Light meters, can someone explain?

[Cocohoney]

Do I need one and how do I use the information they provide?

 

[Denyerec]

I love google!
This short video will explain more, more quickly, than I can type: http://www.wonderhowto.com/how-to-light-meter-161895/

 

[Cocohoney]

Thank you, have bookmarked that for when I am more clearheaded, too tired now!

 

[AliB]

My short answer is that as usual it depends what you shoot.

Have a little look at the Sekonic website. It explains in detail what the differences is reflected vs ambient are and where and when you might use one.

But if you want to shoot in a studio the answer is yes!

 

[awp]

Do I need one and how do I use the information they provide?

You already have one - built into your camera. You probably don't need a separate meter unless you're doing studio work with flash.

 

[AliB]

There are times where your camera meter can easily be fooled though and that's where the incident meter can come in very handy

 

[unclened]

i recently bought one, and I found that the lightmeter told me exactly what the camera did when I was metering off my hand and exposing by +1, but it does help me in tricky lighting situations when I have to balance a few different light sources.

 

[AliB]

I like using it with backlit subjects and window lit portraits. It provides information that the camera meter really does struggle with.

Have a good read up on incident meters and decide if it's something you want to explore

 

[cowasaki]

I didn't have one for ages and for most of the time your camera will be fine without one. The time when the light meter comes into its own is where you are using an off camera flash be this a flash on a radio trigger or a studio flash.

The idea is that you set the light meter to your ISO and shutter speed (the shutter speed is set to open for the duration of the flash) then press the trigger whilst holding the light meter against the thing that you want to shoot and pointing towards the camera. The meter will then give you the correct exposure. This sounds far more complicated than just pointing you camera at the subject but you use it when you want to do something slightly different. (It sounds more complicated than it actually is though!)

This is a shot of my daughter. The camera is hand held with a flash sat on a tripod firing into a FITP softbox......

I have shot this photo at night into a car park at the side of my house. The flash was as near to her as I could get whilst not being in the shot and so the flash lit her up but not the background (which was some way away). The general point of this kind of shot is the pitch black background.

You could add to this with a second flash to the other side with less power just to highlight her edges. Again you could use the light meter just to check that the level is right on that but to be honest on digital you can just check.

If I had not had the light meter then I would have had to guess the exposure because the camera cannot meter for a flash that has no connection with it.

By the way, I am using a Sekonic LS308S which costs about £130

 

[HoppyUK]

You already have one - built into your camera. You probably don't need a separate meter unless you're doing studio work with flash.

There is nothing that you can do with a hand meter that you cannot do with the one in the camera - except manual flash metering. They are very handy for that; it's the only time I ever use one.

Some might argue that a hand meter is more convenient in some situations but that is something of a moot point.

You can take incident readings with camera's meter very easily, and a hand meter is unlikely to have a spot option, and it certainly won't have anything as sophisticated as evaluative/matrix metering. Neither will it allow you to explore more advanced technqiues like Expose To The Right (of the histogram) because it cannot create a histogram in the first place.

There are other drawbacks, which can be significant. A hand meter attempts to measure the light entering the lens, which is not the same as the light recorded by the sensor. There are losses and inaccuracies all the way down the line. Vignetting is one, absorbsion by the glass is another, variances in actual f/number, variances in actual shutter speed, and variances in actual ISO.

For truly accurate exposure, check the LCD and histogram. That is a representation of the actual exposure, rather than a calculated guess at what it might be - which is all any meter can do.

You could say that professional photographers like to be seen using a hand meter because it impresses clients. But I wouldn't dare to suggest such a thing

 

[EOS_JD]

There is nothing that you can do with a hand meter that you cannot do with the one in the camera - except manual flash metering. They are very handy for that; it's the only time I ever use one.

I disagree with that Hoppy. THere's not many situations a reflected reading will be the same as an ambient reading. I understand what you're saying but to use the camera reading you need to understand about exposure and exposure compensation.

With a meter, you take a reading for the light your subject is in (you obvoiusly need to understand what you are metering for), enter the numbers in the camera and that's it - no need to worry about +/- EC.

Some might argue that a hand meter is more convenient in some situations but that is something of a moot point.

More convenient and more consistent.

You can take incident readings with camera's meter very easily

How do you do that? The cameras meter measures reflected light not incident. It will be affected by the brightness of the subject (white/black) and will not provide a consistent result.

and a hand meter is unlikely to have a spot option, and it certainly won't have anything as sophisticated as evaluative/matrix metering.

