Internal vs external light meter reliability

[il_mix]

Hi everyone!

I have a question that bugs me since long.
Internal ligh meters get the light that goes through the lens and filter (if any), and convert the information giving exposure/aperture pair hints.
External light meters get the light directly from the target, and again convert this information.
The question is: isn't it more reliable to get the information from an internal light meter, since we are measuring the "real" light, filtered by the lens and filters, that is going to the sensor/film?
As a simple example, is said that zoom lenses are "darker" than fixed ones ("fixed" isn't probably the right word, I'm sorry. Say a 35mm lens). So an external light meter gives me a hint about the exposure/aperture pair just from the target light, without knowing anything about lens/filters used. An internal one will measure less light if I'm using a zoom lens, or if a filter is used, and gives me exposure/aperture hints for that light.

With these assumption, an internal light meter seems better than an external one.
But since there are 500$ external light meters, maybe I'm missing something...

 

[Overread]

Internal light meters can only read reflected light; that is light reflected off the subject and through the lens.

External light meters though can read the light coming from the light source direct instead of reflected.

This means that if you have a snowy scene lit by the sun the light meter in the camera will read the reflected light off the snow, whilst the external light meter can be held up to face the sun and read the light coming direct from the sun itself. This helps because it means that the light reaching the sensor isn't affected by the reflective properties of the subject.


Another feature is that internal meters can only read constant ambient lighting, whilst an external light meter can read flash lights. This makes them invaluable in studio or controlled lighting situations where you can approach the subject and hold the light meter up to read the light falling on the subject direct from the flash units. That lets you balance the light from the flash units without taking and adjusting photo after photo. Digital and the histogram review feature mean that this isn't "as" critical a feature as it was in the days of film; but its still a major help in lght balancing.

 

[il_mix]

Thanks for the infos. These are some pros about external light meters.
But again, the lens and filter attenuation does not affect the external light meter measure, right?

 

[WayneF]

Thanks for the infos. These are some pros about external light meters.
But again, the lens and filter attenuation does not affect the external light meter measure, right?

There can be a huge difference between reflected meters and incident meters. That is a rather different discussion, with its own learning process.

Reflected meters only see the light reflected from the subject. This light seen depends on how well the individual subject can reflect light. It is white or black color for example? This makes a tremendous difference to the reflected total (and causes metering errors). The reflected meter ASSUMES the scene reflects as if it averages to be about middle gray, regardless if it does nor not.

Only reflected meters can go into the camera (the lens sees light reflected from the subject). Handheld meters can be incident meters, but are not necessarily so. Many handheld meters can read incident or reflected light, your choice.

Between two reflected meters, the camera meter is the way to bet. It does see filters and lens, and only meters what the camera lens can see.

Otherwise, incident meters are aimed at the camera instead of at the subject, and they meter the actual light incident on the subject. Metering at the subject is more awkward, but it is the actual way to bet. Esp in the studio.

Maybe see How light meters work (page 3 is incident metering).

 

[KmH]

Thanks for the infos. These are some pros about external light meters.
But again, the lens and filter attenuation does not affect the external light meter measure, right?

Yes but both light meters are calibrated to compensate for the attenuation each system has.

 

[il_mix]

Well, no doubt that I need to study the reflected vs direct light metering and usage first.
I always used the integrated meter, and never thought that an external one can work in such different way.
Back to the books...

Many thanks to everyone!

 

[bratkinson]

Back when I was shooting 35mm slides eons ago, my meter was frequently used to make sure I got a good exposure. When I was still somewhat in doubt, such as nighttime shots outdoors, I'd bracket the exposure a couple of stops each way just for 'insurance'.

But since switching to digital about 10-12 years ago, the internal camera meter(s) have worked just fine for perhaps 98% of my work, both indoors and outdoors. The times it is not right is when there's one or more bright lights in the scene along with a large, darker area that is really the major focal point(s) of my shot.

The beauty of digital photography is that one can immediately see the results. Looking at the histogram is also a great help in getting the right exposure. In essence, amateur that I am, I've not found a good reason to buy a separate light meter when the one in the camera does very well and it 'considers' the effect of any filters I may have in front of the lens.

 

[il_mix]

I'm shooting 35mm nowadays...

 

[il_mix]

Long story (really) short.

Internal light meters: measure reflected light, fitered by lenses and filters. Measure referenced to a standard gray target. Adjustments must be made when the target scene if far from this condition (e.g. snowy landscape); adjustent experience-based (the number of exposure steps to change vary given the offset from the standard gray condition)
External light meters: measure incident light on the target and/or on interesting areas inside the scene, are usually more accurate that internal light meters, ignores info about lens/filter attenuations. Adjustments must be made according to the light attenuation introduced by lens/target; adjustment depend just on the attenuation factor of the used gears, so it is constant and independent from the scene.

May I send this iper-detailed essay to the ICP?

