Are you suggesting using an IR cutoff filter for everyday indoor photography to eliminate some of the color tweaking necessary? Or is that for special studio stuff only?
We're still weighing solutions, but it's a devil of a problem. In a group shot, you may have two or three (fair-skinned) faces go red and how do you adjust for it? Further, surface veins in the cheeks and blemishes get emphasized.
I'm producing test shots with B+W's 486 filter. From their catalog:
The UV filter 010 blocks ultraviolet radiation which can cause blur or – in color shots – blueness. It is ideal for photos in the mountains, by the sea or in areas with very clear air. Pictures become more brilliant, irritating blue haze is avoided and color reproduction remains neutral. This UV filter is also suitable for use as front lens protection.
The recommended product for digital cameras is the UV/IR blocking filter 486, a steep-flanked interference filter which additionally blocks infrared and prevents blur and color cast with IR-sensitive CCDs.
My studio lights don't produce IR and hence I don't see the problem there. It's most acute under tungsten indoor lighting and on-board flash that isn't bounced off the ceiling.
But I have seen thousands of outdoor shots with a similar problem. Let's take an example from a world-class photographer (Andy) with a world-class camera. Here's what he posted of a fair-skinned caucasian in a forum:
That shot is so far from a legal skin tone it would come straight back to us if printed. The skin at least would have to be much closer to this for any consumer to accept it:
We're still weighing solutions, but it's a devil of a problem. In a group shot, you may have two or three (fair-skinned) faces go red and how do you adjust for it? Further, surface veins in the cheeks and blemishes get emphasized.
I'm producing test shots with B+W's 486 filter. From their catalog:
The UV filter 010 blocks ultraviolet radiation which can cause blur or – in color shots – blueness. It is ideal for photos in the mountains, by the sea or in areas with very clear air. Pictures become more brilliant, irritating blue haze is avoided and color reproduction remains neutral. This UV filter is also suitable for use as front lens protection.
The recommended product for digital cameras is the UV/IR blocking filter 486, a steep-flanked interference filter which additionally blocks infrared and prevents blur and color cast with IR-sensitive CCDs.
My studio lights don't produce IR and hence I don't see the problem there. It's most acute under tungsten indoor lighting and on-board flash that isn't bounced off the ceiling.
But I have seen thousands of outdoor shots with a similar problem. Let's take an example from a world-class photographer (Andy) with a world-class camera. Here's what he posted of a fair-skinned caucasian in a forum:
That shot is so far from a legal skin tone it would come straight back to us if printed. The skin at least would have to be much closer to this for any consumer to accept it:
Wowza, I have taken 5-10,000 candid people shots with my digital and I don't think see this effect outside. This is an extreme example and if I ever see anything like this I would think someone should have set a custom white balance and they did not. Or they shot in the wrong white balance period. This might be a tad underexposed it looks to me as well. It looks like this problem could have been solved in the field by using a diifferent wb setting. I suppose if you don't have live preview you just didn't see this coming though. The only time I see any sort of this stuff is under straight on flash conditions indoors with my hand dandy ef-500 super shot straight on. I only have about 80,000 digital images under my belt, but I dare say I have not seen much, if any of this outdoors. I don't think filters are needed on this one, I think the camera had a problem with the white balance. I just hit this image with one drop of the white eydrpopper and got this:
It took me all of a couple seconds. I could do more here but I just opened the image to see what the problem was and could not resist one quick click with one of the levels eydroppers.
Yup, I am pretty sure now, with my cheap camera I do not get this sort of thing on my outdoor shots when I am properly white-balanced. I dunno on this, Baldy. I only see this with flash and other forms of indoor lighting. My question is why would anyone upload a photo like this without doing some simple corrections.
-don
0
Options
BaldyRegistered Users, Super ModeratorsPosts: 2,853moderator
edited April 14, 2005
With your adjustment, the cyan is in range at 24%. The forehead now measures 47% magenta, 41% yellow, not close to a legal skin tone so this print will come straight back to us unless the consumer chooses autocolor.
With your adjustment, the cyan is in range at 24%. The forehead now measures 47% magenta, 41% yellow, not close to a legal skin tone so this print will come straight back to us unless the consumer chooses autocolor.
phhh...my point was this is a standard color balance problem, I think.
-don
0
Options
BaldyRegistered Users, Super ModeratorsPosts: 2,853moderator
phhh...my point was this is a standard color balance problem
So your mission is to explain why the T-shirt is now white but the infrared-sensitive areas are not in a range any publication or consumer would accept.
So your mission is to explain why the T-shirt is now white but the infrared-sensitive areas are not in a range any publication or consumer would accept.
The nose measures 65% magenta 46% yellow.
No, my point is that the image you presented and corrected is not a very acceptable example of the problem described and the solution in my opinion. I think it is a bit is too yellow-greenish monotonish.
