Camera or Light meter which is right.

  • Thread starter Thread starter whiteflyer
  • Start date Start date
W

whiteflyer

Guest
I recently bought myself a Sekonic L358 Exposure Meter , for shooting in difficult lighting conditions and have noticed it sometimes give completely different settings than what the camera says.

Now I know the camera meters reflected light and the Sekonic meters incidental light, but it can be a couple of stops different.

Anyway, while out with JustDave today I did a quick two shot comparison, there was not a huge differences in the suggested shutter speed, but which is right and can you say which was the camera metered shot.

The shots have had NOTHING done to them, just convert to Jpeg and save for web.

1D4_7484.jpg


1D4_7485.jpg
 
I'd say the second shot is a little under exposed so would guess that is the camera metering due to all the snow.
 
LOL thanks Mark,:thumbsdown: knew it was a bad idea, :thumbsdown: hope I didnt break your camera :lol: for me looking into the sun was a prob, I cant remember which is which but #2 looks best if it could ever :lol: and truer colour, think that was the metered one

Least my cig wasnt in shot
 
You can get variations from the camera's meter just by changing from evalutive, to spot, to centre weighted, so hardly suprising that an incident reading would be different again. No meter is the holy grail of exposure, you need to take into account what part of the shot is your emphasis and interpret the readings appropriatly. For Bird in flight shots I frequently take a reading from average scene to establish a base exposure and set the camera manually to that, if I wave it round the sky it goes all over the place and then you get the black and white bird problem :) Use the readings as an intelligent guide.
 
Just Dave said:
One thing I notice Mark you took the reading fron under my chin, maybe sort of in front of my face would be prefered
Click to expand...

Id guess If the incidental where used like this then effectively your asking for an (18% grey) exposure of his neck to be perfect ...which is what you have in your first shot.

Same as foggy for the second then and I agree with artyman for all the reasons.
 
OK the second shot is from the light meter.

No mater how I use it outside it nearly always says a faster shutter speed than the camera does. The camera Metering Mode was Evaluative Metering. These two shot are only 2/3 of a stop different.


I do know I have more to learn with the meter. :)
 
Last edited by a moderator:
Hum, as said above depends on your camera metering.

Did you use Evaluative, partial, spot or center weighted?

I think by default my camera is on evaluative, so the snow would have caused the camera to underexpose, so I normally over expose by 2/3rds in snow or switch to partial or spot metering. But your pic looks over exposed.

By using a light meter you've got a much more accurate reading from the exact spot you want correctly exposed, hence the better picture.
 
Curious results there Mark.The second shot is definitely the better of the two, but your results are the polar opposite of what I've been finding doing some testing with the G9 and my Sekonic L-558.

The G9 on evaluative metering under-exposes every time, with more visible detail in the snow at the expense of the snow (and any other whites) looking a bit grey.

The L-558 incident reading produces nice white crisp looking snow when the settings are transferred to the G9 in manual mode. There's less detail in the snow but it still produces the nicer looking brighter image.

Just make sure you have the same ISO set in the camera and the Sekonic (obvious I know) and just check that any calibration facility in the meter is at zero. If it's at zero and not producing the result you want there's no reason at all why you shouldn't add exposure comp to the meter in small increments until you have it calibrated to your own liking and thereafter it should be bang on the money.
 
Hmm, ...2/3 of a stop is a very similar amount to the difference between an 18% (film days ...Sekonic too? ) grey metering standard ..and a 14-15% gey standard that seems to be mostly used in digi metered cameras (although that may be changing as stuff improves , dunno) ... Im not sure how much that applies to your problem though ... especially as your images seem to be the opposite of that idea.

:thinking:
 
Thanks Cedric and Adam, a few things to think about.

I did read in the Sekonic manual that I can calibrate the meter by a stop, or add compensation. I'm sure I can do the same on the 1D and set a Cnf to have exposure calibration.

I'll have to take many many more shots just to check things out, but I do think it's more the camera that over exposes than a Sekonic error.

Then again the most likely problem is USER ERROR :)
 
The second one is clearly the better exposure - there is detail in the snow where the image at the top loses it and the skin tones do not look washed out as they do in the top image.

Camera meters can be affected by all sortds of things. - Too much white/black light bouncing into the lens, backlighting affecting the metering..... Reflected meter readings give an indication of what the camera thinks is right.

An incident meter tells you what light is falling on yuor subject - No ifs or buts. you can move the meter to err towards the highlights or shadows and the small white dome will help with that. Correct placement of the meter is crucial.

To check, why don't you take a shot of an 18% grey card. Firsly meter it then shoot. it should come out grey (you may need to do a custom WB which is easy too).

This would give a good idea of the difference in meter and camera exposure - shoot the same scene with camera in your appropriate exposure mode. Both should be the same so long as you fill the scene with 18% grey.
 
I think you are looking at this in the wrong way.