Yes the spot option is on the camera but the only time that is useful to me is where I can't access the area I want to shoot. If shooting a couple why not just walk up and get an accurate consistent meter reading? Even a reflected spot reading needs to be adjusted by the user.

If you need Evaluative/Matrix metering, just use the cameras meter. Many meters will give you an accurate average reading from a scene.

Neither will it allow you to explore more advanced technqiues like Expose To The Right (of the histogram) because it cannot create a histogram in the first place.

The meter doesn't need the histogram. It provides an accurate reading. If you want to overexpose, just add some exposure. However expose to the right is not about overexposing.

There are other drawbacks, which can be significant. A hand meter attempts to measure the light entering the lens, which is not the same as the light recorded by the sensor.

A hand meter measures the light falling on the subject. A cameras meter records the light reflected from it - very different and only one is accurate - so that it needs no adjustment.

There are losses and inaccuracies all the way down the line. Vignetting is one, absorbsion by the glass is another, variances in actual f/number, variances in actual shutter speed, and variances in actual ISO.

You should calibrate your mater when you get it to take account of any differences you may find. That is standard practice - in reality the difference is not much - certainly nowhere near the innaccuracies in using the camera's reflected meter.

For truly accurate exposure, check the LCD and histogram. That is a representation of the actual exposure, rather than a calculated guess at what it might be - which is all any meter can do.

I don't relly get that. It's very difficult to read a histogram accurately as the histogram will look different for every image and each different scene. A reading of the light falling on your subject is the only way to be accurate unless you want to spend time playing around with a histogram for every shot.... I just don't see the benefit in that lost time and effort when one press of a button does it for you.

You could say that professional photographers like to be seen using a hand meter because it impresses clients. But I wouldn't dare to suggest such a thing

Hoppy, having recently converted to M and a meter, I see the benefit and the person that thinks otherwise is a little foolish.


I suppose it's down to how and what you shoot but on a wedding, use of the histogram will not give you consistent exposures shot to shot. An incident meter will.

I respect your views mate but I think you have this seriously wrong.

 

[Richard King]

I brought one for the studio flash, but actually, it is very good at sorting out high contrast situations (like brides and grooms on a sunny day)

while you can use a camera as a spot meter, it isn't really that clear where the spot is or how wide the spot is. you are taking it on faith that it is inthe middle of the frame. with a hand held meter, then yes, you can measure say the exposure needed for the brides face, spot on first time every time

Secondly, it now gets wielded for anything ala-strobist - e.g. outdoor portrait with a couple of SB800's. while commander mode and wireless are a real hoot, actually knowing the exposure needed for ambient, and then that for each of the flashes, helps a lot

___


There is a second school of thought that says don't use one - shoot it and tweak it out - in total reality, I think most of us do a combination of the two. We use the meter to get the initial set up about right, we then make a little tweak after viewing the image. Of course if you are outside in the blazing sun, judging what's on the screen is pretty damn hard

 

[awp]

There are times where your camera meter can easily be fooled though and that's where the incident meter can come in very handy

...which is exaclty why I have one - also for studio and other off camera flash shots when time permits. Otherwise it's guess - check the camera - re-shoot.

 

[MartinH32]

Taking an ambient meter reading and not a reflected reading is always going to give you more consistent results. I think Alison mentioned it further up the thread, it's the only way to get good results in ackward lighting situations - such as backlit portraits (sunny days for example). Photography where you have stark contrasts - bride in white, groom in black against a very bright or very dark background. I trust it far more than I do the cameras meter.

 

[awp]

I think we're starting to confuse things here - the choice is between incident and reflected. Incident readings in theory are going to be more accurate but camera metering systems are very very good these days. Also, unlike in film days, you can check the result on the camera as you go.

 

[cowasaki]

I have only ever used mine in the way I described above ie for off camera flash. I might give it a go in some other situations. I had really major problems at my brother's wedding about a month ago as it was VERY bright sunlight. It might have been interesting to see what difference the meter came up with.

 

[HoppyUK]

I disagree with that Hoppy. THere's not many situations a reflected reading will be the same as an ambient reading.

<snip>

Hoppy, having recently converted to M and a meter, I see the benefit and the person that thinks otherwise is a little foolish.

I suppose it's down to how and what you shoot but on a wedding, use of the histogram will not give you consistent exposures shot to shot. An incident meter will.

I respect your views mate but I think you have this seriously wrong.

I'm not sure you are properly getting how exposure meters work, and as a result not understanding what I said.

I didn't say there was no benefit to using a hand meter. Sometimes it is more convenient, and very useful for flash. I said that.

I also said that (apart from flash) there is nothing you can do with a hand meter that you cannot to with the built-in meter, and that there are quite a few very important things that you cannot do with one.

