It was a bright sunny day today, on the way in to MacWorld, and I was thinking about neutrals as the brightly lit buildings of city of Oakland flashed by. Although I know that the sky is blue, and that shadows are lit by the sky, for the life of me I could not see a hint of blue tint in any of the shadows.
Every single shadow was perfect gray, without a hint of blue. It occurred to me that this is one way of looking at the function of a neutral in an image: we want to duplicate the effect that we see when we are there looking at the scene. We should not, normally, allow the the shadows to appear blue in our photographs.
Another way to look at this is that we should try to recreate the color of the object in the image, and disregard the color of the light falling on the object. There are images where we want to convey a certain mood, or where the light itself is part of the subject (sunsets, for example), but the bedrock of color correcting an image is to compensate for the effects of colored lighting, camera processing, windows, or any other thing that may add a color cast ot the image.
I'd be interested in others' take on this - the concept of a neutral, and of removing a color cast, is so important that I think it's worth looking at different ways to describe it.
Taking BART to MacWorld, thinking about neutrals
Up here in Minnesota, covered by snow 4 months out of the year....I can tell you that shadows often appear to be blue to the eye especially on snow. That being said...
I was always taught that neutral was 18% grey.
That colored light "will" created colored shadows.
That White and Black can function as neutrals. (White Highlights are over exposed neutrals..and Blacks are under exposed...)
My working answer to the "shadows are neutral" is that if they are not created by a colored light source they would be neutral. Example: shadow falling on a painted surface. If the shadow falls on a neutral then the shadow would be neutral if the light source was white light. Another example would be a white ball lit by both a Magenta and Green light from 45 degrees. the ball would appear white...but the shadows would be Magenta and Green depending on what angle you look at the ball from. (Additive property of light creates a neutral when complementary colors shine on the same object with the same intensity) ...but let's say I had a incandescent light.
If it lights the room and it creates a shadow on a white wall that is visable. The shadow would be shifted to the tungsten color we all see. That would in fact be correct for that light source and shadow.
I think your original assumption about neutrals is good. We try to make sense of the image we have captured. If the lighting is neutral and we see a color cast we "know" it's not right. We would then start looking for a way to "fix" the visual problem we have in front of us. Pink hues in the clouds on a sunny day...A greenish tint to a duck on a rock...all point to problems.
When I was in photo school and we were color printing by hand, color analysis was visual...we were told not to wear colored sunglasses on days when we needed to color print...color perception changes very slowly after you have been exposed to a bright light source, if you filter the source with say an amber sunglass lens all your printing that day would be as much as 5C or 5B off. The analog to that today is the monitor. There are many dependencies on the monitor for color correction and perception. Room lighting, monitor calibration...daylight falling on the monitor...all of these can change the way we see the colors before us...
My suggestion would be to visually learn to spot color shifts. (Old School)
One way would be to take an image like a portrait with a color checker square and color correct it so that the grey card is really grey. Then set about to messing it up. Start at a 5 point change in one color. Then 10, 20, 30 then 40. Do that for all colors CMYRGBK. Get a better visual reference. Understand what a 10 blue shift looks like...or a 10- in the K so you know visually that there is a problem. PSE gives you a good visual of this with "Variations".
Yes, the Hue clock can tell you all that, "By The Numbers" but sometimes things need to feel a certain way..the "Bear" image is a great example. Some people made the afternoon light very important, others worried about the color of the bears nose. Myself, I wanted the grass to look normal. Visually I can accept the bear as having golden tones, but the grass needed to be green for me to accept the image as "normal". We used to be alright with interpratation of the image for artistic values...I would say we need to inject some more of that over the course of our exploring this process....Green ducks and all...
Phew...long way to go to say that...I'll sit down now...
Greg
I was always taught that neutral was 18% grey.
That colored light "will" created colored shadows.
That White and Black can function as neutrals. (White Highlights are over exposed neutrals..and Blacks are under exposed...)
