 2011 Share #4 Posted January 10, 2011 Maybe I look at this too simple but I think a light touch of the 'Tint' slider in LR will get you a long way Clive... If you don't mind I had a quick go in LR with it. It's a small jpg so not much leeway to edit. Without knowing the scene it's hard - if not impossible - to judge the tones etc. I played a bit with Tint, Clarity and Sharpness. Richard. Link to post Share on other sites More sharing options...
Clive Murray-White Posted January 10, 2011 Author Share #5 Posted January 10, 2011 G'day Richard, nice to hear from you again - no I don't mind you having a go at it - it all helps - as you can see the picture itself is no masterpiece but it does show the issue pretty well. I'll do my LAB colour homework but ........... does anyone have some advice about setting the camera up prior to taking shots in this kind of environment - like... would a grey card make much difference? I can "work" the room out quite easily prior to taking pictures because its in my house. Thanks again Clive Link to post Share on other sites More sharing options...
lct Posted January 10, 2011 Share #6 Posted January 10, 2011 I would try red saturation -25 or -30. Link to post Share on other sites More sharing options...
farnz Posted January 10, 2011 Share #7 Posted January 10, 2011 Advertisement (gone after registration) Clive, I hope you don't mind but I've given your shot 2 treatments to show some differences. The 1st is the original shot, the 2nd is with the A curve pulled towards the green in LAB, and the 3rd is, at LCT's suggestion, the red channel desaturated in RGB - only I took it to -40 instead of -25 or -30. At first, 2 looks a bit too green until your eye compensates after looking at the magenta original (it helps to take the original off the screen momentarily). Then you can see that the watermelon is about the right red, the hand holding the knife, the auburn-haired girl's temple, the girl with glasses right shoulder etc are all about the right colour (imo:rolleyes:) In 3, the skin still retains a slight magenta hue, the watermelon is way off and the grass in the pictures on the back wall looks unnatural. I'm no expert and not everyone will agree with the colour balance I've produced but it looks pretty good on my screen (which is a colour calibrated, bog standard Samsung 19" flat screen:o). Pete. 1. Original shot Welcome, dear visitor! As registered member you'd see an image here… Simply register for free here – We are always happy to welcome new members! 2. A curve pulled towards green in LAB 3. Red channel desaturated to -40 in RGB Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! 2. A curve pulled towards green in LAB 3. Red channel desaturated to -40 in RGB ' data-webShareUrl='https://www.l-camera-forum.com/topic/141314-pink-nokton-3512-iruv-filter/?do=findComment&comment=1554238'>More sharing options...
Clive Murray-White Posted January 10, 2011 Author Share #8 Posted January 10, 2011 Thanks Pete, I do appreciate all this help with examples so I can actually see the differences of each approach - its turned out to be quite an instructional picture, water melon (in paint colours would need a crimson) and tomatoes (a cadmium red), painted wall is like a yellow ochre cream and exposed wall is pretty much straight yellow ochre (maybe with a bit of warmth in it) and every approach winds up with a trade-off of some kind ie get skin tones more right and the wall goes too green or the beer bottle looses its colour and so on. My current thoughts are to experiment with the lights in the room, there are currently 5 x 120 watt old style hot halogen spots, mainly pointing at the walls but one in the middle of the table, as all lights in Australia have to be energy savers and all older style lights are being phased out I'll have to find a low energy equivalent soon. I personally don't like the light quality of almost all the new lights, cold and miserable by comparison. For skin tones your LAB colour is the clear winner. My guess is that 5 hot spots all around 7-8 feet away from skin is always going to cause the M8 a few IR problems. Other cameras would have as bad a time but probalbly render everything too orange or blue. Clive Link to post Share on other sites More sharing options...
Clive Murray-White Posted January 10, 2011 Author Share #9 Posted January 10, 2011 Had a go with LAB colour + my other tools - much happier now, the original looks pretty right but crunched for posting it did take quite a battering. Skin tones may look at bit odd to some people but both the men shown have quite red faces, the pale skin woman far left is white/pink and the girl with glasses has a very nice tan. Even the bald guy at the back of the room got his neck sunburnt in that afternoon! Clive Welcome, dear visitor! As registered member you'd see an image here… Simply register for free here – We are always happy to welcome new members! Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! ' data-webShareUrl='https://www.l-camera-forum.com/topic/141314-pink-nokton-3512-iruv-filter/?do=findComment&comment=1554380'>More sharing options...
farnz Posted January 10, 2011 Share #10 Posted January 10, 2011 Clive, The skin tones in this version look pretty good. Pete. Link to post Share on other sites More sharing options...
