Advice on using a lightmeter

Status
Not open for further replies.
I shoot medium format film out of choice. I started with digital but preferred pretty much everything about using film, I honestly think it's made me a better photographer, technically at least, than I was before.

Understanding your exposures should never be underestimated, I totally agree with what you're doing and wish you the best of luck with it :)
 
Hoppy I understand what you are saying but surely when you say that you would be daft not to rely on the camera you are then relying on reflected light?

For me I want remove the guesswork and chimping and as much of the pp work. I use custom wb most of the time and want to make sure my exposures are bang on also.

Reflected light is what you see, what the camera records, and the only thing that matters.

What an incident light reading aims to do is take some of the variables out of the equation and make it easier to put a mid-grey tone in the middle of the histogram. It's the same as taking a reflected reading of an 18% grey card. This is perfect for slide film, but not optimum with either neg film or digital, depending on how you want to play it (eg expose to the right) and what your interpretation of correct exposure is. Compared to a 'correct' incident meter reading, with ETTR technique my exposures are usually between one and two stops more than the meter says.


And the mid-grey thing also assumes that a) you get the reading right, and b) that the camera settings are absolutely accurate when in practise they can vary quite a lot from marked settings. Eg, ISO800 on one camera may not be the same as ISO800 on another, f/numbers can be way out compared to actual transmission values and aperures can be both inaccurate and inconsistent especially at higher f/numbers.

There are a number of variables and if you're unlucky and they all compound together, you can easily be a stop out. I recall a thread a few months ago when in a studio situation the poster queried why two cameras and two lenses were giving different results, and neither matched the meter reading by rather more than that.

The histogram and blinkies read the actual image so any variables are 100% netted out and what you see is what you get. If you turn down the JPEG pre-sets (Picture Styles etc) and in particular the contrast setting to minimum, what you get on the histogram/blinkies is extremely close to what's on the Raw file, near as makes no difference. It's as close as you can get.
 
What an incident light reading aims to do is take some of the variables out of the equation
I would say that it takes all of the variables out of the equation, in that an incident light reading reads only the light incident upon (falling upon) the subject and so provides a value that will render all tones correctly. So, the meter reading, if taken correctly, will be right.

But it's just a reading. Just as our car speedometer (or more likely our satnav) will accurately tell us what speed the car is travelling at, it can't and won't tell us whether that speed is safe. It's just data and it needs to be interpreted.

I was running a lighting workshop today. It was an art nude workshop and I tend to use what most people might regard as unconventional lighting, and because it was art nude I often choose to deliberately underexpose images to some degree. Now, most of the shots were fine at the measured exposure, but a couple were way out, in the sense that they simply looked much better at a very different exposure - I'm talking f/5.6 indicated and f10 chosen here.

So, common sense, experience and a degree of subjectivity all have a part to play.
 
I would say that it takes all of the variables out of the equation, in that an incident light reading reads only the light incident upon (falling upon) the subject and so provides a value that will render all tones correctly. So, the meter reading, if taken correctly, will be right.

But it's just a reading. Just as our car speedometer (or more likely our satnav) will accurately tell us what speed the car is travelling at, it can't and won't tell us whether that speed is safe. It's just data and it needs to be interpreted.

I was running a lighting workshop today. It was an art nude workshop and I tend to use what most people might regard as unconventional lighting, and because it was art nude I often choose to deliberately underexpose images to some degree. Now, most of the shots were fine at the measured exposure, but a couple were way out, in the sense that they simply looked much better at a very different exposure - I'm talking f/5.6 indicated and f10 chosen here.

So, common sense, experience and a degree of subjectivity all have a part to play.

It can't take account of all the variables Garry - see para three above.

To continue your car speedo analogy, if you take three cars at random and drive them all at 30mph and then at 100mph as indicated by the speedo, while at the same time comparing them to a sat nav, you will get six different numbers. Sometimes alarmingly different.

The difference being that the speedometer has to estimate a number of variables that it doesn't know anything about, such as tread wear, rolling radius and dynamic diameter. And that's assuming that the correct profile tyres have been fitted (not that they are always exactly what's written on the side*). Likewise, the meter doesn't know anything about T-stops, or the sensor's true sensitivity, or how accurately the lens stops down the the selected aperture.