My working answer to the "shadows are neutral" is that if they are not created by a colored light source they would be neutral. Example: shadow falling on a painted surface. If the shadow falls on a neutral then the shadow would be neutral if the light source was white light. Another example would be a white ball lit by both a Magenta and Green light from 45 degrees. the ball would appear white...but the shadows would be Magenta and Green depending on what angle you look at the ball from. (Additive property of light creates a neutral when complementary colors shine on the same object with the same intensity) ...but let's say I had a incandescent light.
If it lights the room and it creates a shadow on a white wall that is visable. The shadow would be shifted to the tungsten color we all see. That would in fact be correct for that light source and shadow.
I think your original assumption about neutrals is good. We try to make sense of the image we have captured. If the lighting is neutral and we see a color cast we "know" it's not right. We would then start looking for a way to "fix" the visual problem we have in front of us. Pink hues in the clouds on a sunny day...A greenish tint to a duck on a rock...all point to problems.
When I was in photo school and we were color printing by hand, color analysis was visual...we were told not to wear colored sunglasses on days when we needed to color print...color perception changes very slowly after you have been exposed to a bright light source, if you filter the source with say an amber sunglass lens all your printing that day would be as much as 5C or 5B off. The analog to that today is the monitor. There are many dependencies on the monitor for color correction and perception. Room lighting, monitor calibration...daylight falling on the monitor...all of these can change the way we see the colors before us...
My suggestion would be to visually learn to spot color shifts. (Old School)
One way would be to take an image like a portrait with a color checker square and color correct it so that the grey card is really grey. Then set about to messing it up. Start at a 5 point change in one color. Then 10, 20, 30 then 40. Do that for all colors CMYRGBK. Get a better visual reference. Understand what a 10 blue shift looks like...or a 10- in the K so you know visually that there is a problem. PSE gives you a good visual of this with "Variations".
Yes, the Hue clock can tell you all that, "By The Numbers" but sometimes things need to feel a certain way..the "Bear" image is a great example. Some people made the afternoon light very important, others worried about the color of the bears nose. Myself, I wanted the grass to look normal. Visually I can accept the bear as having golden tones, but the grass needed to be green for me to accept the image as "normal". We used to be alright with interpratation of the image for artistic values...I would say we need to inject some more of that over the course of our exploring this process....Green ducks and all...
Phew...long way to go to say that...I'll sit down now...
Greg
The core of Dan Margulis's philosophy, in correcting "by the numbers" is that there are some basic principles to color correction. Shadow, highlight, and neutral are so important to get right that I want to focus on getting these concepts solid, before we start to emphasize that there are exceptions to these rules.
It' like tempo in music. When starting a musical instrument, the concept of keeping an even beat is so important, that it overshadows the later importance, in a skilled musician, of varying the tempo to add mood and expression to the performance. In most cases, that same musician is well able to play the same piece with a perfectly even beat, but chooses not to do so.
In the same way, the ability to remove a color cast is very basic, and so many other concepts depend on this ability, that I'd prefer to leave the nuances of leaving a color cast in an image until later on, until after the ability to control color casts is established.
It's true that shadows and highlights are generally neutral, but they are more likely to be under or overexposed, and in any case contain less color information than a middle gray, and for this reason I recommend relying more on the midtone for setting a neutral when it is available.
To your other points about developing a sense of color, etc. I would add that, although vision is of course important for the final aesthetic judgment, to the extent that our eyes are so easily deceived, it is all the more important to have numerical methods, such as setting a neutral, to help us through the forest of color correction. Dan Margulis has said that "we are all color blind", and this is what he was referring to.
It' like tempo in music. When starting a musical instrument, the concept of keeping an even beat is so important, that it overshadows the later importance, in a skilled musician, of varying the tempo to add mood and expression to the performance. In most cases, that same musician is well able to play the same piece with a perfectly even beat, but chooses not to do so.