I think that is the answer the OP needs to hear. Listen to Andrew/awp if you don't want to believe me. He also understands how expsore meters actually work. You are confusing terminology, ambient (as opposed to flash) and incident (as opposed to reflected). Just a slip I'm sure but it's important to be clear.

All any meter does is measure light, it then converts it into camera settings. What matters is what you point it at, how you convert the reading, and how you interpret that in relation to your particular camera/lens/sensor and requirements.

I am aware on the drawbacks of taking a reflected reading from the subject, and how an incident reading eliminates that variable. It does this by measuring the light source, and applying a conversion factor (ie the invercone/white diffuser) which reduces the light to the equivalent of 18% grey.

In that sense it is no different to taking a reflected light reading off an 18% grey card, or by taking a similar reading off the palm of your hand and adding 1.3 stops, or off a white card and adding three stops, etc etc - any tone will do so long as it is known and the appropriate compensation factor is applied. That is exactly the same as taking an incident reading with a hand meter, or if you prefer, you can use one of the many diffusers sold to fit over the lens - that converts the camera to an incident meter directly, without any compensation necessary. Or you can just hold three layers of Tesco Value copier paper over the lens - same difference.

While an incident reading should always be consistent, and will peg 18% grey to the middle of your tone range, that might not always be where you want it for optimum exposure. It's always going to be pretty damn close of course, and is perfect for shooting slide film for which it was originally invented (because with slides there is no opportunity to modify anything in printing/post processing). But it is usually not the best exposure for shooting negative film, and neither is it usually absolutely optimum for digital - hence the Expose To The Right technnique which will quite likely yield exposure settings two stop higher than an incident meter will indicate. Two stops is a heck of a lot to throw away if you want the absolutely best result the sensor can record and deliver.

You can attempt to apply a correction factor to an incident reading to better exploit the potential of ETTR, but that is a very dangerous game as you're pushing things to the limit and every subject is different. The only way to know that you have not exceeded the limit is to read and understand the histogram/blinkies. Which brings us back to the LCD, which is an absolute indicator of actual exposure, with no guesswork involved, rather than the 'best estimate' which is all the meter can ever do, incident or reflected. That is the ultimate arbiter of 'optimum' exposure, however you want to define that. It is not always easy to read the LCD/histogram/blinkies, but if you understand what's happening, all the necessary information is there.

Other variables. Vignetting - shoot wide-angle zoom at low f/numbers and the corners of the image will be two stops down on the centre, so what is the correct exposure for that? In theory, f/16 is one stop down from f/11 - except that if the diaphragm blades are fractionally out, like 1/10th mm through mechanical inaccuracy, it could be half a stop different. Does f/5.6 with an L prime look the same as f/5.6 on a consumer zoom? No, f/numbers are theoretical, not actual - add a third of a stop. What is ISO400? The subject of some debate between manufacturers, but basically there is no agreed ISO standard and some of them stretch the truth to the point of porkies - half a stop potential varianace there.

What this boils down to, is that the LCD image and its associated aids takes everything into account. It is an actual recording. If you like what you see, then lock the settings in manual. You just don't need a hand meter to get you there.

It's important not to confuse consistency with accuracy. Or to think that an incident reading is any different to a reflected reading from a known tone. Or to confuse convenience with capability.

 

[Denyerec]

Hoppy, I know you're a stickler for accuracy, which is why I'd like to direct you here in case it was previously unknown to you:
http://www.bythom.com/graycards.htm

Keep up the detailed posts, they're very worthwhile!

 

[HoppyUK]

Hoppy, I know you're a stickler for accuracy, which is why I'd like to direct you here in case it was previously unknown to you:
http://www.bythom.com/graycards.htm

Keep up the detailed posts, they're very worthwhile!

Cheers Denyer

I am aware of that, and have referred to it before. But I don't bring it up much because whether exposure meters are calibrated to 12% or 18% grey only adds confusion and, ultimately, it's irrelevant if you work to the LCD/histogram (which is the key to optimum exposure, however you care to define that). There is no such thing as 'correct' exposure - there is acceptable/good exposure, which is what you'll always get using an incident light reading, but there is better/optimum exposure if you work on it a bit, according to the subject and the result you want (we're talking quite fine differences here!).

It does however suggest that with a digital sensor that has more dynamic range than seven stops (the standard used as a basis for film exposures) then working from 12% exploits a bit more headroom beyond seven stops than 18% does, kind of a bit of built-in 'expose to the right'.

