Upload & Sell: Off
| p.1 #3 · anyone recommend using the SpyderCube for white balance? |
The cube design cleverly simulates a full range scene from specular highlight to black void but lacks black and white textured surfaces which are a better visual indicator of when exposure is optimal.
The basic shortcoming of digital is it can't reproduce a full sunny scene as seen by eye in all conditions. These shots illustrate that and the most practical workaround.
The only differences in the two shots is the angle of the light at 11AM and the size of the more important content, the target. In the first the target is facing East and mostly highlighted by the sun, and in the second it is turned facing West and entirely in shadow. The knowledge of what the content should look tempers your expectations and judgement of which is the more "correct" exposure.
Technically they are both the same, exposed to keep the white towel under clipping where hit by the sun, the criteria selected for the test. Had the criteria been to reproduce the gray card at 18% on a print made from both shots I would have needed to expose the second with 2-3 stops more exposure which would have blown the highlights. It's a dilemma photographers have faced outdoors since the invention of color film; what to clip on the ends of the tonal scale to reproduce the midtones like shaded faces correctly.
The workarounds are to avoid crossed lighting when a full range is needed or use flash. The first shot where the sun was more or less over my shoulder and flat lighting the target used the sun as the "key" light and put everything important in it. I'd use the same strategy if shooting a soccer game by chosing which side of the field to stand on then exposing per the clipping warning on the white uniforms knowing the detail in the black ones would also be mostly in the sun and reproduced by the range of the camera.
This shot illustrates how flash can be used in backlight to overcome the short range of the camera in backlight by changing the contrast on the shaded side .
Here's a candid single flash portrait with the same flash strategy:
No expensive targets needed, just knowledge of the limitations of the recording medium and how to work around them. Simply expose sunny highlights under clipping and fill via flash or reflectors until the shadows are lifted to where the camera can record detail in the shadows also.
Color fidelity is something best controlled at capture via Custom WB with a gray card. That eliminates the need to click correct the test shot then copy / paste the settings into the other RAW files. The actual color rendering of the camera is affected by the style selected. To judge how colors change it's best to use a standard color chart like the MacBeth Color Picker.
Below is an illustration I created in Photoshop from the RGB values listed in the documentation for my Mini Color Checker:
Here's what it actually looks like when capture and different styles are applied to the RAW capture in PP:
When different styles are selected it alters the color patches but not the neutrals as a TEMPERATURE (blue/yellow) or TINT (green / magenta) slider correction would. That's ideal when you want a bit more saturation in skin tones without making the subject's white clothing off color.
In Photoshop the Styles are applied on the "Camera Calibration" tab:
Personally I find the "Portrait" style oversaturates reds in skin too much. I'd never use "Landscape" for portraits either because as seen from the chart comparisons and below it exaggerates saturation of the RGB primariles even more:
The advantage of having the color chart in a comparison like that is being able to know what RGB values are in each color block in the chart via eye dropper readings and compare numerically how the styles change the RGB readings of each color.
The colors on the chart are selected by MacBeth to represent light / dark skin, "grass" green, "sky" blue, etc... Reference colors a viewer will relate to as "normal" or not. The gray scale isn't as white or black as possible, more in the range of where detail should be recorded. So a correctly exposed target will not be one that clips the white patch at 255, but rather one that reproduces it a 243 per the documentation. When the white patch is exposed for 243.243.243 (about 1/3 stop under 255 clipping) the darkest patch should read 52.52.52 not 0.0.0. 50.50.50 is the level at which detail above solid black void can usually be seen in a photo. If exposure pegs the white at 243 and the black patch isn't 52 it indicates the key:fill ratio is not fitting scene to sensor optimally and more or less fill is needed.
After trying different options I stick with "Neutral" at capture, set Custom WB off a plain 8 x 10 gray card to get gray scale neutral SOOC, then shoot the subject holding a similar test card and color charts as editing references. I will first open that test shot with the Neutral / Neutral SOOC baseline in the Camera Calibration tab, see how it looks, try different SELECTIVE color changes, then decide by before / after subjective visual comparison which best suits each subject.
In addition to the "canned" styles include in ACR by Adobe to match those the camera creates there are the sliders on the left to tweek the color. Once the ideal mix for that subject is found it can be saved as a pre-set and will be automatically applied to the other RAW files in the batch as opened so there's no additional PP overhead.
So while the SpyderCube is a good tool for what is designed to do, it doesn't provide the complete set of information needed to precisely control the process and there are less expensive means to that result. I use a $17 Kodak Gray Card, the Mini-Color Checker, and a couple of towels: white to evaluate highlight exposure visually and back to evaluate when the fill is sufficient to lift the shadows to were the camera sensor range can record detail:
I first used the white towel back in 2005 when trying figure out how Canon ETTL flash ratios worked compared to setting flashes manually as I had for many years. I due to the pre-flashes couldn't meter them conveniently and needed some visual way to tell when FEC was exposing the highlights correctly for detail and how once the correct FEC was set based on that criteria what A:B ratio was needed for a shot like this with a black suit and white shirt:
I also use a white towel to calibrate my Sekonic meter to the actual ISO of my camera. I meter the lighting (e.g. f/5.6) then shoot a bracketed series in 1/3 stop increments from f/4 ... f/8, open them on the computer and look at the image of the towel and decide visually which is best exposed. Doing that I discovered my Meter was out of sync with my camera by 1/3 stop and needed a compensation factor added to the meter.
For a while after that I'd meter then check the reading by either having a the subject hold a towel in a test shot, or just put it on a stand where there face would be before they even arrived:
As you can see in that shot the contrast of the rim light is visible on both the foreground side of the towel and the background and the rim lit parts still have detail; all set via the clipping warning on the camera because after metering then taking test shots to verify I discovered it was faster and more accurate to just shoot and adjust based on the clipping warning.
Aperture is selected for DOF and fill raised to record detail in shaded black content (represented by the draped black towel). The rim light is raised until clipping is seen then backed down until it disappears (1/3 stop) the Key light in front is adjusted until it clips then backed down 2/3 stops. I adjust the background similarly 2/3 below clipping to retain the ambience of the rim lighting and match the "white" on the foreground side, the same balance perceived by eye in backlit situations.
Outdoors the workflow is similar for exposure:
Start by getting the rim light (provided by direct sun) under clipping:
Add centered chin level fill just to reveal the shadow detail at the f/stop selected:
Add a second raised "Key" flash to mimic the natural downward modeling of the skylight the first flash used as fill killed:
After all the lights are set I have the subject hold the target, set custom WB off the gray card then shot them holding the target with Custom WB as my baseline Neutral Style / Neutral Gray Scale reference for editing in ACR as shown above.
If that approach seems logical to you in terms of the goal of making a reproduction seem as "seen by eye" real as possible given the technical limit of the process you might want to skip the SpyderCube and instead get:
1) A 8 x 10 gray card large enough for setting custom WB. The Koday Card kit cost $17 at B&H last time I checked.
2) A Macbeth Color Checker: I now use the mini version attached to the gray card so set Custom WB and have the color reference on one card.
3) White and Black Terry Cloth: Draping them over as stand to simulate range is the same concept as the cube, but uses the more meaningful evaluations points of highlight DETAIL and shadow DETAIL as the guide points. They don't need to be large. I keep them in my bag on top of the cameras, which adds padding.