how does one control an incident meter if there are distinct shadows and highlights?
You wouldn't control the incident meter. Instead, you would control, or manipulate, the light falling into the scene. Maybe you add reflector to bounce more light into the scene, maybe you add a black reflector to block light from falling into the scene, or maybe you add fill flash to add light somewhere in the scene. Or, maybe you deal with the issues in post-production.
But, again, you don't deal with the meter. You deal with the light coming into the scene when you're using an incident meter. The incident meter just does its job, i.e., measures light falling into the scene. It doesn't fix problems, it only gives you data about the light levels/ratios falling into the scene.
The photographer is what solves the problems, , using his/her understanding of the way light falls into a scene, instinct, intuition, modifiers, post-production, and etc.
That said, a photographer using an incident meter is usually going to be more concerned with achieving contrast equivalent to the scene they actually see or that they are artificially modifying/creating using flash photography, reflectors, or etc.
What if the sun is very bright, but there are also many shadow areas that I want to capture? Surely in these kinds of conditions, an incident meter will give a reading that needs to be adjusted, not simply one that represents the way the eye sees.
The answer to that line of questioning depends on where you take the meter reading from. Remember, in my previous post where I used the example of the black cat, the white cat, and the gray cat, all cats were under the
SAME level/quality of light
(i.e., ISO 200, 1/60th, F/4 to refresh your memory). They were NOT in a scene with MIXED light levels
(e.g., shade in one part of the scene and sunlight in another part of the same scene, as you suggest in your questioning).
Again, when you do have mixed light levels in the same scene, the incident exposure reading will depend on where you take the meter reading from and the level of light in that area. But, believe it or not, the incident meter will give you a reading that will render the area in the scene that you're metering similar to what your eyes see. For instance, let's put our black cat and white cat near a tree together. Let's say that the white cat is lit by sun light while the black cat is lit by shade. Taking a meter reading in front of the white cat lit by sunlight will give you a reading that will render the white cat bright white, similar to how it would appear to the eye. Taking a meter reading in front of the black cat lit by shade light will give you an incident reading that will render the black cat dark, similar to how it would appear to the eye.
Because each subject in the scene is being lit by DIFFERENT light sources, you WILL get different readings, as opposed to the example in my previous post where the cats were all lit by the SAME light source. If there are any "problems," then the photographer using the incident meter would manipulate the light until he/she gets desired light ratios.
Regardless, with an incident meter, a black cat will always be black when you're metering the light falling upon the black cat. And, a white cat will always be white when you're metering the light falling upon the black cat.
Thus, if you meter a black cat in shade light, it will appear to be a black cat in shade in your photo. If you meter a black cat in sun light, it will appear to be a black cat in sun light in your photo. If you meter a white cat in shade, it will appear to be a white cat in your photo. If you meter a white cat in sun light, will appear to be a white cat in sun light in your photo.
But, again, you don't adjust the incident meter because all it is supposed to do is give you feedback on how light is falling into the scene. The adjustments are made to the scene itself or to the image in post-production.
The meter I use allows the user to save several meter readings into memory so that they can be "averaged" into one reading. In my experience, it has proven very useful in some cases but not in all cases. In every real world situation, there's no guarantee that resulting average exposure is going to render the scene exactly the way you want it to look. So, it's not something you can blindly rely upon. You, sort of, have to develop your own intuition and metering techniques that suit your needs.
How does one adjust if there is no reference point such as a portion of the subject set to zone xyz (in a reflected light metering)?
The question is unanswerable
(with respect to the type of answer you seem to be looking for).
AFAIK, there is no full proof metering trick that you can use to "adjust" the incident exposure in any given contrasty lighting scenario to an exposure value that you will always approve of. So, you have to experiment with an incident meter, learn how it responds in real world situations, and then develop an intuition for complex lighting scenarios.
But, most importantly, you have to have a really good reason to use an incident meter in the first place (rather than a reflective meter), otherwise you will waste time and get confused. Someone using an incident meter is more likely going to be more concerned with achieving
"contrast relative to the scene" rather than
"a perfect exposure." And if you're planning on learning how to use an incident meter, begin learning with simple, evenly lit scenarios -- don't jump right into a complex scenario without getting a feel for what an incident meter does in a simple scenario first.
When using an incident meter, the reference point cannot be established from the subject or any other object in the scene.
The reference point is the light falling into the scene (e.g,. shaded light, sun light, cloudy light, tungsten light, fluorescent light, candle light, bounced light, and etc. <--- those are your references).
If you're taking a photo of a yellow taxi cab on the street, the incident meter should recommend a meter reading that renders the yellow (of the taxi cab) as it appears under the illumination of the light falling upon it (depending on where you take the meter reading).
If, for instance, the front end of the taxi cab is lit by sun light and if you take a meter reading in that area, the meter will suggest a reading that will give you a picture similar to the way you see that front end of the taxi cab with your own eyes. In your photo, you'll see bright yellow and the presence of sun light. The incident meter won't "care" about the other parts of the scene (e.g., the shadows) because the meter can only evaluate one place at a time.
However, if the back end of the taxi cab is lit by shade light and if you take a meter reading in that area, the meter will suggest a reading that will give you a picture similar to the way you see the back end of the taxi cab with your own eyes. In your photo, you'll see darker yellow and the presence of shady light. And, again, the incident meter won't "care" about the other parts of the scene (e.g., the highlights) because the meter can only evaluate one place at a time. So, the front end of the taxi cab may show blown highlights if it falls outside of the dynamic range of your capabilities.
What a photographer with an incident meter does in such a scenario is measure the shadows (i.e., the light falling upon the back end of the cab) and records the reading. Then, the photographer measures the highlights (i.e., the light falling upon the front end of the cab) and records the reading. With that data, the photographer can manipulate the lighting in the scene or make adjustments during post-production. But, it is the ratio of light that the photographer is concerned about, not the reflectivity of the objects in the scene.
Because of the limitations of incident meters, it makes more sense to use a reflected meter when using the Zone System in many cases. The type of meter I use depends on what I'm shooting and how I shoot it; so I use both meters.