 

[KmH]

The in-the-camera reflected light meter usually offers 3 or more modes - 1 that averages the light in the entire scene (Evaluative/Metrix), 1 that provides an weighted average of the central potion of the scene (a weighted average of about 70%) (center-weighted), and 1 mode that meters a very small portion (about 2%) (Spot).
Spot metering mode is the mist accurate.
The in-the-camera reflected light meter usually has a default setting that allows adjusting exposure triad settings in 1/3 stop steps. DSLR cameras usually allow changing that to 1/2 stop or full stop steps.

Digital hand held meters bt default commonly display meter readings in 1/10 stop increments, and when used in reflected mode does not offer measuring modes comparable to in-the-camera Spot, Center-weighted, or Evaluative.

 

[hirejn]

A hand-held meter, which is precise down to 1/10 stop, is technically capable of more accuracy than a camera's reflective meter, and finding the exact correct exposure with a reflective meter, even with a supposed 18% gray card, still has a small margin of error, whereas an incident meter will measure the exact amount of light every time. It's true the built-in camera meter will automatically adjust for filters. However, lenses don't affect metering. The reading that you get on a hand-held meter can be applied to any camera and lens combo, from pocket cams to large format.

One of the most important differences is that the camera meter changes depending on where you point it, so it's difficult to tell exactly what the correct exposure is. Since it measures light reflected off of surfaces, the reading changes because different objects reflect different amounts of light, even if the light source is constant. However, if the source is constant, the exposure should also be constant, and if you could find that exposure, you could get accurate shots of everything under the same light with a single setting. That's what incident metering does. Incident metering finds the midtone exposure for the given light, rendering all objects under that light tonally accurate. Reflective metering finds the midtone of the object, averaging all objects or scenes to gray, which is definitely not tonally accurate in every case since not all objects are gray. Reflective metering was meant to be a convenience for those who didn't want to carry and use an incident meter; but you must understand how it works and its limitations.

You're definitely missing some of the things hand-held meters do, such as display the dynamic range of the scene, calculate differences between readings, trigger and adjust power on remote flashes, and of course meter flash and help establish precise lighting ratios, none of which the camera's meter can do. But there's no one tool or method that's "best" for everyone. Photographic equipment is just a tool, like a paint brush or palette knife. How you use it, why and whether you like it are completely subjective.

 

[WayneF]

A hand-held meter, which is precise down to 1/10 stop, is technically capable of more accuracy than a camera's reflective meter, and finding the exact correct exposure with a reflective meter, even with a supposed 18% gray card, still has a small margin of error, whereas an incident meter will measure the exact amount of light every time. It's true the built-in camera meter will automatically adjust for filters. However, lenses don't affect metering. The reading that you get on a hand-held meter can be applied to any camera and lens combo, from pocket cams to large format.

Overall, no disagreement, but some details seemed unclear. You imply the camera meter is less accurate, and the hand held is more accurate. I think that is far from true. It is not about accuracy, but about precision. Nikon meters to 1/6 stops units, but the camera can only be adjusted in 1/3 stops (flash power can be set more fine). That is the difference in precision and accuracy. Even if the hand held meter reads out in tenth stops, the camera and flash can only be adjusted to third stops. This is not about accuracy. We really don't have a clue about accuracy.

The meter accuracy is not at all in question, but the reflected scene certainly is in question.. With any reflected meter, an all white scene will reflect much light, and and the camera will see a lot of light, and will underexpose it. An all black scene will Not reflect much light, the meter sees much less, and the camera will overexpose it. Yes, both are trying for the middle gray result, which is all the meter knows to do. The actual scene is an unknown to the meter. The dumb meter cannot recognize a black cat in a coal mine from a polar bear on the snow, and both will come out middle gray, if the meter is accurate.

But the lens definitely does has effect, generally a good effect. For example if using a telephoto lens, metering the scene with a wide angle hand held reflected meter can be pretty far off, being affected by stuff the lens does not even see (like for example, bright sky about the telephoto view of the smaller subject). TTL means Through The Lens, which means the meter only meters what the lens sees, which is all that is important, and is a big plus over handheld meters (in that one aspect, comparing reflected meters). Due to this lens angle, I would always bet on the camera meter being more accurate than a hand held reflected meter, but any reflected meter is suspect. It is not about accuracy of the meter, it is about any reflected meter not having any clue about what it is seeing.

The difference is NOT about hand held vs camera meter (except the viewing angle of one does match what the lens sees, which is a big plus). The significant difference is reflected meter vs incident meter. The incident meter is about the light falling on the subject. The reflected meter is about the unknown reflectance of the subject. The camera meter can only be reflected. The hand meter might be incident, but is not necessarily so, but many offer either choice (and can meter flash too, which is the major importance of a handheld meter).

The advantage of an incident meter is that it meters the light, not the subject. It is aimed away from the subject, at the camera. It measures the actual light ON THE SUBJECT. Any subject (black or white) in that light will come out pretty close (will appear black or white, instead of middle gray). That is not about meter accuracy, it is about subject independence. Incident metering is independent of the subject. Other forms of subject-independent exposure are Sunny 16, or flash Guide Number, for example.