But it's all far from clear, and that link doesn't explain much. References to Ansel Adams, who shot on large format negative film, are completely irrelevant to digital. What calibration figures are manufacturers actually using? I believe they are coy about it because if by default they are adding additional exposure in their suggested settings, then the impression that users might get is that the true ISO is lower than it is. Which might not be a correct impression but it's still bad marketing.

At the end of the day, getting the exposure right is about creating a good looking image, the best you can make it. A meter is only a tool, only a rough guide really, to help you get that. The histogram tells you exactly what has been actually recorded, and the more data you can cram on to it, the more photons and the less noise (higher signal to noise ratio) you can get, the better. When you look at the graph, all the black areas above the graph are wasted areas of data collection, and there is massively more recording potential on the right hand side than there is on the left - it is not a linear scale. So pushing the histogram to the right with more exposure will record more information. Of course the image will then probably look too bright, but when you darken it down again in post processing, you get better tone sepatation, more shadow detail, and less noise. (Just to be clear, you need to shoot Raw with ETTR technique, and post process the files to darken them.)

Given the foregoing it might be hard to believe, but I really don't want to go on about it. I just wanted to correct the idea that somehow a hand meter was some kind of magic weapon, and that incident readings are always abosolutely optimum. To repeat, there is nothing that a hand meter can do that you can't do with the built-in meter (manual flash metering excepted) and there are lots of things it cannot do. And an incident reading will always be good (especially if you only shoot JPEGs) and always consistent, but rarely optimum.

 

[EOS_JD]

I'm not sure you are properly getting how exposure meters work, and as a result not understanding what I said.

I think I do Although always open to new ways of thinking but I think your arguments are flawed.

I didn't say there was no benefit to using a hand meter. Sometimes it is more convenient, and very useful for flash. I said that.

Almost essential for flash particularly with multi flash setups - And certainly a requirement to get bang on consistent results even with ambient light.

I also said that (apart from flash) there is nothing you can do with a hand meter that you cannot to with the built-in meter, and that there are quite a few very important things that you cannot do with one.

Not sure I agree but I'm reading one and answering as I go.

All any meter does is measure light, it then converts it into camera settings. What matters is what you point it at, how you convert the reading, and how you interpret that in relation to your particular camera/lens/sensor and requirements.

I agree with that.

I am aware on the drawbacks of taking a reflected reading from the subject, and how an incident reading eliminates that variable. It does this by measuring the light source, and applying a conversion factor (ie the invercone/white diffuser) which reduces the light to the equivalent of 18% grey.

As you know though importantly it measures the light falling on the subject. Not reflected from it.

In that sense it is no different to taking a reflected light reading off an 18% grey card, or by taking a similar reading off the palm of your hand and adding 1.3 stops, or off a white card and adding three stops, etc etc - any tone will do so long as it is known and the appropriate compensation factor is applied.

THe difference here is when you take your reading from the card you then need to see where the exposure is and adjust it. THis can take time - at a wedding I'd rather press a button and be told than have to flaff around. Also taking readings helps you understand the light.

That is exactly the same as taking an incident reading with a hand meter, or if you prefer, you can use one of the many diffusers sold to fit over the lens - that converts the camera to an incident meter directly, without any compensation necessary. Or you can just hold three layers of Tesco Value copier paper over the lens - same difference.

What covers are you talking about? I've not seen these? Only cover I have is the expodisk and I use that for WB correction. I can't see how a reflected meter can be changed to an incident one though?

While an incident reading should always be consistent, and will peg 18% grey to the middle of your tone range, that might not always be where you want it for optimum exposure. It's always going to be pretty damn close of course, and is perfect for shooting slide film for which it was originally invented (because with slides there is no opportunity to modify anything in printing/post processing). But it is usually not the best exposure for shooting negative film, and neither is it usually absolutely optimum for digital - hence the Expose To The Right technnique which will quite likely yield exposure settings two stop higher than an incident meter will indicate. Two stops is a heck of a lot to throw away if you want the absolutely best result the sensor can record and deliver.

I mentioned that above I think - Expose to the Right is easily done in the meter. If you know you have 1-2 stops extra, just calibrate the meter to +1 or whatever exposure and every exposure you take will be that consistent amount over the correct exposure. Only problem there is there are times when you don't want to take a chance with this - i,e when shooting weddings!

]You can attempt to apply a correction factor to an incident reading to better exploit the potential of ETTR, but that is a very dangerous game as you're pushing things to the limit and every subject is different. The only way to know that you have not exceeded the limit is to read and understand the histogram/blinkies.