In the same way, the ability to remove a color cast is very basic, and so many other concepts depend on this ability, that I'd prefer to leave the nuances of leaving a color cast in an image until later on, until after the ability to control color casts is established.
It's true that shadows and highlights are generally neutral, but they are more likely to be under or overexposed, and in any case contain less color information than a middle gray, and for this reason I recommend relying more on the midtone for setting a neutral when it is available.
To your other points about developing a sense of color, etc. I would add that, although vision is of course important for the final aesthetic judgment, to the extent that our eyes are so easily deceived, it is all the more important to have numerical methods, such as setting a neutral, to help us through the forest of color correction. Dan Margulis has said that "we are all color blind", and this is what he was referring to.
-
imported_ganna
- Posts: 460
- Joined: Tue Nov 07, 2006 11:41 am
Oh I'm all for learning to crawl before I walk and walking before I run....
I just think that strictly by the numbers can make this process all lot more difficult. You end up chasing perfection where there might not be any. I guess I settle for a bit less than perfect numbers.. I understand the concept but have a real tough time "not being satisfied" with less than "correct numbers". I usually put alot of effort into the image and hate to feel less than satisfied with my choices. If I bet the image on my choices and not the numbers at least I feel better when I close the program at the end of the day...
One of the things I see, is that the difference between what I see in PS and what I see on the web is huge. Personally I correct so that my hard copy prints look right, regardless of the numbers...
I'm not sure if that philosophy belongs in a curves class but I try hard not to let it cloud what I am helping each class discover about the images...
It's 9:00 in the morning and i feel like we should have a good nut brown ale or stout between us discussing this...lol
Greg
I just think that strictly by the numbers can make this process all lot more difficult. You end up chasing perfection where there might not be any. I guess I settle for a bit less than perfect numbers.. I understand the concept but have a real tough time "not being satisfied" with less than "correct numbers". I usually put alot of effort into the image and hate to feel less than satisfied with my choices. If I bet the image on my choices and not the numbers at least I feel better when I close the program at the end of the day...
One of the things I see, is that the difference between what I see in PS and what I see on the web is huge. Personally I correct so that my hard copy prints look right, regardless of the numbers...
I'm not sure if that philosophy belongs in a curves class but I try hard not to let it cloud what I am helping each class discover about the images...
It's 9:00 in the morning and i feel like we should have a good nut brown ale or stout between us discussing this...lol
Greg
-
mikemeister_admin
- Posts: 4927
- Joined: Fri Sep 20, 2013 8:29 pm
Obviously each of us are into photography for different reasons. I have three printers, however I only have so many walls to hang prints on. I no longer print copies. When I come back from a trip with 500-1000 images I am not about to spend a whole lot of time going into minute detail to preserve every part of a image. There are times where one image may "pop" out at me and I will spend more time with it. A image first must have a good subject and second if it does not have reasonable exposure I move on.
After adjusting images I now incorporate them into slide shows. At this point when viewed on a 46 inch HD tv the results are not spectacular. I find that I make these for myself. Others viewing them might say "nice pictures" but after a few minutes its more like "nice transition".
I enjoy working on the images but to a point, otherwise I would not be in this class.
Thanks Joe s
After adjusting images I now incorporate them into slide shows. At this point when viewed on a 46 inch HD tv the results are not spectacular. I find that I make these for myself. Others viewing them might say "nice pictures" but after a few minutes its more like "nice transition".
I enjoy working on the images but to a point, otherwise I would not be in this class.
Thanks Joe s
-
mikemeister_admin
- Posts: 4927
- Joined: Fri Sep 20, 2013 8:29 pm
-
mikemeister_admin
- Posts: 4927
- Joined: Fri Sep 20, 2013 8:29 pm
Return to “January 2008 Curvemeister 101”
Who is online
Users browsing this forum: No registered users and 20 guests