The LCD/histogram/blinkies take all these things into account, in the same way that the sat nav does - if it's ultimate accuracy that you want.

* Sadly, I have checked all these things in a previous life :D
 
A light meter, and it's correct use, can show you where the highlights are, shadows, mid tones, in fact every value in the shot. Where they are, not simply the fact they are there, a la histogram...

Use the tools you've got, I agree, but theres nothing wrong with using a light meter, understanding and visualising an image before hitting the shutter should be encouraged, and that includes lighting not just composition.
 
I'm intrigued by the pointing the meter at the camera option, that would have some major problems as far as I can see, however I'm happy to be wrong and to learn in the process :)

Take this recent maternity shot, this is a single light to the side and behind the subject and one on the background. By pointing the meter at the camera I would be totally blown out as the dome wouldn't be reading the light??? I think, or am I missing something?

Tasha-and-Bump-0002.jpg
 
I'm intrigued by the pointing the meter at the camera option, that would have some major problems as far as I can see, however I'm happy to be wrong and to learn in the process :)

Take this recent maternity shot, this is a single light to the side and behind the subject and one on the background. By pointing the meter at the camera I would be totally blown out as the dome wouldn't be reading the light??? I think, or am I missing something?

Common sense has to play a small part in the equation. You took the decision to correctly expose the dome, and therefore metered for it. If someone else wanted the exact same setup (lighting and pose) but wanted the camera side of the subject correctly exposed, and the edge of the dome effectively rimlit (or blown out, as you eluded to), then the meter reading would be taken from the subject facing the camera.
You meter for the effect and result you have in mind.
 
I'm with that, just making sure I didn't miss something. My usual method is point towards the 'main' light source (usually sun for me or light direction if I'm in shade), if something is fully back lit then it's towards the camera.

And I assume you get the results you are after. (y)

Which is what it's all about.
I meter to camera, generally speaking. I meter to light when appropriate. I don't meter as often on location, as it's often "seat of the pants" stuff. :LOL:
 
Having just really started to use a meter i have really enjoyed reading through the posts,so there are a lot of different options using a meter,great posts.(y):)
 
mark1616 said:
I'm intrigued by the pointing the meter at the camera option, that would have some major problems as far as I can see, however I'm happy to be wrong and to learn in the process :)

Take this recent maternity shot, this is a single light to the side and behind the subject and one on the background. By pointing the meter at the camera I would be totally blown out as the dome wouldn't be reading the light??? I think, or am I missing something?

Mark, I agree with you lol,
A light meter should be pointed at the light lol. But if pointing a meter at camera works for others, then cool, as long as it works (y)
 
GGRRRRRR LOL!

:bang::bang::bang::bang::bang::bang:
As Michael Winner says in the car insurance advert...
"CALM DOWN DEAR"
It's just the internet, everyone is entitled to an opinion even if they don't understand the issues and even if they choose to ignore the facts...
 
Gary, I honestly want to learn, so if you have some of the physics behind things I would like to know. I point at the light source (assuming it is illuminating a part of the subject in question) rather than pointing at the camera as in my maternity shot posting.

Happy to learn :)

OK, an explanation from Garry...:LOL:
If you point the meter at the light from the subject position then the meter measures the light incident upon the meter sensor. let's call that f/11
But (usually) the light isn't where the camera is.

The light strikes the subject and bounces off at the same angle (the angle of reflectance = the angle of incidence) and although some or most of the light will bounce towards the camera, not all of it will. Therefore, the camera will receive less light than you expect.

But if you point the meter towards the camera then the meter sensor (receptor) will be at (roughly) the same angle as the camera and so will receive roughly the same amount of light that the camera receives. If the camera is in a position that receives only half as much light, the meter reading will take account of that and f/11 becomes f/8, and f/8 will be the technically correct exposure - or as near to it as you can get.

Hope that helps, Garry
 
Oops..... I dropped an R sorry!