If you are correct in saying that you have the latitude of 1-2 stops (which I'm not arguing with), use the incident meter to expose for the highlights in a high contrast scene and you should be fine. As you say it's understanding what you point at that allows you to understand the process. I do not advocate changing the meter in this way and it should be used more where you have the time to tweak exposures - not on a busy shoot where you want consistent exposures.

Which brings us back to the LCD, which is an absolute indicator of actual exposure, with no guesswork involved, rather than the 'best estimate' which is all the meter can ever do, incident or reflected. That is the ultimate arbiter of 'optimum' exposure, however you want to define that. It is not always easy to read the LCD/histogram/blinkies, but if you understand what's happening, all the necessary information is there.

If you meter for the highlights, you won't get blinkies. To use a reflected meter, you therefore need to expose for the highlights and then guess the amount of EC, check and reshoot to check.... and so on (depending on your knowledge) - Use a Light Meter, expose for the highlights and go (flash may be required for fill in a high contrast situation but this would depend on what you shoot).

Other variables. Vignetting - shoot wide-angle zoom at low f/numbers and the corners of the image will be two stops down on the centre, so what is the correct exposure for that? Correct exposure will be what the incident meter tells you. Vignetting is an unfortunate downfall of some lenses although it is much reduced with crop sensors.

In theory, f/16 is one stop down from f/11 - except that if the diaphragm blades are fractionally out, like 1/10th mm through mechanical inaccuracy, it could be half a stop different.

Does f/5.6 with an L prime look the same as f/5.6 on a consumer zoom? No, f/numbers are theoretical, not actual - add a third of a stop. What is ISO400? The subject of some debate between manufacturers, but basically there is no agreed ISO standard and some of them stretch the truth to the point of porkies - half a stop potential varianace there.

How would most of us know this? And how often do you think anyone considers it?

Whilst I'm not arguing these points, I don't see that it's "that" relevant in most situations. Half a stop is not a lot and with consistent exposures be the +/- a half stop, a quick batch in Lightroom will sort it and you can adjust the next time with that setup. If you do notice consistent under/over exposure with a particular lens or camera it's easy to bring this into the equation by adding/reducing exposure by calibrating your meter.

What this boils down to, is that the LCD image and its associated aids takes everything into account. It is an actual recording. If you like what you see, then lock the settings in manual. You just don't need a hand meter to get you there.

But it seems getting to that exposure is the issue. Fine if you have the time to play around but in the speed of an event it doesn't make sense and is just much quicker and erasier to press a button and go. Know your equipment and all will fall into place.

It's important not to confuse consistency with accuracy. Or to think that an incident reading is any different to a reflected reading from a known tone. Or to confuse convenience with capability.

This is a point that I could also partly agree with. But your view of accuracy may not be mine. You are in some cases looking to over expose certain images and bring that exposure back in post processing to reduce noise (the basis of ETTR) - Not saying that's wrong but your idea of an accurate exposure may be different to others. I do agree with your points about that understanding what you are shooting and knowing the scene and reading the histogram etc etc but by doing this you will waste a lot of time.

For me therefore the important thing is to trade off the points you note. Accuracy/convenience, Time/Money etc etc....Software is so advanced now that a slight under/over exposure is rarely an issue.

 

[EOS_JD]

I have only ever used mine in the way I described above ie for off camera flash. I might give it a go in some other situations. I had really major problems at my brother's wedding about a month ago as it was VERY bright sunlight. It might have been interesting to see what difference the meter came up with.

When you shoot in any semi auto mode (as you may have done) the camera reads every scene differently and produces different eexposures every time. Putting the camera in manual and using an incident meter will get you quickly to more accurate results so long as you know what you are metering for. As Hoppy says there are other ways to get this - but imo none as quick and convenient.

 

[Lensflare]

Isn't a hand meter just a bit bigger than a hand yard?

And about 3x bigger then a hand foot?

 

[Jayst84]

I prefer taking photographs.

In the majority of situations I find my eye and a quick check of the first shot on the LCD / with the histogram is plenty good enough.

 

[HoppyUK]

I think I do Although always open to new ways of thinking but I think your arguments are flawed.


Almost essential for flash particularly with multi flash setups - And certainly a requirement to get bang on consistent results even with ambient light.

Not sure I agree but I'm reading one and answering as I go.

I agree with that.

As you know though importantly it measures the light falling on the subject. Not reflected from it.



THe difference here is when you take your reading from the card you then need to see where the exposure is and adjust it. THis can take time - at a wedding I'd rather press a button and be told than have to flaff around. Also taking readings helps you understand the light.



What covers are you talking about? I've not seen these? Only cover I have is the expodisk and I use that for WB correction. I can't see how a reflected meter can be changed to an incident one though?



