G
Guest
Guest
Here's mine.
Photoshop Gurus Forum
Welcome to Photoshop Gurus forum. Register a free account today to become a member! It's completely free. Once signed in, you'll enjoy an ad-free experience and be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox!
CORRECTION FOR FLOURESCENT LIGHTING
- Thread starter Gadgetgal
- Start date
! Your step-by-step pictorial really helps to see how you achieved same! Here's my second try. I used Erik and Wbiss methods combined. I also tweaked levels and curves on individual colors, sharpened, and applied AutoFX Dream filter just slightly. My background changed colors a bit, but there's not as much yellow.
G-Man, I had not seen your post when I did this version. I am definitely going to try yours tonight. It looks really good. The step-by-step will help. Thanks for posting.
I know I am not very good at this . . . YET!! I am determined if it takes years I am going to be good. Thanks to each of you for sharing and teaching. I will post round three tomorrow.
G-Man, I had not seen your post when I did this version. I am definitely going to try yours tonight. It looks really good. The step-by-step will help. Thanks for posting.
I know I am not very good at this . . . YET!! I am determined if it takes years I am going to be good. Thanks to each of you for sharing and teaching. I will post round three tomorrow.
G
Guest
Guest
The fastest way to remove the yellow light; open the original and go to Image/Adjustments/Selective Color...
Use the settings as shown in the attachment.
Use the settings as shown in the attachment.
It is indeed the fastest method. But, just like changing the contrast is less subtile than levels which, in their turn are less subtile than curves, this selective colour is less subtile than levels and curves. Just think how many "yellows" there are, and how PS most certainly has some preset yellow that can, but does not have to correspond with the typical yellow we are now trying to eliminate.
I would not go to deep into the LAB method, but believe me: it offers more subtilities than any other one. And this is not because I posted it. One can also change the lightness and the red-green channel, and choose for the yellow any possible yellow a comp can reproduce.
Just try it out before you discard it. LAB is incredibly versatile.
I would not go to deep into the LAB method, but believe me: it offers more subtilities than any other one. And this is not because I posted it. One can also change the lightness and the red-green channel, and choose for the yellow any possible yellow a comp can reproduce.
Just try it out before you discard it. LAB is incredibly versatile.
G
Guest
Guest
Ok, that is all nice in theory Erik, but the fact is that but no-one seems to be able either to use it and show results. Why? Because it's too difficult too use. Yes, I agree, in professional hands they WILL deliver the best results, but not in the hands of the average photoshopper. Just do some research among photoshoppers to find out how many are able to do colour corrections by the numbers...not a lot I tell you.
It's in my opinion a myth that everything should be done with curvers and different color modes either.
Some people always yell "curves!" when levels will do just fine and that's the same with a lot of other tools. The writers of the book Photoshop Artistry 6 show in detail how it's the combination of those different tools that deliver the best results. In this photograph we're dealing with a global yellowish light, so selective color will do just fine, in my opinion that is.
If someone can show Gadgetgal a faster result than using my method, go ahead, I will would like to see it
Until now I have mostly seen retouched picures that are either too blue or too flat.
Just my 2 cents
It's in my opinion a myth that everything should be done with curvers and different color modes either.
Some people always yell "curves!" when levels will do just fine and that's the same with a lot of other tools. The writers of the book Photoshop Artistry 6 show in detail how it's the combination of those different tools that deliver the best results. In this photograph we're dealing with a global yellowish light, so selective color will do just fine, in my opinion that is.
If someone can show Gadgetgal a faster result than using my method, go ahead, I will would like to see it
Until now I have mostly seen retouched picures that are either too blue or too flat.
Just my 2 cents
Like I wrote: your method is the fastest. But you are mixing things up. There is no way to say something is too blue because you have no point of reference. Only when the subject/model is sitting there you have a reference. And even then it depends on the light.
When I wrote that one cannot do colour corrections by numbers, I also reacted to the idea of using cmyk. (see thread in Gare's forum)
Curves and levels are different as I do not work with "this hue should be x more Red, Y more green and Z more blue", no: what I do is create myself a palette that offers me, in this case, 256 different shades of yellow because I know that no colour cast is limited to one specific hue of yellow. And I do not like generalisations like Photoshop does:"let's assume that there is one yellow...". If I restore a painting, I never can use that one single yellow straight out of the tube. I need a palette of yellow hues. And that is what curves offers me.