That makes sense. What happens in the situation like my maternity shot where no light would hit the dome as it is behind the subject, would you then turn fully to the light or take an angle so it is partly towards the light or partly towards the camera?
 
r, so you did - why didn't I notice that?:)

The light wasn't behind the subject in that maternity shot, it was in a typical rimlight position - at the side and a bit behind, and a meter placed against the part of the body that you wanted to render correctly in terms of exposure and pointed at the camera would have received plenty of light and would have produced a correct reading, or at least as close to correct as these things can do.

But pointing it directly at the light would work too, in that it would measure the light. As a photographer, you then need to interpret that data and make allowance for the light that has shot off at an angle not received by the camera. My guess is something like 2 stops. Am I right in that guess?

I think it comes back to the point I made much earlier. The meter is a tool which, when used correctly, produces a theoretically correct reading that amounts to no more than data which we, using our experience and understanding of basic physics, then interpret.
 
I think this discussion needs to start off with "what does properly exposed mean" and then "what does correctly illuminated mean"

Both are more than a simple setting (i.e. F8, 1/60 ISO100)... you have to consider the desired artistic result, the tonal range of the subject, the tonal limitations of the film or sensor (dynamic range), the colour and reflectivity of the subect, etc. etc.

Incident metering measures light falling on the subject, so for example if you shoot a bride a groom and a best man next to each other, each holding a similar bunch of flowers - one with a white dress, one with a grey suit and one with a black suit, in theory, the incident reading will be "about right", the black, grey and white will be rendered well, and the flowers will all be rendered the same, and the colour will look good

In practice, you still may need to tweak the incident reading to accommodate the limitations of your camera sensor or choice of film

The reflected reading, will vary between the white dress, the grey suit and the dark suit and the flowers, so with that measurement, you can take your pick. Meter for the dress, and under expose everything else, meter for the black suit, and over expose everything else

Enter the 18% grey card... In this instance, we attempt to measure the reflected reading of of something that "generally makes things look OK"

Here is the rub - camera's internal sensors measure the reflected reading. If you pop it on single spot metering, you can do exactly what I said, and get the same results for the black, grey and white clothing, and then if you are frustrated you can also pop a grey card in the scene to try and improve things

Here is the second rub - if you meter reflected light, it is the same for the spot you metered at any distance (just about) If you are metering a light source, the ISL applies. so if you take a reflective meter reading of a big green tree as f8, and walk 100 meters away, it will still expose properly if your camera is set to f8. If you consider a street lamp next to the tree, and meter it to be say F32 (to expose the bulb perfectly), you will notice that if you meter it again at 100M, it will have a considerably different reading, as the light falling on the meter's sensor will have diminished. In this example, if 10M you expose the bulb in the light perfectly, of course the tree will be black, if you did the same at 100M away, the tree will likley be a dark green grey (not so under exposed

Hence my observation about "what does properly exposed mean"

A light-meter / flash meter is just another tool for measuring this with, understanding the results are a different kettle of fish

Going back to my black, grey and white dress and 3 similar bunches of flowers, with modern cameras, and difficult or wrong lighting, it is very easy to both blow the dress and loose the details in the shadows. This is where a meter really helps. If your film or sensor has a dynamic range of say 8 stops, and the objects you are photograph meter out with a range of say 12 stops, you need to do something, and when you have done something, you can check again and get your exposure set so the darks and lights are not destroyed, or if they are, you know why and where
 
Last edited:
I agree to a certain extent Garry but I am also an advocate of metering towards the light. If for example an accent/kicker light is catching the right side of a subjects head then there will be a highlight there if brighter than the key. Now if meter that towards camera, my meter might not pick up the point on the head where that is brightest, my eyes cannot always see that . If I meter towards the light from the point of highest contrast, the meter reading I will get will be accurate as it is measuring the light hitting the subject at that point. Now the difference maybe enough that I could overexpose the highlight if I only take the reading of the light towards camera whereas if I meter towards the light and use that reading then assuming everything else being equal that reading might not be right ... ( for the given exposure I want so as not to get involved in a what is a correct exposure) .

I like the discussion in here :)
 
Gary, what you're saying is that you have to interpret the results. And that's right.
But I'm guessing (from your other active thread) that you just prefer to do it that way and that you know that the physics is wrong.

Photography is a strange, eclectic mixture of physics, experience, practice and personal preference. Different people place different emphasis on different parts of that mix, and there's nothing wrong with that.
 