I mentioned that above I think - Expose to the Right is easily done in the meter. If you know you have 1-2 stops extra, just calibrate the meter to +1 or whatever exposure and every exposure you take will be that consistent amount over the correct exposure. Only problem there is there are times when you don't want to take a chance with this - i,e when shooting weddings!


If you are correct in saying that you have the latitude of 1-2 stops (which I'm not arguing with), use the incident meter to expose for the highlights in a high contrast scene and you should be fine. As you say it's understanding what you point at that allows you to understand the process. I do not advocate changing the meter in this way and it should be used more where you have the time to tweak exposures - not on a busy shoot where you want consistent exposures.

If you meter for the highlights, you won't get blinkies. To use a reflected meter, you therefore need to expose for the highlights and then guess the amount of EC, check and reshoot to check.... and so on (depending on your knowledge) - Use a Light Meter, expose for the highlights and go (flash may be required for fill in a high contrast situation but this would depend on what you shoot).



How would most of us know this? And how often do you think anyone considers it?

Whilst I'm not arguing these points, I don't see that it's "that" relevant in most situations. Half a stop is not a lot and with consistent exposures be the +/- a half stop, a quick batch in Lightroom will sort it and you can adjust the next time with that setup. If you do notice consistent under/over exposure with a particular lens or camera it's easy to bring this into the equation by adding/reducing exposure by calibrating your meter.



But it seems getting to that exposure is the issue. Fine if you have the time to play around but in the speed of an event it doesn't make sense and is just much quicker and erasier to press a button and go. Know your equipment and all will fall into place.



This is a point that I could also partly agree with. But your view of accuracy may not be mine. You are in some cases looking to over expose certain images and bring that exposure back in post processing to reduce noise (the basis of ETTR) - Not saying that's wrong but your idea of an accurate exposure may be different to others. I do agree with your points about that understanding what you are shooting and knowing the scene and reading the histogram etc etc but by doing this you will waste a lot of time.

For me therefore the important thing is to trade off the points you note. Accuracy/convenience, Time/Money etc etc....Software is so advanced now that a slight under/over exposure is rarely an issue.

I'm not going to go through this point by point. You are not understanding what I've said, and are dismissing other factors because they don't apply to your (very narrow) area of working, ie weddings, or to the equipment you use.

You already have an ExpoDisc - check it out. It converts your camera into an incident meter, if that's what you want. On their website it says "The ExpoDisc's 18% light transmission allows the ExpoDisc to be used as an incident ambient exposure tool (as you would with an 18% gray card), as well as for white balance." You can buy the same thing on ebay for a fiver.

Now you can throw away your hand meter and do a custom white balance at the same time! http://www.expoimaging.com/product-detail.php?cat_id=1&product_id=2&keywords=ExpoDisc_Neutral

 

[classcams]

i recently bought one, and I found that the lightmeter told me exactly what the camera did when I was metering off my hand and exposing by +1, but it does help me in tricky lighting situations when I have to balance a few different light sources.

Only if you are an Anglo

 

[classcams]

Cheers Denyer

I am aware of that, and have referred to it before. But I don't bring it up much because whether exposure meters are calibrated to 12% or 18% grey only adds confusion and, ultimately, it's irrelevant if you work to the LCD/histogram (which is the key to optimum exposure, however you care to define that). There is no such thing as 'correct' exposure - there is acceptable/good exposure, which is what you'll always get using an incident light reading, but there is better/optimum exposure if you work on it a bit, according to the subject and the result you want (we're talking quite fine differences here!).

It does however suggest that with a digital sensor that has more dynamic range than seven stops (the standard used as a basis for film exposures) then working from 12% exploits a bit more headroom beyond seven stops than 18% does, kind of a bit of built-in 'expose to the right'.

But it's all far from clear, and that link doesn't explain much. References to Ansel Adams, who shot on large format negative film, are completely irrelevant to digital. What calibration figures are manufacturers actually using? I believe they are coy about it because if by default they are adding additional exposure in their suggested settings, then the impression that users might get is that the true ISO is lower than it is. Which might not be a correct impression but it's still bad marketing.

At the end of the day, getting the exposure right is about creating a good looking image, the best you can make it. A meter is only a tool, only a rough guide really, to help you get that. The histogram tells you exactly what has been actually recorded, and the more data you can cram on to it, the more photons and the less noise (higher signal to noise ratio) you can get, the better. When you look at the graph, all the black areas above the graph are wasted areas of data collection, and there is massively more recording potential on the right hand side than there is on the left - it is not a linear scale. So pushing the histogram to the right with more exposure will record more information. Of course the image will then probably look too bright, but when you darken it down again in post processing, you get better tone sepatation, more shadow detail, and less noise. (Just to be clear, you need to shoot Raw with ETTR technique, and post process the files to darken them.)