As for using lab: it encompasses all hues that can ever be displayed, and it has some great channels. In the first posting I only wanted to indicate a possible road. A road I always follow myself and one that I trust because it's wide and offers more options than any other I've ever tried out.
Of course, besides the b channel also the a channel and the lightness can be used. And do tell me: which other colour space allows you to see the greys separated from the colours? Which one is completely device-independant?
Yet finally, the result will always be a personal interpretation. And it has never been my intention to brag that my interpretation was, is or ever will be the best. In fact it's fascinating to see how many different "natural" skin colours can be seen in this thread. We all interpret differently, like I already said in the color correction by numbers thread. And if you're satisfied with one yellow as Photoshop's architects chose for you, go ahead. But do know that every skin colour needs yellow, and that you remove that also. On most monitors this will may indeed make little or no difference. Most aren't well calibrated. On a desktjet printer it matters even less as perhaps only two percent give exactly the colours that were seen on the monitor. And most people don't even know that they're interpreting, not even when they say: "this looks very natural to me." So they may not even see it.
People are free to use whatever method they choose. No problem there. But don't mix using curves in lab with using numbers in cmyk.
As for the argument that the method is too difficult for so called "amateurs": these forums are put online for people to learn and practise. Results do not have to be professional, but the better methods can be learned from the start, which, in due time, will become a great time saver, and give a better insight in quality. Which is all that counts.
Yet: speaking about numbers: there is indeed one reference we can look at: the white logo on his tee shirt. Perhaps that one could be approximately set to 255,255,255?
But yes: I need three minutes to get a result that pleases my eye. That extra time is, in my opinion, not wasted.
When I wrote that one cannot do colour corrections by numbers, I also reacted to the idea of using cmyk. (see thread in Gare's forum)
Curves and levels are different as I do not work with "this hue should be x more Red, Y more green and Z more blue", no: what I do is create myself a palette that offers me, in this case, 256 different shades of yellow because I know that no colour cast is limited to one specific hue of yellow. And I do not like generalisations like Photoshop does:"let's assume that there is one yellow...". If I restore a painting, I never can use that one single yellow straight out of the tube. I need a palette of yellow hues. And that is what curves offers me.
As for using lab: it encompasses all hues that can ever be displayed, and it has some great channels. In the first posting I only wanted to indicate a possible road. A road I always follow myself and one that I trust because it's wide and offers more options than any other I've ever tried out.
Of course, besides the b channel also the a channel and the lightness can be used. And do tell me: which other colour space allows you to see the greys separated from the colours? Which one is completely device-independant?
Yet finally, the result will always be a personal interpretation. And it has never been my intention to brag that my interpretation was, is or ever will be the best. In fact it's fascinating to see how many different "natural" skin colours can be seen in this thread. We all interpret differently, like I already said in the color correction by numbers thread. And if you're satisfied with one yellow as Photoshop's architects chose for you, go ahead. But do know that every skin colour needs yellow, and that you remove that also. On most monitors this will may indeed make little or no difference. Most aren't well calibrated. On a desktjet printer it matters even less as perhaps only two percent give exactly the colours that were seen on the monitor. And most people don't even know that they're interpreting, not even when they say: "this looks very natural to me." So they may not even see it.
People are free to use whatever method they choose. No problem there. But don't mix using curves in lab with using numbers in cmyk.
As for the argument that the method is too difficult for so called "amateurs": these forums are put online for people to learn and practise. Results do not have to be professional, but the better methods can be learned from the start, which, in due time, will become a great time saver, and give a better insight in quality. Which is all that counts.
Yet: speaking about numbers: there is indeed one reference we can look at: the white logo on his tee shirt. Perhaps that one could be approximately set to 255,255,255?
But yes: I need three minutes to get a result that pleases my eye. That extra time is, in my opinion, not wasted.
My thanks to Erik
I have to admit I do not know enough about color correction to even fully understand the discussion...I found this forum by google because I wanted to color correct a photo. I am glad I found it! I will certainly learn a lot here.
I just wanted to than Erik for his post, the step by step was clear and allowed me to correct a very yellowish flourescent photo I took recently. It looks great to my eye and it only took me a few minutes to do.
Thanks!
I have to admit I do not know enough about color correction to even fully understand the discussion...I found this forum by google because I wanted to color correct a photo. I am glad I found it! I will certainly learn a lot here.
I just wanted to than Erik for his post, the step by step was clear and allowed me to correct a very yellowish flourescent photo I took recently. It looks great to my eye and it only took me a few minutes to do.
Thanks!