I suppose ezprints won't print this one either? If not, I don't see what good softproofing with the ezprints profile or the CMYK profile does here. Your saying what you see won't get printed. What??? No redheads or sunburn prints here at smugmug? Geesh my wife is a very white redhead for gosh sake! You telling I have to paint her green??
My monitor out of cal??? What gives here? I get prints from stuff that looks like this at walgreens and don't have a problem except the fact that they charge me too much when I pick them up.
You gonna tell me ezprints won't print something like this either, Baldy?
0
Options
BaldyRegistered Users, Super ModeratorsPosts: 2,853moderator
edited April 14, 2005
EZ Prints will print it, of course. But no magazine will accept it. Nor any consumer because it's not a skin tone that occurs in nature.
Your adjustment is on the left, below. The same man's photo is on the right, taken under studio lights that emit no infrared. Everyone here, who is looking at the man, says he doesn't look anything like the shot on the left.
If you print them, they'll turn out okay unless you choose the true color option. If the lab is faithful to their true color mantra, the left will look nuclear-radiated because the skin measures 7-20% more magenta than yellow.
EZ Prints will print it, of course. But no magazine will accept it. Nor any consumer because it's not a skin tone that occurs in nature.
Your adjustment is on the left, below. The same man's photo is on the right, taken under studio lights that emit no infrared. Everyone here, who is looking at the man, says he doesn't look anything like the shot on the left.
If you print them, they'll turn out okay unless you choose the true color option. If the lab is faithful to their true color mantra, the left will look nuclear-radiated because the skin measures 7-20% more magenta than yellow.
I don't buy it. The man cleary looks to have a shiney red nose when the outdoor shot was taken. He is full of stubble and has clear signs of reflection and the sun on his forehead. There is no way that gentleman's face looked anywhere near that "studio light " color out on the street. Nobody's face in the sun looks monoyellowtone like that face in the studio shot is. The object is not to try to make the man look studio yellow I don't think. I would think in this snapshot you would actually want the man to look like he did.
All Walgreen's prints pass through autocolor. I had understood you were a champion of true color?
My bad, for the record at times I use Target's free standing kiosks, not walgreens. I have a walgreens 1 block away and made a misspeak. Target's kiosks let you diasable autocolor when you do the prints yourself. I have done many on-the-spot concert shots on the kiosks. Quick 35 dollar 8x10's to band members and family type stuff.
Baldy, are you with me here? Is this monitor out of cal or something? What do those last street shots look like to you? Or anyone else? I pulled out my backup monitor and things are looking jacked up. I need to do some checking, it's possible I could be eating some words here tonight...On my backup monitor your edit looks good and mine look like *hit. So the moral of the story might be shut your face if you are not using a calibrated monitor.:flush
Consumers will not accept prints where the yellow value isn't at least 3% higher than magenta, especially when cyan is as low as 28% of the magenta value.
Consumers will not accept prints where the yellow value isn't at least 3% higher than magenta, especially when cyan is as low as 28% of the magenta value.
It's a lovely shot, however, and will print nicely with autocolor on.
Glad to see you are still up, Baldy. So you are saying the girl would not like this print? I looked at smugmug auto adjust and autotan and she turns unnaturally yellow. She was in the late may sun and had a slight reddish glow to her when I photographed her. This photo looks exactly like she did then. I am looking at 5 monitors here and she looks good on 4. I also asked nik and to his untrained eye and uncaled monitor he says she looks nice and pale. He also printed it on a simple inkjet and he said she looks just like she should after getting a little late spring sun. After a long winter indoors here in Illinois, all the fair skinned girls that don't go to the tanning booth during the winter and go straight out into the late spring or early summer sun get that nice soft reddish hue for a bit here.
Now I have not printed the girl myself but I have printed the bandshots. Did you say those bandshots would print and be acceptable, or not?
As far as the edit you did that I was jumping on you about it does look better on 4 of my 5 monitors here. Probably acceptable and I apologize. The monitor I was using here to do the edits and view your edit is acting very strange. However, it does still look a bit unnaturally yellowgreenish to me on the other monitors, probably because of the surrounding grass in the shot. It does still look better than my edits however, and I will recalibrate my ps rig later tonight.
The shot below is a little dark, but I would like to know what you think the customer would say on the fleshtones on this one along with those other two band shots I posted earlier.
Now I think I know what you are going to tell me and this should be interesting.
Once again, sorry for dogging your edit so bad, that junk monitor I was using(my main) was mucking me up. I'll fix that...
I don't buy it. The man cleary looks to have a shiney red nose when the outdoor shot was taken. He is full of stubble and has clear signs of reflection and the sun on his forehead. There is no way that gentleman's face looked anywhere near that "studio light " color out on the street.
