Is your goal reducing production time or maximizing quality?
If in a position to use a handheld incident meter, it will result in both.
Not necessarily. To minimize post processing time, you want to get the brightness correct in-camera.
This is incorrect. It depends on the shot you're after. If you know the shot you're after--which is normally the case in portraiture--and you know enough about light to be able to work with it, and how to use an incident meter. It will result in getting the shot right in camera, thus minimizing the need for post production work. Image quality is subjective for the most part. Imagine what you want, use the light to create it. If the image matches what your mind's eye saw, that would be a quality image.
The exposure differences we are talking about affect noise levels, blown highlights, a detail lost in shadows. In the context of this discussion, these are the issues I am referring to when I talk about "Image Quality".
As to what brightness level a skin tone should be in a JPEG, that's partially an artistic question. As to minimizing noise level in the image, that's an engineering issue we can measure.
It's also an artistic question as to whether or not we want to maintain details in the shadows and highlights. However, if we do want to retain those details, we can say that we have higher "image quality" if we have retained those details, and if we have minimized image noise.
Artistically, we may actually want noise in the image, and/or blown out highlights. While such an image may be wonderful and amazing, it may not score highly with the narrow definition of "image quality" we are using here.
Maximizing in camera quality results in less post production time.
Strangely this isn't true. The exposure that produced the absolute maximum quality typically will require post processing.
Again, "quality" is subjective. When using a handheld incident meter--and knowing the shot you want--many if not most of the resulting pics need very little if any post processing with the exception of, say, some skin smoothing, cropping, things like that. Abbreviated workflow: Meter off the model > set my camera settings > snap the model holding a color chart > take a bunch of pics to my heart's desire. Post processing: use the pic of the model holding the color chart to sync white balance to all the pics from that session. Crop the keepers if I feel it's needed (rarely) > off to PS for skin smoothing if it's needed and any other artistic embellishments. Done. Rarely any tweaking of shadows, highlights, exposure or color. I'm the kind of person who actually likes tweaking my photos in post production. Often times I'm disappointed at the lack of tweaking needed in portrait shots.
For the purposes of this discussion, Image Quality refers to maintaining details while minimizing noise.
Knowing how to use a meter, in many circumstances, is going to be faster than snapping, chimping, adjusting, snapping, chimping some more, etc. Especially if, for example, you're in a bright daylight environment. Squinting at a tiny LCD in the bright sun sucks.
If those are the only two choices than an external meter might be a better choice.
Many people have the option of using the in-camera meter. Many find this easier and more reliable than an external meter. Shooting with the internal meter saves the metering step. A quick glance at the histogram and a check for blinkies tells you if you need exposure compensation. It is not a slow nor a crazy workflow.
Sure, it may be a better choice for some. That's called winging it: getting as much quality data into the file as possible, dealing with it in post production. There's nothing wrong with that. I do it all the time. Hell, when shooting landscapes I bracket all the time and then merge as needed in post production, etc.
But I don't do it for portraiture (formal people pics--stand here while I take your pic). The OP specifically mentioned portraits. When I'm taking portraits, I do glance at the LCD the first couple of shots to see if the overall image more or less resembles what I had in mind; shadows, contrast, etc. After that I don't look at it all. Many of the shots, if I were a histogram chimper, would make me cringe because the histogram is "technically" way out of whack; piled up on one side or the other--or both--but not clipping. So what? My lighting and the shot are what I saw in my head before hand.
Your method certain works, and it can produce excellent quality images. However, it may not produce the minimal noise levels, and it may result in loss of shadow or highlight detail.
Assume that we take two portraits of a model standing in a studio with an open window behind her. The model only lit by constant interior light.
We take two photos. One at noon, where the exterior is illuminated by bright sunlight, and one at night, with the exterior illuminated by moonlight.
What should the exposure be? More importantly, as the model is the important part of the image, should the exposure be the same for both, or should the two images have different exposures?
Let's suppose we meter off the model, and use that. We can process both images with the same parameters, and we will get prints where the model looks good, but the window is either all white or all back. The problem is that we have lost the background details.