Given the foregoing it might be hard to believe, but I really don't want to go on about it. I just wanted to correct the idea that somehow a hand meter was some kind of magic weapon, and that incident readings are always abosolutely optimum. To repeat, there is nothing that a hand meter can do that you can't do with the built-in meter (manual flash metering excepted) and there are lots of things it cannot do. And an incident reading will always be good (especially if you only shoot JPEGs) and always consistent, but rarely optimum.

Another chapter from "Gone with the wind"

 

[AliB]

Sorry Richard but that is a totally inaccurate. An incident meter records the light falliing on the subject and should be metered from the subject towards the light source. b****r all to to with the camera.

 

[HoppyUK]

Sorry Richard but that is a totally inaccurate. An incident meter records the light falliing on the subject and should be metered from the subject towards the light source. b****r all to to with the camera.

Ali, please pay attention The light falling on the subject is not the same as the light recorded by the sensor.

It has to go through the lens, which a hand meter knows nothing about, while the on-board TTL meter does. The TTL meter therefore compensates for any variables there, light lost through absorbsion (f/numbers are theoretical, not the same as actual transmission), internal reflection and vignetting.

Furthermore, if you read from the LCD/histogram (the actual exposure, no guesswork metering going on here) that takes into account any variables introduced by mechanical innaccuracies in the diaphragm on stopping down (especially at high f/numbers) and it also compensates for a sensor that is over-optimistic on it's ISO rating (common). Shutters also used to be inaccurate at high speeds but they're pretty good now (although leaf shutters are less accurate - a Hasselblad at 1/500sec is nearer 1/350sec).

With high end equipment these variables will probably not be significant, but try it with a wide consumer zoom at low f/number (vignetting), try a couple of different lenses at f/22 (innacurate diaphragm), see what ISO400 looks like with a few different cameras. Then you'll see what I mean.

Why are you guys so dead set on using an incident reading hand meter for everything? It's great for your situation, weddings, and photographers have been using it for that for ever. I would too! For all sorts of reasons it's a very good and practical solution for that - pretty accurate, consistent, 'safe' for that dress!

But it's not infallible, in other situations it could be be quite a long way out (eg long lens), sometimes impossible to use, and rarely absolutely optimum. The on-board system addresses all those shortcomings, AND it can do anything that a hand meter can do.

 

[HoppyUK]

Classcams is back!

Only if you are an Anglo

Actually, the palm of your hand is quite a consistent tone between races. A 'black' hand has a slight advantage in that it usually requires zero compensation.

Another chapter from "Gone with the wind"

That's another classic

 

[unclened]

as self proclaimed judge of this debate, call this one a tie!

I agree with both sides of the argument. Put it this way... a beginner with little knowledge of metering could use a lightmeter and get an exposure pretty much spot on. However.... a more experienced user who understands metering can acheive the same result too, without the aid of a light meter.

Its just down to personal choice and what route people want to take to get the decent exposure.

 

[unclened]

Actually, the palm of your hand is quite a consistent tone between races. A 'black' hand has a slight advantage in that it usually requires zero compensation.

I've heard racists comments before, but calling someone 18% grey is just bang out of order Richard!

 

[AliB]

lol I am paying attention Richard, particularly to the bit where you say an incident meter is trying to measure the light entering the lens which is utter poppycock (always wanted to use that word!) It's measuring the light falling on the subject as metered with the meter pointing towards the light source.

If I don't have it on me then I sometimes use a technique where, still using evaluative metering, I zoom right in on the part of the image I want correctly exposed and meter off that. But for things like window lit portraits, give me a blooming incident meter any day!

 

[Denyerec]

Ali I think you mis-read Hoppy. He did not suggest that an incident meter was trying to measure light entering the lens.
The implication is that an incident meter is metering light falling, well, incidentally on the meter itself. It's providing an exposure value suitable for the meters own internal calibration, but that doesn't necessarily tie into the camera's calibration.

The light falling on the camera sensor/film can be quite different, for the reasons elaborated on by Hoppy.

I think you're both on the same page, but the book is in different languages!


Hoppy: Can you suggest what I should see on my histogram if I shoot a piece of CEILAB-calibrated pure-white material with my camera's spot meter suggested exposure? The spotmeter should, as I understand it, render what it's targetted at as an "18%" luminance level. Ergo should one expect to see the white rendered in the image as "18% Grey" in such a scenario? I've been conducting some tests but am not sure I'm fully equipped to interpret the results.