I agree to a certain extent Garry but I am also an advocate of metering towards the light. If for example an accent/kicker light is catching the right side of a subjects head then there will be a highlight there if brighter than the key

Is that because you are lighting to a ratio?

The confusion sometimes is that while the ratio may be 3:2 you may still want to set the camera to the meter equivalent of 3 (or 2 or 2.5)
 
Not at all Garry, I totally understand the physics but I do work on ratios, whatever I want them to be, if I only meter towards camera I may miss that highlight and therefore cock up my exposure. I am not alone in this thinking but we all have methods that work for us.
 
Sorry, but you're wrong.
If you meter towards the light then you theoretically get a true reading of the output of that light, but if the camera is in a different position to the light then less light will reach the camera. Read up on cosine law if you want to understand the theory.
So, if you want the exposure to be closer to correct, point the meter towards the camera, because it's the amount of light that reaches the camera, not the amount of light that reaches the meter, that matters.

Some people believe that it's best to point the meter midway between camera and light, and they may be right - if using a flat receptor. But as I keep saying, it normally makes very little difference when using a dome receptor.

OK, so I'm going to follow your theory here and meter towards the camera and follow your argument about physics as you're a trainer and are teaching, perhaps this will be something new to me.

I'm also going to meter towards the light for my setup - which is exactly as I've been taught and how Mr Elliott is positing.

Time to see which one is right.........


Split-lit-meter-to-camera by matty27272, on Flickr

This is badly blown all down the left side when I metered to camera.


Split-lit-meter-to-light by matty27272, on Flickr

This is NOT badly blown when I metered towards the light source.

CONCLUSION of the simple experiment

If you want the meter reading to be right and not give you blown highlights, meter towards the lightsource and not the camera. Any argument that you only need to meter to the light source in certain situations is going to set you up to fail when you forget which situations to do it in. Your meter may well have a dome which reads light from 180 degrees, but unless you ALWAYS have your lights in the same position, you're going to forget and end up with a potentially expensive mistake.

If you don't want to believe me, just set up your lights (or single light) and prove it to yourself. Move your light to different positions and do 2 shots on each light position. One towards camera and one towards light source. Keep doing this until you run out of space on your memory cards or patience for the experiment.

My mentor tells me how to do things, shows me them working and then sets me homework to do which involves me proving the things to myself afterwards. I suggest that this is the way to do things. See the setups and arguments put forth by people who are offering training, watch their proofs and then prove it to yourself.
 
I think this discussion needs to start off with "what does properly exposed mean" and then "what does correctly illuminated mean"

One decides what type of result one desires and then one gets ones meter out and points it in the correct direction.

In a discussion of which direction to point your meter, what properly exposed means is absolutely irrelevant. It has already been decided before you get the camera out of the bag, or off the shelf, or off the floor where you put it after the last shot, or wherever it might be placed.
 
A challenge! I like a challenge:)

I'll take some example shots and post them, to illustrate why you got the results you did.

Watch this space...
 
A challenge! I like a challenge:)

I'll take some example shots and post them, to illustrate why you got the results you did.

Watch this space...

Please do. Once you get them, please also post the setup and an explanation of why I got the results I did. I know why I got them, but I'm sure other readers of the topic will enjoy the explanation too :)
 
One decides what type of result one desires and then one gets ones meter out and points it in the correct direction.

In a discussion of which direction to point your meter, what properly exposed means is absolutely irrelevant. It has already been decided before you get the camera out of the bag, or off the shelf, or off the floor where you put it after the last shot, or wherever it might be placed.

We seem to be going round in circles on this one. I will hazard a guess and say that some people use a meter to determine exposure, and others use a meter to check the image matches the exposure they were aiming for. Others use the meter for checking a lighting ratio, and others to check is a background is evenly illuminated. For others it is a combination of most of the above

Whatever the reason a few things shine out of the comments
- there often is more than one way of doing things
- often the reading you end up with ends up being "interpreted" or adjusted anyway
 
OK, so I'm going to follow your theory here and meter towards the camera and follow your argument about physics as you're a trainer and are teaching, perhaps this will be something new to me.

I'm also going to meter towards the light for my setup - which is exactly as I've been taught and how Mr Elliott is positing.

