Photographer
Robert Randall
Posts: 13890
Chicago, Illinois, US
Lighting seems to be a buzz issue today, and in viewing the work of others that claim they did it all with great lighting, I've run across a lot of telltale signs of bad lighting being fixed with Photoshop. For instance, when an image is so poorly lit and exposed that you have to crank on the exposure slider to save that raw file, it leaves a weakness in the tones that separate from highlight to midtone, and midtone to shadow. Usually that manifests itself in a tone that seems artificially transparent due to a corrupted black point. Do you know of a way to fix that look and salvage what the world views as a lost effort. I'll give you a hint, we did a similar exercise when changing black nylons to flesh colored nylons. Edit... The problems I'm referring to are images that were under exposed by at least three stops.
Photographer
Star
Posts: 17966
Los Angeles, California, US
Robert Randall wrote: Lighting seems to be a buzz issue today, and in viewing the work of others that claim they did it all with great lighting, I've run across a lot of telltale signs of bad lighting being fixed with Photoshop. For instance, when an image is so poorly lit and exposed that you have to crank on the exposure slider to save that raw file, it leaves a weakness in the tones that separate from highlight to midtone, and midtone to shadow. Usually that manifests itself in a tone that seems artificially transparent due to a corrupted black point. Do you know of a way to fix that look and salvage what the world views as a lost effort. I'll give you a hint, we did a similar exercise when changing black nylons to flesh colored nylons. really, if I inverse and then use that as a base layer I can skin graft back on top and save the blown areas?
Photographer
Robert Randall
Posts: 13890
Chicago, Illinois, US
Star wrote:
really, if I inverse and then use that as a base layer I can skin graft back on top and save the blown areas? That and setting the layer to difference are good for blown areas, and I guess I wasn't clear enough in my OP. Most of the problems I see are with under exposure, by at least three stops.
Photographer
slave to the lens
Posts: 9078
Woodland Hills, California, US
Robert Randall wrote: Lighting seems to be a buzz issue today, and in viewing the work of others that claim they did it all with great lighting, I've run across a lot of telltale signs of bad lighting being fixed with Photoshop. For instance, when an image is so poorly lit and exposed that you have to crank on the exposure slider to save that raw file, it leaves a weakness in the tones that separate from highlight to midtone, and midtone to shadow. Usually that manifests itself in a tone that seems artificially transparent due to a corrupted black point. Do you know of a way to fix that look and salvage what the world views as a lost effort. I'll give you a hint, we did a similar exercise when changing black nylons to flesh colored nylons. Edit... The problems I'm referring to are images that were under exposed by at least three stops. Yes, but it's definitely not the right way. I run a curve on the whole thing, focusing on the midtones. Using the exposure slider in ACR seems to cause more issues than it helps. I've tried just tweaking the g or b channel, but can't seem to get it where I want it. I'd love to hear the best way, or even a better way... cause I know what I'm doing isn't it. EDIT: 3 stops? Uh-uh. No clue how to bring that back.
Photographer
Sean Baker Photo
Posts: 8044
San Antonio, Texas, US
The question causes me pause enough to wonder whether disabling the base tone curve in the raw conversion would minimize the flattening which would be applied to the data by default (especially at -3EV); allowing reconversion & application of a base tone curve after adjustment of the exposure. Come to think of it, I'm not sure why ACR doesn't do this by default (Rick?). But raw engine tweaking doesn't strike me as Bob's style, and I've lost my copy of the original nylon file so I'll just subscribe and anxiously await someone divining the answer .
Retoucher
James Minshall
Posts: 218
Bedford, Indiana, US
underexposed 3 stops? find a new photographer! fify anyway, I'd gently raise exposure by tweaking all relevant settings as much as possible in raw, then maybe try the shadows/highlights thingy in PS. It works pretty well, especially if you do separate channels and set the right blending mode. Dunno what it is.. been a while since I've had to perform such a miracle.
Photographer
Star
Posts: 17966
Los Angeles, California, US
Robert Randall wrote:
That and setting the layer to difference are good for blown areas, and I guess I wasn't clear enough in my OP. Most of the problems I see are with under exposure, by at least three stops. sorry, I am more likely to overexpose since most of my information is in the brightest channel i am deathly afraid of underexposing.
Photographer
Phil Drinkwater
Posts: 4814
Manchester, England, United Kingdom
Robert Randall wrote: Edit... The problems I'm referring to are images that were under exposed by at least three stops. Three stops!!
Photographer
Roy Whiddon
Posts: 1666
Ottawa, Ontario, Canada
I will often correct one stop of underexposure in Photoshop (better than blowing highlights), but three stops? That's a discard in my book.
Photographer
Phil Drinkwater
Posts: 4814
Manchester, England, United Kingdom
RW Photo Art wrote: I will often correct one stop of underexposure in Photoshop (better than blowing highlights), but three stops? That's a discard in my book. Assuming you're shooting RAW, you usually have a stop latitude on the highlight end. I'd prefer to overexpose by 1/2 stop and keep the shadow quality.