Clive Murray-White Posted January 10, 2011 Author Share #11 Posted January 10, 2011 Pete: Without your help and the introduction to LAB colour I wouldn't have been able to do it, so - thank you very much indeed. Just on the finer points, I switched it back to RGB after fiddling with it in LAB, do you reckon that's the way to go? Thanks again - Clive Link to post Share on other sites More sharing options...
farnz Posted January 10, 2011 Share #12 Posted January 10, 2011 Yes definitely, Clive, because other software won't recognise LAB. Also there's no penalty for switching between LAB and RGB, ie no loss of data. Pete. Link to post Share on other sites More sharing options...
wstotler Posted January 11, 2011 Share #13 Posted January 11, 2011 My current thoughts are to experiment with the lights in the room, there are currently 5 x 120 watt old style hot halogen spots, mainly pointing at the walls but one in the middle of the table, as all lights in Australia have to be energy savers and all older style lights are being phased out I'll have to find a low energy equivalent soon. Clive, I've shot extensively with the Nokton and a 486 B+W filter. Your photo shows a WB problem rather than "tinting contamination" introduced by the filter. E.g., it's not the filter tearing up your skintones. It's WB being off for the scene. Future Fix * Try setting the WB in your M8 manually to Tungsten. * Shoot RAW. * (optional) Snap a shot of the greycard in the trouble light, if you have a greycard. Like the WhiBal G7 Pocket Card. * Adjust WB in post by sampling your greycard (and then using that value for all shots) OR just sample known grey/netural color spots in the photo. E.g., sample from the plates' shadows, white shirt shadows, etc. There shouldn't be a need to do anything other than adjust the WB on the original RAW image to "fix" this shot--sampling from a known (or close to known) neutral gray--during the RAW converter stage. Thanks, Will P.S. If you upload a RAW image version of this image on Sendspace and post the link so it can be downloaded, I'll have a look. Link to post Share on other sites More sharing options...
jaapv Posted January 11, 2011 Share #14 Posted January 11, 2011 It does indeed look like horrible light - but energysavers will be worse.. The only real remedy I found is to set a manual WB by shhoting a sheet of white paper and warming up slightly in postprocessing. Link to post Share on other sites More sharing options...
WestMichigan Posted January 11, 2011 Share #15 Posted January 11, 2011 I want to Chime In to second the others in saying you've got a Color Balancing issue. You've received some great input on the color correcting side of things, but -in the end- those tools are presciptive remedies to fix the image you & your camera came away with. The simpler route is to have a preventative remedy as a part of your photographer's toolkit in use ie: Setting a Custom Color Balance. The where-as-its and how-to-its in the Leica M world aren't part of my skill set - Yet! Perhaps other members might offer input on it, if you request. Sincerely, Richard in Michigan Link to post Share on other sites More sharing options...
adan Posted January 11, 2011 Share #16 Posted January 11, 2011 This is not ONLY a WB problem (although that is always a question). It is also a profiling problem. Tungsten or Halogen lights don't put out much blue light. To correct neutrals, therefore, one is adding a lot of blue to the image globally - regardless of whether one is using AWB, or a custon WB, or correcting the WB in raw development.. Strong reds, however (watermelon, tomatoes), don't contain much blue, so they reproduce more or less correctly under tungsten light (WITHOUT WB correction) and the blue added when correcting the overall WB turns reds pink/magenta (and also desaturates them a bit - blue being a complementary color to the yellow in the reds, so blue + yellow = gray or neutral = less saturated). Ideally you should make a personal camera profile specific for tungsten (or any yellowish) light, using a MacBeth ColorChecker or similar target. In most cases this ends up shifting the red hue more towards yellow than magenta compared to a daylight/full-spectrum profile, among other changes. Link to post Share on other sites More sharing options...