However, we can use a lower exposure for the daytime exposure, this will help us retain some of the sunlit details in the window. We can use a higher exposure for the nighttime shot. This can help us retain some of the details of the moonlight exterior, without blowing out the details of the model.
The important point is that the mapping of the model's skin tone to numeric values should nor be fixed. It should vary depending on the needs of the image.
When the image is processed, the model's skin tones can be adjusted to the level we want, and the scene outside the window can be adjusted as well. This can be done manually, or through various automated processes.
If your goal is to maximize the useful detail captured, then you want to maximize exposure without blowing out important highlight details. This is much easier to do with the in-camea meter/histogram than with an external meter.
If you want to regularize the values in the raw file (i.e. an 18% skin tone should always give the same values), then you may have a simplified workflow, but you are giving up some quality.
Let me repeat, that for many images the quality difference between the two methods is small. Modern cameras are so good, that you may not need to squeeze the last drop of quality out of them.
...
However, vis a vis the original OP, for most portraiture, indoor, outdoor, strobed or naturally lit; positioning people to stand while photographing them, knowledgeable use of a handheld incident meter is almost always going to be superior to using a camera's reflective meter. And it will be quicker. No matter how advanced the camera's "tools" are.
Maybe, maybe not. Whether you are using an external meter or the camera's meter, the goal is to get the right amount of light on the sensor. The in-camera meter is in an excellent position to do that.
In portraiture, for me anyway, the goal is to get the shot as I envision it. Envision it, create it, shoot it. The in camera meter is in an excellent position to give you an image that is a nice average of everything. The handheld incident meter and an understanding of how light works is a nice way of getting the shot as you envisioned it inside of your camera.
The camera does not record the light falling on the subject. The camera records the light reflecting off the subject to the camera.
When you measure incident light, you need to make assumptions as to how that light will reflect. If the too much light reflects, then you blow out highlight detail. If not enough light reflects, then you will lose shadow detail.
Remember, if you are shooting a very bright white wall, or a dark back wall, you will get a better quality image by letting the sensor think it is mid-gray, and then adjusting the brightness in post. If you base your exposure on an incident light meter, you may end up with a blown out white wall with no detail, or a pure black dark wall with no detail.
The notion that capturing a white or black wall as mid gray and fixing it in post is best is, frankly, goofy in my opinion. Again, I'm talking portraiture. I used to do just that. Then I learned how to use my handheld meter and manipulate my light as needed (still learning as it's always a learning process) and/or move the subject to get the background I want. Making those needed adjustments takes a few moments. Using post production to create a bg over tens if not hundreds of images is far more time consuming and difficult. Well, unless you're simply doing a global adjustment which I can't imagine that being a viable option. Not for me anyway. You want that white wall? Bump up brightness, lose some detail in the subject. Or, use layers to separate the bg from the model, go through that whole thing. No, thanks.
It's only "goofy" if you want to match a particular work flow, and don't mind giving up a little quality.
Let's go back to shooting a very white wall. Use an incident light meter, and you might get blown out highlights. Use in-camera metering/histogram, and you retain detail.
Open the raw file in a program that auto-adjusts brightness, and both give you an 18% wall. However the incident light meter version has fewer details.
Open the raw file in a program that allows you to adjust the brightness, and both give a white wall, but the in-camera/histogram metered one has more details.
At the end of the day, in camera metering has a place, absolutely. But a lot of people actually create more work for themselves and limit themselves in many respects by thinking it is the end all be all.
I am tempted to say the same thing about external light meters.
Again, the notion that in camera metering and the "tools" built within the camera have rendered knowledgeable use of a handheld incident meter obsolete is absurd.
I did not say that incident meters were obsolete. What I said is that they are not required for most types of high end photography.
There are many situations where one can use either incident meters or in-camera metering. It's a matter of personal choice. However, in-camera metering and tools have reached a point, where there are very few situations where incident meters are needed.
thank you for you answer