Photographer
Kevin Connery
Posts: 17824
El Segundo, California, US
Phil Drinkwater wrote:
Robert Randall wrote: Edit... The problems I'm referring to are images that were under exposed by at least three stops. Three stops!! One of the shots in my MM portfolio was 4 stops or so underexposed; the flash didn't fire. If memory serves, there were fewer than 32 shades in each channel. Pushing Exposure in ACR to the max, it was still a very dark image, but... I typically just delete such images, but I needed an example for a lesson, and played with it--and liked the results. It'll never be a clean shot, but grungy was possible.
Photographer
Ex Voto Studio
Posts: 4985
Columbia, Maryland, US
Sean Baker wrote: The question causes me pause enough to wonder whether disabling the base tone curve in the raw conversion would minimize the flattening which would be applied to the data by default (especially at -3EV); allowing reconversion & application of a base tone curve after adjustment of the exposure. Come to think of it, I'm not sure why ACR doesn't do this by default (Rick?). But raw engine tweaking doesn't strike me as Bob's style, and I've lost my copy of the original nylon file so I'll just subscribe and anxiously await someone divining the answer . I love when you guys talk!!! *waits to see what Bob says*
Photographer
Andrew Thomas Evans
Posts: 24079
Minneapolis, Minnesota, US
This is just off the top of my head. - How about using one of the channels as a new layer or mask? Usually one of them has enough info in it to fill in blown out areas. (sometimes I've heard this works if an image is under exposed, just replace the other channels with a one that isn't as dark.) - Or just re-do the raw conversion, then mask out areas you don't want. - Or what the hell, make a copy of the image, set it to soft light, and then desaturate it back to normal. (I'm bored at work)
Photographer
Skydancer Photos
Posts: 22196
Santa Cruz, California, US
Ex Voto Studio wrote: I love when you guys talk!!! *waits to see what Bob says* Bob is out of commission for 3 days.
Photographer
Star
Posts: 17966
Los Angeles, California, US
Robert Randall wrote: For instance, when an image is so poorly lit and exposed that you have to crank on the exposure slider to save that raw file, . ah-ha here is where i got confused. I am sorry, for some reason I read this as being cranking on the recovery. I did that cause i am a dork.
Photographer
Lumigraphics
Posts: 32780
Detroit, Michigan, US
Kevin Connery wrote:
One of the shots in my MM portfolio was 4 stops or so underexposed; the flash didn't fire. If memory serves, there were fewer than 32 shades in each channel. Pushing Exposure in ACR to the max, it was still a very dark image, but... I typically just delete such images, but I needed an example for a lesson, and played with it--and liked the results. It'll never be a clean shot, but grungy was possible. This image https://www.modelmayhem.com/pic.php?pid=3369008 Was terribly underexposed as the flash didn't fire. I almost trashed it but after some playing in Photoshop was able to rescue what is now one of my favorite shots ever.
Photographer
Mike Stalnaker
Posts: 1881
Sarasota, Florida, US
RW Photo Art wrote: I will often correct one stop of underexposure in Photoshop (better than blowing highlights), but three stops? That's a discard in my book. I agree.
Digital Artist
Koray
Posts: 6720
Ankara, Ankara, Turkey
I usually add snake skin and motorcycle parts here and there...and liquify the body and remove the eyes!
Photographer
Lumigraphics
Posts: 32780
Detroit, Michigan, US
Mike Stalnaker wrote:
I agree. Read my post, one above yours, and look at the linked image. WAY more than one stop under, more than three stops under.
Photographer
Mike Stalnaker
Posts: 1881
Sarasota, Florida, US
I saw that, a great job of recovering a very unexposed image. I don't want to get into a critique of it here. I just find that if it's more than 2 stops under exposed it's not worth trying to fix. I have done some that were almost 3, but never liked the final product. I am watching this post closely to see if any photoshop wizards have any secrets to share though :-)
Photographer
Leggy Mountbatten
Posts: 12562
Kansas City, Missouri, US
Star wrote: sorry, I am more likely to overexpose since most of my information is in the brightest channel i am deathly afraid of underexposing. If your most important information is in the highlights, isn't overexposure a greater risk? Blow a channel and that information is gone. Underexpose a little bit, and you can always bring it back up.
Photographer
Star
Posts: 17966
Los Angeles, California, US
Leggy Mountbatten wrote: If your most important information is in the highlights, isn't overexposure a greater risk? Blow a channel and that information is gone. Underexpose a little bit, and you can always bring it back up. no. if you underexpose a stop the information is gone. The greatest, not most important, but the greatest amount of information is in the brightest stop of the dynamic range. Please google this information. i feel that perhaps the technical information is not making sense to you. "This is important because the typical digital camera can capture roughly six stops of dynamic range - and of these six stops, half of the information captured (independent on the bit-depth of the camera's encoding) is in the brightest stop. " http://www.flickr.com/groups/ngproinvit … 502013483/
Photographer
Mike Stalnaker
Posts: 1881
Sarasota, Florida, US
Star wrote:
no. if you underexpose a stop the information is gone. The greatest, not most important, but the greatest amount of information is in the brightest of the dynamic range. Please google this information. i feel that perhaps the technical information is not making sense to you. "This is important because the typical digital camera can capture roughly six stops of dynamic range - and of these six stops, half of the information captured (independent on the bit-depth of the camera's encoding) is in the brightest stop. " http://www.flickr.com/groups/ngproinvit … 502013483/ That was my thinking and experiance. I was always told to shoot to the right.