Time to see which one is right.........


Split-lit-meter-to-camera by matty27272, on Flickr

This is badly blown all down the left side when I metered to camera.


Split-lit-meter-to-light by matty27272, on Flickr

This is NOT badly blown when I metered towards the light source.

CONCLUSION of the simple experiment

If you want the meter reading to be right and not give you blown highlights, meter towards the lightsource and not the camera. Any argument that you only need to meter to the light source in certain situations is going to set you up to fail when you forget which situations to do it in. Your meter may well have a dome which reads light from 180 degrees, but unless you ALWAYS have your lights in the same position, you're going to forget and end up with a potentially expensive mistake.

If you don't want to believe me, just set up your lights (or single light) and prove it to yourself. Move your light to different positions and do 2 shots on each light position. One towards camera and one towards light source. Keep doing this until you run out of space on your memory cards or patience for the experiment.

My mentor tells me how to do things, shows me them working and then sets me homework to do which involves me proving the things to myself afterwards. I suggest that this is the way to do things. See the setups and arguments put forth by people who are offering training, watch their proofs and then prove it to yourself.

You make a point, however your experiment is slightly flawed, and you haven't mentioned the effect on the other tones in the scene (i.e. the black)

It would be more representative and useful to shoot a group of objects with more tonal range in them. Hence my comment way back about deciding what correct exposure actually means

Being devils advocate, one could argue that maybe the dark surface in the scene in the first shot is correctly exposed and the white is blown because the scene has a wider dynamic range than your sensor can handle. That's the issue with exposure arguments, sometimes the answers also are interpretative
 
Please do. Once you get them, please also post the setup and an explanation of why I got the results I did. I know why I got them, but I'm sure other readers of the topic will enjoy the explanation too :)

I'm planning to set up a couple of different lighting arrangements, metering both to the light(s) and to the camera on each one, I'm planning on showing setup shots to avoid any misunderstandings and then just publish the results on a blog - should be interesting.

We seem to be going round in circles on this one. I will hazard a guess and say that some people use a meter to determine exposure, and others use a meter to check the image matches the exposure they were aiming for. Others use the meter for checking a lighting ratio, and others to check is a background is evenly illuminated. For others it is a combination of most of the above

Whatever the reason a few things shine out of the comments
- there often is more than one way of doing things
- often the reading you end up with ends up being "interpreted" or adjusted anyway
Basically yes, and I've already said so. But the laws of physics are immutable and so measuring the light in accordance with those laws, instead of disregarding them, makes sense to me.

Once we have true data, we can interpret it and make adjustments to get the effects we want
 
Matty, totally agree with your way of thinking...looking forwards to you arguments Mr Edwards....

I'm not actually offering any arguments.
From memory, it was that nice Mr. Newton who came up with the argument, and as he was a lot brighter and better informed than me I have the good sense not to try to re-invent a perfectly good wheel.

But Mr. Newton didn't have access to a blog and I do, so I'll be sure to post the results of my verification of Mr. Newton's research:)
 
We seem to be going round in circles on this one. I will hazard a guess and say that some people use a meter to determine exposure, and others use a meter to check the image matches the exposure they were aiming for. Others use the meter for checking a lighting ratio, and others to check is a background is evenly illuminated. For others it is a combination of most of the above

Whatever the reason a few things shine out of the comments
- there often is more than one way of doing things
- often the reading you end up with ends up being "interpreted" or adjusted anyway

The points you're raising are clouding the issue.

The question raised in the thread is which direction is the correct one to point one's lightmeter in when measuring flash.

Any other discussion is beyond the scope of that and just clouds the answer to that question.

Proof of which direction to point the lightmeter is above. When the lightmeter is pointed at the lightsource, the subject isn't blown out. When the lightmeter is pointed at the camera with neither the subject or the light being adjusted, the meter reading gives a blown out exposure. It really is that simple.
 
You make a point, however your experiment is slightly flawed, and you haven't mentioned the effect on the other tones in the scene (i.e. the black)

It would be more representative and useful to shoot a group of objects with more tonal range in them. Hence my comment way back about deciding what correct exposure actually means

Being devils advocate, one could argue that maybe the dark surface in the scene in the first shot is correctly exposed and the white is blown because the scene has a wider dynamic range than your sensor can handle. That's the issue with exposure arguments, sometimes the answers also are interpretative

The argument is flawed in what way? The post answers the question of which way to point the lightmeter very simply.