 

[HoppyUK]

lol I am paying attention Richard, particularly to the bit where you say an incident meter is trying to measure the light entering the lens which is utter poppycock (always wanted to use that word!) It's measuring the light falling on the subject as metered with the meter pointing towards the light source.

If I don't have it on me then I sometimes use a technique where, still using evaluative metering, I zoom right in on the part of the image I want correctly exposed and meter off that. But for things like window lit portraits, give me a blooming incident meter any day!

Then you're not paying enough attention Ali And putting your own (wrong) interpretation on what I'm saying. You're seem to be obsessed with how with an incident reading you point the meter at the camera from the subject (or the light source - opinions vary) as if this makes some kind of fundamental difference. It's just a different way of doing the same thing as a reflected reading off a grey card. And that may not be optimum either

But that's not the point I was making there. The point is, a hand meter measures the light - by whatever means - that will enter the lens. It does not, and cannot, take into account the unknown variables that happen after it enters the lens. Whereas the on-boad system can, and does. And they exposure settings you get are often different.

Why don't you guys try some of the things I've suggested instead of arguing the toss? I've spent half my life doing it as a magazine technical editor. I'm not making this stuff up, or making academic points for the sake of it.

 

[The23rdman]

I prefer taking photographs.

In the majority of situations I find my eye and a quick check of the first shot on the LCD / with the histogram is plenty good enough.

Hail to that.

 

[The23rdman]

And all I'll add is that the only times I've used a light meter (in the studio) because someone told me I absolutely had too everything came out too hot for me. I'd rather check the histogram and chimp.

 

[Lensflare]

The difference between exposure choices is what provides different results from different people. Quite the "I am right and you are wrong".

The flexibility surrounding the mid point exposure means that you can be a stop either side and still get a perfectly reasonable result.....if you only want the highlights "correct" or are happy for the highlights to blow and the shadows are more important....depends on the subject, but more importantly, the mood of the photographer and how they want to portray what is in front of them.

An incident reading is taken by measuring the light falling on the subject.
A reflected reading is taken be measuring the light reflected off the subject.

Simple as that, whether you use a hand meter or a built in one. It is a bit difficult to take a true incident reading with a built in meter though....metering off a grey card is a reflected reading. Full stop. It is measuring the reflected light off the card, so is NOT an incident reading, it is a reflected reading - just from a known mid tone, thats all. You could use concrete wall, hand, face, grey jumper, grass - all much the same and certainly close enough!

 

[HoppyUK]

And all I'll add is that the only times I've used a light meter (in the studio) because someone told me I absolutely had too everything came out too hot for me. I'd rather check the histogram and chimp.

Thank God for an outbreak of common sense.

The difference between exposure choices is what provides different results from different people. Quite the "I am right and you are wrong".

The flexibility surrounding the mid point exposure means that you can be a stop either side and still get a perfectly reasonable result.....if you only want the highlights "correct" or are happy for the highlights to blow and the shadows are more important....depends on the subject, but more importantly, the mood of the photographer and how they want to portray what is in front of them.

An incident reading is taken by measuring the light falling on the subject.
A reflected reading is taken be measuring the light reflected off the subject.

Simple as that, whether you use a hand meter or a built in one. It is a bit difficult to take a true incident reading with a built in meter though....metering off a grey card is a reflected reading. Full stop. It is measuring the reflected light off the card, so is NOT an incident reading, it is a reflected reading - just from a known mid tone, thats all. You could use concrete wall, hand, face, grey jumper, grass - all much the same and certainly close enough!

Incident refers to way the meter is used, not to the reading it takes. Used in incident style, the invercone or diffuser absorbs sufficient light to bring the tone down to 18% grey, which is obviously the same as a reflected reading directly from a grey card, where the grey tone does the same amount of light absorbsion. Different method of measuring the light, same result.

And in both cases, the same basic assumption is being made - that pinning 18% grey to the middle of the tone range, will give a good result. Which it will, but 18% grey might not be the middle of the sensor's dynamic range (in the same way as it is with slide film). These days, digital sensors often have a couple of stops more dynamic range available. So good exposure, but not optimum.

The only way to get optimum exposure is the interrogate the histogram.

 

[awp]

With meters like the Sekonic L-758DR you can set the meter up to match the camera sensors dynamic range. You can store up to 3 sets of camera data.

 

[Kev M]

The only way to get optimum exposure is the interrogate the histogram.

Even if it's only a JPEG interpretation of the data and won't actually look much like the RAW file and is also subject to having the correct white balance set (something incident meters don't care about).