 

[il_mix]

This thread is not about which light meter is better, of course. I know that internal and external one have different pros and cons (that I'm still learnig about...) and are more useful in one context or another.
I have a couple of questions regarding, again, how lenses interfere with light meters.

Reading your posts it seems that there is no light attenuation introduced by the lenses, or better that there are no difference (or not appreciable difference) in light attenuation between different lenses. So, the argument that a zoom lens is "darker" than a fixed focal length one is just a urban myth?

Second thing. WayneF said that a different focal length, and so the different field of view af lenses, change the internal light measure of the scene since a different amount of reflected light goes through the lens. This is stated as a pros of internal light meters, that measure the light in the area of interest. But, since external one can measure the light that strike the target, I am, as said, measuring the light that strikes the target (say, I'm placing the light meter on the head of the subject of my portrait to measure the light), so you can't compare the two situations. This, if I've well understand how an incident light meter works... Maybe you are talking about using an external meter "like" an internal one? (e.g. mounting the light meter on the body of the camera like this).

Thanks again

 

[WayneF]

Reading your posts it seems that there is no light attenuation introduced by the lenses, or better that there are no difference (or not appreciable difference) in light attenuation between different lenses. So, the argument that a zoom lens is "darker" than a fixed focal length one is just a urban myth?

Light passing through any glass causes a little loss, primarily due to a little of the light bouncing back off of the lens surface instead of just passing through. It was a big issue in the early days, when multiple lens glass elements all added up multiplying this loss. However today, lens have modern anti-reflection coatings, which greatly reduces this loss to nearly zero (also minimizes internal visible reflections and flare). But still, some modern zoom lenses with say 15 glass elements (30 surfaces) does lose a bit more than a simple lens with say 4 glass elements. Worst case today probably won't total to be 1/3 stop, which is the minimum adjustment the camera can make anyway. Hollywood movies do use lenses calibrated in T-stops instead of f-stops, which is about the actual transmission instead of the general math. But movies might switch lenses in one continuous scene where 1/3 stop might be visible, where still photos taking only one picture will never "see" a switch.
Good modern coatings make it be an extremely minor thing TODAY, generally not of any interest now (it is solved). Lenses have modern coatings today (cheap filters may not be so good).

Second thing. WayneF said that a different focal length, and so the different field of view af lenses, change the internal light measure of the scene since a different amount of reflected light goes through the lens. This is stated as a pros of internal light meters, that measure the light in the area of interest. But, since external one can measure the light that strike the target, I am, as said, measuring the light that strikes the target (say, I'm placing the light meter on the head of the subject of my portrait to measure the light), so you can't compare the two situations. This, if I've well understand how an incident light meter works... Maybe you are talking about using an external meter "like" an internal one? (e.g. mounting the light meter on the body of the camera like this).

Hand held meters can be reflected or incident. Beginners asking which (camera or handheld) generally don't know about incident meters, which is the actual advantage of a better hand held meter.

Also the overwhelming advantage of most hand held meter use is to measure flash in the studio (camera meter cannot, excepting TTL, which is very special purpose and limited).

So, a hand held meter can be an incident meter (which camera meter is not), and a hand held meter can be a flash meter (which the camera meter cannot, generally).

If metering your studio portrait flash picture, the camera meter cannot do it (the Nikon Commander TTL excepted, but limited). Flash surely represents the large majority of hand held meter usage today.

If just metering your landscape picture, the way to bet in general about metered "accuracy" is:

1. Best: Incident meter, which must be hand held, and is metered at the subject, meter aimed at the camera. Meters actual light level. Light falling outside the actual lens view is of no interest or importance. Best because it meters actual light level on the subject, and it is independent of the unknown subject reflection characteristics.

2. Reflected meter in the camera, behind the shooting lens (meters only what the lens will see).

3. Worst: Hand held reflected meter, which is generally wide angle, as compared to the probable lens used.


EDIT: For example of flash meter in the studio, with four manual lights, main, fill, background, and hair light. With a hand held incident flash meter, we meter at the subject (flash intensity falls of with inverse square law). We meter each light individually, at its subject (background fabric in that case), and set its manual flash power level to give the level we want (the ratio we want). Maybe we set main light to meter f/8. Fill light to be maybe f/5.6. Background light maybe f/8, depends). Hair light depends on hair color, we sort of set it by inspection, black hair maybe a stop more than main, blonde hair maybe a stop less than main light. Then we meter the main light and fill light together to set camera aperture. Then we are certain what the lights are doing.

The camera meter cannot do any one step of that, but the incident flash meter makes it be trivial (and repeatable). So hand held meters are quite popular, but possibly not for newbies still trying to learn the what's what about the camera and the exposure. Beginners ought to try to learn how the reflected light meter actually works though.

 

[il_mix]

Maaaaaaaany thanks WayneF. Now the topic is more clear, and I can make my personal decision on the device to use.
Since I'm mainly interested in landscape/architectural pictures, rarely in portraits, and I don't use flash at all (I decided to shot only daylight outdoor photos for a couple of years, and then start studying flash photography), my preference goes to the internal meter.
Now, I just have to let it work correctly... (old camera issues).