Here is another shot of the same man in the same location, taken about 5 minutes later with another Canon. This one had an L lens with anti-infrared coating.
Here is another shot of the same man in the same location, taken about 5 minutes later with another Canon. This one had an L lens with anti-infrared coating.
The T-shirt is equally neutral.
Which camera produced the correct skin tone?
Come again, the dudes hair is green on the pic to the right!?! And the lighting conditions are not nearly the same. The first pic clearly has some bright stuff hitting his head and many other lighting differences. And you can still see the guys nose was/is reddishpurplish like I mentioned before.
-don
0
Options
BaldyRegistered Users, Super ModeratorsPosts: 2,853moderator
edited April 15, 2005
? I see a green leaf in the sky behind his head. The one on the left is the one you adjusted. His hair is being dyed on the right one.
The light is the same. One camera is pointing down (the one on the right) while the one on the left is pointing up towards the sky.
? I see a green leaf in the sky behind his head. The one on the left is the one you adjusted. His hair is being dyed on the right one.
The light is the same. One camera is pointing down (the one on the right) while the one on the left is pointing up towards the sky.
ok, I can accept the green dye as I have seen it around anyway...It's still not a good comparison because the lighting is much different. The camera either did, or should have done at least few things differently when exposing those two shots. No way that is fair comparison for much, I don't think. You can't shoot one shot into bright light and expect it to look like a shot not shot into bright light. And don't forget the edit on the left was done on a bunk monitor, I mentioned that earlier.
-don
0
Options
BaldyRegistered Users, Super ModeratorsPosts: 2,853moderator
You can't shoot one shot into bright light and expect it to look like a shot not shot into bright light.
The skin tone changes as the light behind the subject gets brighter?
— Actually, I can see how that would happen if the camera had auto white balance on. It would be influenced by the sky.
But the point remains, the customer will accept the skin tone on the right because it's in a legal range of skin tones. On the left, you have to choose autocolor for them to accept that shot.
The skin tone changes as the light behind the subject gets brighter?
The color temperature and metering information the camera was seeing and trying to capture in those two shots is quite different. The shot into the shadows is balanced nicely because it is an easy shot. The other one, with the backlighting on the top left, combined with grass on the bottom left, and the reflection and glare off of the glass on the right, and then mix that with the rest of the darker stuff on the right side of the image and in the shadows caused the camera to have a problem with the color/white balance and exposure, it looks like to me. It's a much tougher shot for the camera and/or the photographer to get correct. I say metering and white balance and the exposure differences caused this problem. That's still the way I see this one. Show me a shot with that same lens and same camera settings shot into the backlighting, then we have a real discussion here, I think.
The skin tone changes as the light behind the subject gets brighter?
— Actually, I can see how that would happen if the camera had auto white balance on. It would be influenced by the sky.
Yes, that and other reasons...and
let's quit talking about crappy photos and crappy edits for a bit
Now, if you would please, start giving me your opinion on the 3 concert shots and the 8 streetshots I posted. Will customers accept them, or do you say they are too crappy?
Thanks
0
Options
BaldyRegistered Users, Super ModeratorsPosts: 2,853moderator
The band member you posted is fine. The others are red x's.
refresh your browser once or twice, the images are there, well at pbase I mean...I gotta get the rest of my stuff off pbase, you just can't trust that place anymore...I'll start moving the links to somewhere else.
0
Options
BaldyRegistered Users, Super ModeratorsPosts: 2,853moderator
edited April 15, 2005
Red x's. PBase must not like me tonight. But you have exactly the same tools I do: the eyedropper tool and the legal range of skin tones.
Scott Kelby and Dan Margulis talk of skin tone ranges in their books if you'd rather take it from them (they list the same values we do, except Scott doesn't mention cyan, which is critical).
Red x's. PBase must not like me tonight. But you have exactly the same tools I do: the eyedropper tool and the legal range of skin tones.
Scott Kelby and Dan Margulis talk of skin tone ranges in their books if you'd rather take it from them (they list the same values we do, except Scott doesn't mention cyan, which is critical).
I moved all but one of the images to one of my servers. I have made my last direct link to pbase. I am asking you if you think these prints are acceptable by your standards.... You still have not addressed the sunny pink issue on the very first image I posted way back when yet either. The girl did have that skin color and tone. I use 5 monitors and my printers and my eyes and memory to get the colors correct. I do not usually go by any standard anything to make adjustments by. I do not want my sun soaked ladies and girls to get the jaundice!! I say that system is not foolproof and works for many folks, but not all folks on all images all of the time. It just makes some images look unrealistic at times when they don't have to look that way.
I don't mean to be knocking the gurus, but blanket techniques are hard for me to get used to and I don't always trust them. So if you would, you can use the dropper and charts and I will use my eyes and we will see how many images we come to the same conclusion on. Fair enough? Then if there is a big discrepancy I will print them (some again) and we can see how they look and what is up.
