khorinek

Thanks, I generally shoot in Tv, (shutter Priority) mode and manual Mode when I'm getting blown skin tones. When I see I'm getting skin tones with no definition, I switch to manual mode and underexpose. My issue is, when the camera says I have correct exposure, I get overexposed skin tones. Not sure why this is.

tomcat

These are great ideas and some great advice. The one caveat he pointed out is that he switched to a new model of camera and started getting these hot skin tones. As I mentioned not too long above, his situation reminds me of the experiences that I had when I first switched to the Nikon D3--hot cheeks, foreheads, and noses. And a recent post just above ours suggested to decrease exposure for him. That's the only way I was successful at reducing the highlights and I always shot in manual, just like you do. But your suggestions are great for all of us to remember and keep in mind. I find that my attentiveness is waning and I need to heed your advice and go with what you get for stage lights. I recently shot a Christmas play at a local high school and the blue light traces from a sidelight gave a beautiful cast to the performance but left me with a blue streak on the sides of the student actors. No way to remove the lighting, but to just use it.

RRS Loc: Not sure

When I look back at the picture you sent to show your problem I would have thought that you'd be shooting in Av (aperture priority) to control the DOF. Is the EC (exposure composition) set to (0). If you are shooting about 500 shots a day and if in the same light I'd be shooting in manual, you then control the DOF and needed shutter speed. You could dial in a negative EC to help with the skin tones.

RRS Loc: Not sure

Can you go into LR or PS to colors and decrease the blue color without changing any of the other colors in the shot. If the color blue is in other parts of the shot this will not work this way. Just a thought.

burkphoto Loc: High Point, NC

This is probably more than you need to know in response to your question, but it speaks to the root issues you may face.

Automation. It's great — except when it's fooled by what amateur photofinishers used to call "subject failure".

Subject failure is really the failure of a subject to appear as a perfectly balanced shade of gray reflecting between 13% and 18% of the light falling on it (depending on manufacturer calibration references).

Reflected light meters and AWB (really, reflected color temperature meters) are both dumb. Stupid. Annoyingly predictable. They make anything you point them at look gray, if you take the reading at face value. So it behooves us as photographers to evaluate scene values and brightness ranges. There are plenty of times when a scene is far from the average gray reflectance! That's when automatic modes fail to differing degrees.

Let's assume for the following examples that you are working in JPEG mode, at camera default settings, which include AWB (automatic white balance). You are NOT using full (NO auto ISO) manual mode, which means you ARE using either Av, Tv, Program Auto, iA, or M with auto ISO.

Point an in-camera meter at a mostly black scene, and overexposure will be the result. Point it at white, and underexposure will be the result. Point it at a predominantly red scene, and the rest of the scene will be tinted cyan (blue-green). Point it at a predominantly blue scene, and the rest of the scene will be tinted yellow. Point it at a predominantly green scene, and the rest of the scene will be tinted magenta (dark pink). Point it at a predominantly cyan scene, and the rest of the scene will be tinted red. Point it at a predominantly magenta scene, and the rest of the scene will be tinted green. Point it at a predominantly yellow scene, and the rest of the scene will be tinted blue.

What's worse, THAT IS PERFECTLY NORMAL, BUT IT IS ALSO PERFECTLY WRONG AND UNACCEPTABLE.

That is why professionals so often rely on the use of exposure targets or hand-held incident light meters, and why we also use suitable white balance targets to perform a custom (or preset, or manual) white balance. We may even use a tool called the "ColorChecker Passport," that allows a Lightroom plug-in to profile the camera for the exact scene, yielding extremely accurate color.

Out of the box, cameras are set to a "normal" JPEG look. Users can alter all sorts of properties of JPEGs — "Picture Style" is sort of like a film stock simulation (Normal, Portrait, Landscape, Standard...). But you can adjust various Picture Style properties like hue, saturation, color tone, sharpness, contrast, dynamic range compensation...

I encourage new users of any camera to TEST ALL SETTINGS at each "tick" throughout the range of control:

Step ONE: Set up a scene under CONTROLLED, CONSTANT LIGHTING, with lots of colored and highly detailed objects. Set your nominal exposure by reading ONLY an exposure and white balance target held at the exact plane of exposure. If you have one, include a Color Checker Chart, and/or Kodak or Tiffen Q-13 or Q-14 21-step gray scale, and/or another standard color and gray scale device. Make several HUNDRED exposures, varying each control on the entire camera until you have altered JUST ONE AT A TIME (from standard, out-of-the-box settings). Write the frame numbers in a notebook, noting the settings for each frame! If you don't do this, you will have to read the EXIF metadata table for each image, later. If you can find a model who isn't bored easily, have a real live person in each scene, posed essentially the same way.