You need to take some trial images of known tonal values and know where they will fall on the histogram, then adjust your ISO or calibration to get the consistent results you require.

Actually you should do this for your camera meter too if you want to be sure of your exposure when working at speed and don't have time to check the histogram between every shot.

The use of digital RAW capture and matrix metering has made the task of needing to know how your meter responds less of a requirement than it needs to be if using film.
 
stuart anthony said:
to make two comparison shots they have to be composed the same, these are not. there is less snow in shot two on the lower part of frame, and subsequently more dark area in the upper part of frame.
Click to expand...

:) There is about 3 of 4 inches in the vertical composition of the two shots, I doubt that would make any difference.
 
I did not mean to complain by starting the thread, more just to point out I was have different readings, but it's been very good to hear different suggestions.

Lots to take onboard, but as I said it's more probably to be my user error.
 
Looking at the two images again I see that the total area of the darker areas is greater than the total area of the snow, which is why the camera meter has over-exposed. I suspect if you took the shot again against a complete snow background Mark, you'd get the under-exposure we'd typically expect from the camera meter left to it's own devices, and would be a better indication of just how useful incident metering is in this situation.
 
Neither is "Right". Use either as a consistent guide, and your eyes and experience to judge which you prefer.

I cant remember the last time I didn't use some exposure compensation one way or the other. And the beauty of digital is that you can chimp and adjust.
 
whiteflyer said:
OK the second shot is from the light meter.

No mater how I use it outside it nearly always says a faster shutter speed than the camera does. The camera Metering Mode was Evaluative Metering. These two shot are only 2/3 of a stop different.


I do know I have more to learn with the meter. :)
Click to expand...

No doubt about it mate, its knackered. I'll give you a fiver for it.......:)
 
Last edited:
CT said:
Looking at the two images again I see that the total area of the darker areas is greater than the total area of the snow, which is why the camera meter has over-exposed. I suspect if you took the shot again against a complete snow background Mark, you'd get the under-exposure we'd typically expect from the camera meter left to it's own devices, and would be a better indication of just how useful incident metering is in this situation.
Click to expand...

I used the meter again today and I'm happy with the results, so feel it was worth buying, and will be very useful in the right situations.

Just took the reading dialled it in and shot away, I'm sure if I was in AV I would have had to use some compensation to get the runner right against such a leaden sky.

No highlight or shadow alert in DPP on this :) and I've done nothing to it, so I'm quite happy now.

1D4_8525.jpg
 
Last edited by a moderator:
No flashed used at all on this shot, just the 1D4 and 70-200 f2.8, like you say maybe 50 feet away.
 
Last edited by a moderator:
Just out of interest, how does using a meter now affect your viewing of images on the LCD? Are you trusting the meter reading and discounting what you see, or are you still relying on blinkies and histogram to check the exposure?
 
specialman said:
Just out of interest, how does using a meter now affect your viewing of images on the LCD? Are you trusting the meter reading and discounting what you see, or are you still relying on blinkies and histogram to check the exposure?
Click to expand...

Not really seeing any blinkies at the moment, but sure there will be in some shots, but still checking the histogram.

I did do a set of shot with JUST the meter and they came out fine.
 
Excellent reult there with your runner Mark - couldn't really be any better.
 
whiteflyer said:
I did not mean to complain by starting the thread, more just to point out I was have different readings, but it's been very good to hear different suggestions.

Lots to take onboard, but as I said it's more probably to be my user error.
Click to expand...


Not user error, almost certainly just a calibration issue as mentioned earlier. DSLRs appear to be calibrated to 12% or 15% grey rather than 18% for film, although I've never read anything official from manufacturers.

Digital is different to film, it needs to be treated differently, and when I compare an incident reading from my Sekonic, it usually under exposes compared to the camera. And when I compare it to the exposure I finally end up using, after consulting the histogram/blinkies and pushing that to the right a bit, it's often two stops or more adrift!

There's no such thing as 'correct' exposure, but if you want optimum exposure, or best exposure (depending on JPEG or Raw capture) then the histogram is the place to check that. It is an actual exposure, with all the many and unpredictable variables factored in; the exposure meter, any meter, is just guessing.
 
The problem with using a meter that isn't integrated into the camera is that light conditions can change very quickly, and since your eye adjusts automatically you may not realise.

Really I don't see any reason to do it. It's an extra thing to carry, will take longer to use and prevents you from being able to recompose into a differently lit area without re-metering.

Better of learning how the meter in the camera works and learning to adjust accordingly. If you have snow in the shot it's almost always going to be blown or nearly blown, if you don't want that then dial it down 1/3rd-2/3rds...

Only exception I can think of is when you want to meter a long night shot (ie >60 seconds) and the camera meter won't go that far.
 
loplyg said:
The problem with using a meter that isn't integrated into the camera is that light conditions can change very quickly, and since your eye adjusts automatically you may not realise.

Really I don't see any reason to do it. It's an extra thing to carry, will take longer to use and prevents you from being able to recompose into a differently lit area without re-metering.