Photographer
Leggy Mountbatten
Posts: 12562
Kansas City, Missouri, US
Star wrote: no. if you underexpose a stop the information is gone. The greatest, not most important, but the greatest amount of information is in the brightest stop of the dynamic range. Please google this information. i feel that perhaps the technical information is not making sense to you. "This is important because the typical digital camera can capture roughly six stops of dynamic range - and of these six stops, half of the information captured (independent on the bit-depth of the camera's encoding) is in the brightest stop. " http://www.flickr.com/groups/ngproinvit … 502013483/ Oh yes, I'm well aware of how the information is encoded (Adobe's white paper on digital capture was an interesting read), but in the real world, the point is to get your exposure correct, rather than risk blowing a channel, which is what sounds like you're doing. There is a lot of information in our RAW files these days, particularly when most of them are at least 14 bit. Just like with film, it's a matter of choosing where the most important information for your particular photos is. If you overexpose, you risk losing the highlights. If you underexpose, you'll lose the shadows. The correct exposure is a balance of these interests.
Photographer
Lumigraphics
Posts: 32780
Detroit, Michigan, US
Mike Stalnaker wrote: I saw that, a great job of recovering a very unexposed image. I don't want to get into a critique of it here. I just find that if it's more than 2 stops under exposed it's not worth trying to fix. I have done some that were almost 3, but never liked the final product. I am watching this post closely to see if any photoshop wizards have any secrets to share though :-) It wasn't really recovering it, more a case of making it a photomanipulation. I wouldn't ever hold that up as an example of how to correctly light a shot! And yes, I think I got lucky with it, usually when it is that far under its total garbage.
Retoucher
Star the retoucher
Posts: 437
Los Angeles, California, US
Leggy Mountbatten wrote:
Oh yes, I'm well aware of how the information is encoded (Adobe's white paper on digital capture was an interesting read), but in the real world, the point is to get your exposure correct, rather than risk blowing a channel, which is what sounds like you're doing. There is a lot of information in our RAW files these days, particularly when most of them are at least 14 bit. Just like with film, it's a matter of choosing where the most important information for your particular photos is. If you overexpose, you risk losing the highlights. If you underexpose, you'll lose the shadows. The correct exposure is a balance of these interests. I am sorry but you are wrong in your understanding of the information. If you underexpose by one stop you lose 1/3 of your information. If you overexpose by one stop you lose 1/16 of your information. Again, i believe your technical understanding is flawed.
Photographer
The Bald Photographer
Posts: 518
High Wycombe, England, United Kingdom
Star the retoucher wrote:
I am sorry but you are wrong in your understanding of the information. If you underexpose by one stop you lose 1/3 of your information. If you overexpose by one stop you lose 1/16 of your information. Again, i believe your technical understanding is flawed. Isn't it a case of what you perceive in the result, rather than how much information is technically available to see? If you underexpose you perceive noise and maybe some quantisation. If you blow the highlights you see distortion in the colours (depending on how many colours are 'flat lining') or no detail whatsoever. So isn't it actually a case in practice of a choice of what perceived compromises you want to risk in the resulting image?
Photographer
Star
Posts: 17966
Los Angeles, California, US
The Bald Photographer wrote:
Isn't it a case of what you perceive in the result, rather than how much information is technically available to see? If you underexpose you perceive noise and maybe some quantisation. If you blow the highlights you see distortion in the colours (depending on how many colours are 'flat lining') or no detail whatsoever. So isn't it actually a case in practice of a choice of what perceived compromises you want to risk in the resulting image? no
Photographer
Photons 2 Pixels Images
Posts: 17011
Berwick, Pennsylvania, US
The Bald Photographer wrote: Isn't it a case of what you perceive in the result, rather than how much information is technically available to see? If you underexpose you perceive noise and maybe some quantisation. If you blow the highlights you see distortion in the colours (depending on how many colours are 'flat lining') or no detail whatsoever. So isn't it actually a case in practice of a choice of what perceived compromises you want to risk in the resulting image? Star wrote: no Maybe I can help. I'll explain the way I understand it. I'm a simple guy. Think of the difference between a printed page and a displayed page on a monitor. The monitor starts black and has light (information) added to it. "Shadow areas" are basically a lack of information in a certain area. With the printed page it's the opposite. The page starts white and has pigment (information) added to it. In this instance, the "Shadow areas" are where more information is because you are basically saturating that area with the pigment. Now change printed page with film and monitor with digital. Since we're not discussing film, the important thing here is that with digital you will have more information in the higher end (overexposed) areas. Thus why it's more difficult to bring back an image that is severely underexposed as opposed to overexposed. Thus the reason for Mr Randall starting this thread.
|