Having a more varied range of tones in the subject would do what exactly? Please explain this as I'm missing any point you're trying to make with the post. I'm very happy to learn something new.

Your devils advocate thing is just clouding the question. What possible reason is there for talking about exposing the background when the point made by the post is about blowing highlights on the subject? Please educate me.
 
OK, so I'm going to follow your theory here and meter towards the camera and follow your argument about physics as you're a trainer and are teaching, perhaps this will be something new to me.

I'm also going to meter towards the light for my setup - which is exactly as I've been taught and how Mr Elliott is positing.

Time to see which one is right.........


Split-lit-meter-to-camera by matty27272, on Flickr

This is badly blown all down the left side when I metered to camera.


Split-lit-meter-to-light by matty27272, on Flickr

This is NOT badly blown when I metered towards the light source.

CONCLUSION of the simple experiment

If you want the meter reading to be right and not give you blown highlights, meter towards the lightsource and not the camera. Any argument that you only need to meter to the light source in certain situations is going to set you up to fail when you forget which situations to do it in. Your meter may well have a dome which reads light from 180 degrees, but unless you ALWAYS have your lights in the same position, you're going to forget and end up with a potentially expensive mistake.

If you don't want to believe me, just set up your lights (or single light) and prove it to yourself. Move your light to different positions and do 2 shots on each light position. One towards camera and one towards light source. Keep doing this until you run out of space on your memory cards or patience for the experiment.

My mentor tells me how to do things, shows me them working and then sets me homework to do which involves me proving the things to myself afterwards. I suggest that this is the way to do things. See the setups and arguments put forth by people who are offering training, watch their proofs and then prove it to yourself.

Be nice to know where you metered from (also the position & size of the light source) for the first image. Based on where the shadows are falling it looks as though the light source was slightly to the rear of the subject. So if metered from front/centre it would have been in partial shadow so wouldn't be able to meter correctly.

Paul
 
I'm planning to set up a couple of different lighting arrangements, metering both to the light(s) and to the camera on each one, I'm planning on showing setup shots to avoid any misunderstandings and then just publish the results on a blog - should be interesting.

Basically yes, and I've already said so. But the laws of physics are immutable and so measuring the light in accordance with those laws, instead of disregarding them, makes sense to me.

Once we have true data, we can interpret it and make adjustments to get the effects we want

True data? I'm sorry, but you'll have to explain to this simpleton what you mean by that. If you mean that you'll post lighting setups that others can put into practice themselves to see why things work the way they do, then I welcome that.

The laws of physics are immutable, yes. That much is very true. Measuring the light in accordance with them is exactly what I did. I measured how much light was hitting my subject. I saw no reason to point the meter at anything other than the light which is pointed at the subject as no other direction, such as towards camera, is going to give me a good measure of how much light is hitting the subject and thus reflecting off and hitting my sensor.

How you can explain that pointing the meter in a direction other than at the lightsource is going to give you a reading of how much light is hitting the subject from that source is beyond my understanding. I'm really looking forward to this.
 
Be nice to know where you metered from (also the position & size of the light source) for the first image. Based on where the shadows are falling it looks as though the light source was slightly to the rear of the subject. So if metered from front/centre it would have been in partial shadow so wouldn't be able to meter correctly.

Paul

Light source was a large dish in same plane as subject. Aimed slightly forward to feather the light across the subject and not hit the background (ambient in the room gave some small illumination on that).

Metering was taken from the nose of the polystyrene head pointing towards the lightsource and then from the same position towards the camera.

Your point that the front of the subject would be in partial shadow is absolutely correct and illustrates EXACTLY why the lightsource is the direction the meter needs to be pointed. This point is why I gave the head a split-lit (close enough for a quick and dirty point-prover) pattern instead of something more frontal.

When the meter isn't pointed at the lightsource, it isn't picking up the light which is hitting the subject and thus reflecting off to the sensor.

Metering really IS that simple
 
Status
Not open for further replies.
Back
Top