I'm out for the night, it's sleepy time and then taxola time for me. :uhoh :cry
Comments
I'm producing test shots with B+W's 486 filter. From their catalog:
The UV filter 010 blocks ultraviolet radiation which can cause blur or – in color shots – blueness. It is ideal for photos in the mountains, by the sea or in areas with very clear air. Pictures become more brilliant, irritating blue haze is avoided and color reproduction remains neutral. This UV filter is also suitable for use as front lens protection.
The recommended product for digital cameras is the UV/IR blocking filter 486, a steep-flanked interference filter which additionally blocks infrared and prevents blur and color cast with IR-sensitive CCDs.
My studio lights don't produce IR and hence I don't see the problem there. It's most acute under tungsten indoor lighting and on-board flash that isn't bounced off the ceiling.
But I have seen thousands of outdoor shots with a similar problem. Let's take an example from a world-class photographer (Andy) with a world-class camera. Here's what he posted of a fair-skinned caucasian in a forum:
That shot is so far from a legal skin tone it would come straight back to us if printed. The skin at least would have to be much closer to this for any consumer to accept it:
It took me all of a couple seconds. I could do more here but I just opened the image to see what the problem was and could not resist one quick click with one of the levels eydroppers.
Yup, I am pretty sure now, with my cheap camera I do not get this sort of thing on my outdoor shots when I am properly white-balanced. I dunno on this, Baldy. I only see this with flash and other forms of indoor lighting. My question is why would anyone upload a photo like this without doing some simple corrections.
-don
-don
The nose measures 65% magenta 46% yellow.
I suppose ezprints won't print this one either? If not, I don't see what good softproofing with the ezprints profile or the CMYK profile does here. Your saying what you see won't get printed. What??? No redheads or sunburn prints here at smugmug? Geesh my wife is a very white redhead for gosh sake! You telling I have to paint her green??
My monitor out of cal??? What gives here? I get prints from stuff that looks like this at walgreens and don't have a problem except the fact that they charge me too much when I pick them up.
-don
Your adjustment is on the left, below. The same man's photo is on the right, taken under studio lights that emit no infrared. Everyone here, who is looking at the man, says he doesn't look anything like the shot on the left.
If you print them, they'll turn out okay unless you choose the true color option. If the lab is faithful to their true color mantra, the left will look nuclear-radiated because the skin measures 7-20% more magenta than yellow.
-don
-don
-don
Looking forward for the new features!
Cheers!
Consumers will not accept prints where the yellow value isn't at least 3% higher than magenta, especially when cyan is as low as 28% of the magenta value.
Here are the acceptable ranges:
http://www.smugmug.com/help/skin-tone
It's a lovely shot, however, and will print nicely with autocolor on.
Now I have not printed the girl myself but I have printed the bandshots. Did you say those bandshots would print and be acceptable, or not?
As far as the edit you did that I was jumping on you about it does look better on 4 of my 5 monitors here. Probably acceptable and I apologize. The monitor I was using here to do the edits and view your edit is acting very strange. However, it does still look a bit unnaturally yellowgreenish to me on the other monitors, probably because of the surrounding grass in the shot. It does still look better than my edits however, and I will recalibrate my ps rig later tonight.
The shot below is a little dark, but I would like to know what you think the customer would say on the fleshtones on this one along with those other two band shots I posted earlier.
Now I think I know what you are going to tell me and this should be interesting.
Once again, sorry for dogging your edit so bad, that junk monitor I was using(my main) was mucking me up. I'll fix that...
-don
Thanks
-don
The T-shirt is equally neutral.
Which camera produced the correct skin tone?
-don
The light is the same. One camera is pointing down (the one on the right) while the one on the left is pointing up towards the sky.
-don
— Actually, I can see how that would happen if the camera had auto white balance on. It would be influenced by the sky.
But the point remains, the customer will accept the skin tone on the right because it's in a legal range of skin tones. On the left, you have to choose autocolor for them to accept that shot.
let's quit talking about crappy photos and crappy edits for a bit
Now, if you would please, start giving me your opinion on the 3 concert shots and the 8 streetshots I posted. Will customers accept them, or do you say they are too crappy?
Thanks
Scott Kelby and Dan Margulis talk of skin tone ranges in their books if you'd rather take it from them (they list the same values we do, except Scott doesn't mention cyan, which is critical).
I don't mean to be knocking the gurus, but blanket techniques are hard for me to get used to and I don't always trust them. So if you would, you can use the dropper and charts and I will use my eyes and we will see how many images we come to the same conclusion on. Fair enough? Then if there is a big discrepancy I will print them (some again) and we can see how they look and what is up.
I'm out for the night, it's sleepy time and then taxola time for me.
Thanks for the discussion.
-don