wstotler Posted January 11, 2011 Share #17 Posted January 11, 2011 This is not ONLY a WB problem (although that is always a question). Most respectfully, his reported (and visible) "reddish cast" has a WB problem at its root--95% worth. More below. The main reason I commented is that what has happened in his image is not some strange "IR" problem causing the reddish cast with this filter and lens. It's WB. Adjusting the WB on the RAW image by sampling a middle grey tone in post will correct the issue *almost* fully. Enough, anyway, that it won't be a noticeable issue. The image will be "fixed enough" that we're into calibrated monitors and ideal proofing conditions to be able to positively spot the trouble. And playing with color once it's been balanced "enough" becomes subjective, really--the camera says one thing, the monitor another, what I think about the color is something else, somebody else thinks something else. Etc. But the main point is that after WB the images that the photos aren't "noticeably reddish" any longer. Ideally you should make a personal camera profile specific for tungsten (or any yellowish) light, using a MacBeth ColorChecker or similar target. In most cases this ends up shifting the red hue more towards yellow than magenta compared to a daylight/full-spectrum profile, among other changes. I would say that this level of rigor is commendable and, yes, "ideal." For most, I think a simple WB adjustment in post on the RAW will suffice. Shooting a card in the light to ease adjusting the WB in post will take it one level further. Your solution goes the whole way in terms of getting to "accurate" color. I'm of the mind that "accurate" color is fine as a goal, but I'm also completely OK with color I can perceive is "good" or adds to the scene--the "reddish cast" does not fall into an "OK" category. But if everything in this photo was pushing "warm" but not pushed "red"? Eh. I'm OK with that. But the "reddish cast" has to go. And, in this case, setting WB will fix that problem. Thanks, Will Link to post Share on other sites More sharing options...
Clive Murray-White Posted January 11, 2011 Author Share #18 Posted January 11, 2011 Its just as well I enjoy the detective work! Re WB, white plates, shirts, picture mounts all looked within the acceptable range to me straight out of the camera, I do of course know that none may have been "white card" white - this is also confirmed by using white eye dropper in Lr. But when I get a spare evening I'll set up up a whole lot of stuff in the room - and see what the WB differences may be between Auto, Tungsten and white card custom. Strong reds, however (watermelon, tomatoes), don't contain much blue, so they reproduce more or less correctly under tungsten light (WITHOUT WB correction) and the blue added when correcting the overall WB turns reds pink/magenta (and also desaturates them a bit - blue being a complementary color to the yellow in the reds, so blue + yellow = gray or neutral = less saturated). Adan This reminds me of one of the more difficult concepts to get across to first year art students in regard to mixing colour and why in your set of paints you had to have 2 blues, 2 reds and 2 yellows and if you "accidentally" mixed the blue best suited to making purple and the yellow suited for making orange you get a pretty muddy green. Of course the poor old camera system can't just grab the right tube of blue!!! it would make life so much easier! Thanks again Clive Link to post Share on other sites More sharing options...
wstotler Posted January 11, 2011 Share #19 Posted January 11, 2011 Its just as well I enjoy the detective work! Re WB, white plates, shirts, picture mounts all looked within the acceptable range to me straight out of the camera, I do of course know that none may have been "white card" white - this is also confirmed by using white eye dropper in Lr. Clive--there appears to be a bit of confusion in your post. If there's not, disregard the remainder of this post. Thanks. You're seeking an 18%ish grey value when you white balance using the WB dropper in Lightroom (or Aperture), not white. - Set your WB pulldown to "as shot." - Then, using the dropper, select a "middle grey" color. I suggested shadows on white surfaces as a quickie fix. This is not ideal, but works very well in many cases. You can then *gently* play with the Temp and Tint sliders to make fine adjustments. More here with the technique: How to Correct White Balance Tutorial Part One: Lightroom 2 | neutralday Cheers, Will Link to post Share on other sites More sharing options...
Clive Murray-White Posted January 11, 2011 Author Share #20 Posted January 11, 2011 Thanks Will, I've just moved over to Lightroom and my previous, either Bibble 5 or Olympus Studio - can't remember which, had a very useful "click white" option - I just assumed the same for Lr. Thanks again Clive - you never stop learning do you? Link to post Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.