Step TWO: Use a hardware and software kit (from X-Rite or DataColor) to calibrate the best monitor you have or can afford. This will be necessary for a color-accurate workflow anyway, and will keep your test evaluation HONEST.

Step THREE: Go through your test images, and see what happened. This will give you MANY clues as to what your camera can do, or is doing.

I did this test back in 2003 with a Nikon D100, and have done it with every camera I've had since. It opened my eyes to what you can, and cannot do with JPEGs. Results with most of the Canons I used were the same or nearly identical, even from one generation to the next, and one sample to the next. Once you eliminate the variables of automation and scene changes, you can see what is actually happening.

I love automatic exposure and white balance when working in raw mode, but never, ever use them for JPEG work. That's because JPEG latitude is about +/- 10 points on color, and + 1/3 stop, - 2/3 stop on exposure, before visible problems result that cannot be corrected in post-production. Most often, when recording JPEGs at the camera, I'm working under controlled, consistent lighting, and I'm using FULL manual exposure with a custom (manual, preset...) white balance.

tomcat

Thanks for the suggestion. The blue is an electric blue color, a blended blue, and trying to desaturate it in LR left me with some weird skin patches, so I left it alone. Reminds me of the times when someone wearing a white shirt would have green reflected from tree leaves.

E.L.. Shapiro Loc: Ottawa, Ontario Canada

SUBJECT FAILURE! This is so true! If you use ANY kind of exposure or color analysis automation SUBJECT FAILURE will become the bane of your existence! That is, unless you know what it is and how to manage it. This gremlin occurred in automatic printers back in the analog color lab scene, it can come to haunt you in automated flash exposure and I am sure it is alive and well and comes back to bite us in the rear-end in many aspects of digital photography. This is simply where the device, wahtever it is, makes some kind of integrated reading or samples a ligh value or color value other that the main subject or target and makes a correction based on bad information. It can be intermittent or constant and depends on the circumstances.

In my old lab at my studio we used the Kodak "S" printers and others made by Pako to produce proofs and what labs used to call "economy" or machine prints. The machine would integrated all the colors in a negative and produce a color balance based on a "mishmash" of the colors. If the image was a close up where the dominant color was a skin tone, it could work. If however it was a full length portrait or a group made out of doors and there was a dominance of green foliage, the skin tone MAY have gone magenta. We would make quick set prints of a section of the image and the color corrector would mark the corrections for the final print. The density was also tweaked and adjured accordingly. I suspect, the internal workings of a digital camera may be affected by the same issue in terms of exposure and color balance. For final CUSTOM prints, made on the enlargers, we used calibrated color analyzers which were set up to probe the diffuse highlights in a skin tone, as to color and density- or a gray card and if all systems were working properly we would get a great skin tone an all the other colors would fall into place.

To test this theory, try a close up shot of a person where the face and skin tone are dominant in the frame then back up to include more background, preferably with a NON-neutral color or shade, and see of the camera is consistent .

Exposure wise, the same problems can happen in the camera, especially with TTL flash where the the camera is making a matrix reading.In a large function room or a church, the inverse square law kicks in and the camera SEES the dark abyss in the background and overexposes the subject beyond correction.

Theses scenarios may seem very detached from the problem at hand but it may provide some insight into the theory. So many folks, when they complain about theses inconsistencies, they say the CAMERA did this or didn't do that etc. This indicates to me that there are times when the automation needs to be bypassed or circumvented and the photographer needs to make the necessary adjustments. In the case of the flash scenario, I simply memorize my flash exposures, learned to estimate distances and set the aperture manually. I never experience subject failure and my exposure are tack-on! If I am not setting up a custom white balance and using the camera automatic system and am concerned about skin tones, I take all my readings in close up of the subjects face, lock everything in and recompose as required.

This, along with Burkophot's detailed and exceptional explanation of the theory should get you on the right track to solving theses issues. You will also have to experiment with and without some of theses menu selections such as "portrait" etc. Some of theses setting may alter the color or density to some sort of engineering preconception but will seem unacceptable to you. If you are getting "brown" skin tomes PERHAPS the setting you have selected is producing a sepia-like semi-monochromatic image that is too warm. Sound like a simulation of the old Ektacolor (CPS) film before the later Varicolor and Portra incarnations were introduced. The film was so red biased that it produced gray grass. I used to joke that if I wanted sepia prints, I would shoot black and white and tone my prints.