Better of learning how the meter in the camera works and learning to adjust accordingly. If you have snow in the shot it's almost always going to be blown or nearly blown, if you don't want that then dial it down 1/3rd-2/3rds...

Only exception I can think of is when you want to meter a long night shot (ie >60 seconds) and the camera meter won't go that far.
Click to expand...

If I had to estimate a night exposure, I'd rather have a go with the camera than a hand meter! But far better than both is to take a snap and check the LCD/histogram.

The only time I use a hand meter is when setting up multiple studio flash heads. You can do it without, but it's heck of a palava. A meter will set the ratio of each light very quickly, but then I always refer to the histogram/blinkies for the final exposure setting, when any changes to f/number or ISO will change the overal exposure, but maintain the same lighting ratios.

The only other time I can think a hand meter would be useful would be wedding photography, and that's just a convenience thing too. That invloves a rare combination of a big white dress and dark suits that will often fool an on board meter's reflected light reading. You can in theory take an incident reading with the camera, but that invloves pointing it at the sky with a diffuser/invercone and that's not really practical.

But if you have the camera on a tripod or a relavively fixed position, and are shooting a varienty of different group combinations when the light is changing due to wind and clouds, then an incident reading with a hand meter (that has been pre-calibrated to give the result you want) will keep track of that very quickly and exactly and you can change the exposure manually with confidence and know it's not being inluenced by the changing nature of the subject, but only the changing light level.
 
shot 2 looks better to me but both are within the realms of post processing
a light meter will be set at pre arranged levels and then translated onto a dial or readout...this may vary from what the camera is set at
regarding incident readings....imho they are only a starting point for outside readings
fine in a studio or where you want unambiguous settings or for black people and snow
spot metering is far more consistent with what we want to photograph...highlights need to be controlled and shadow areas assessed relative to the subject and the intention of the final image..
 
With digital you need to be aware of the reduced dynamic range of digital media, so exposure for highlights is a good idea, shadows can always be lifted. Blown highlights can not be rescued.
 
mrcrow said:
shot 2 looks better to me but both are within the realms of post processing
a light meter will be set at pre arranged levels and then translated onto a dial or readout...this may vary from what the camera is set at
regarding incident readings....imho they are only a starting point for outside readings
fine in a studio or where you want unambiguous settings or for black people and snow
spot metering is far more consistent with what we want to photograph...highlights need to be controlled and shadow areas assessed relative to the subject and the intention of the final image..
Click to expand...

Not sure I agree with that. A spot meter reading is still a reflected reading and you still have to add/subtract EC unless you are metering an average grey from a small area of your subject? You can take an incident meter reading towards the highlight and one on the shadow area just the same as with any meter but with an incident meter you know you are measuring the light on the subject with no requirement to guess EC.

There are times when spot metering is required and useful but not sure your point comparing makes sense to me.The incident reading is the most consistent - not saying that it's the best as Hoppy I know likes to meter for the highlights and that works well but does require more work to get consistent results.

For an event/wedding/portrait tog, consistancy is king and an incident reading measuring the light falling on (not reflected from) your subject gives the most consistent results.

All down to personal preference. I use more than just one of the above but do like my L358 :)
 
mrcrow said:
a light meter will be set at pre arranged levels and then translated onto a dial or readout...this may vary from what the camera is set at..
Click to expand...

That's pretty much the whole point of an incident reading!
 
mrcrow said:
a light meter will be set at pre arranged levels and then translated onto a dial or readout...this may vary from what the camera is set at
Click to expand...

the camera's meter does the same and you can also use your camera to take an incident reading by adding something like the Expodisk (I think Hoppy told me that and it works well).
 
EOS_JD said:
the camera's meter does the same and you can also use your camera to take an incident reading by adding something like the Expodisk (I think Hoppy told me that and it works well).
Click to expand...

He might have done ;)

But yes, you can use something like an Expodisc, or others use a Pringles lid, or a coffee filter, or shoot an 18% grey card, or take a reading off the palm of your hand and add about 1.3 stops (depending on your personal calibration). Last time I checked, I think three sheets of Tesco Value printer paper over the lens gave a perfectly calibrated incident light reading.

They are all effectively the same thing - a method of measuring the brightness of the light source by referring to a constant and known tone, and converting that into a reading.
 
EOS_JD said:
Not sure I agree with that. A spot meter reading is still a reflected reading and you still have to add/subtract EC unless you are metering an average grey from a small area of your subject? You can take an incident meter reading towards the highlight and one on the shadow area just the same as with any meter but with an incident meter you know you are measuring the light on the subject with no requirement to guess EC.

There are times when spot metering is required and useful but not sure your point comparing makes sense to me.The incident reading is the most consistent - not saying that it's the best as Hoppy I know likes to meter for the highlights and that works well but does require more work to get consistent results.

For an event/wedding/portrait tog, consistancy is king and an incident reading measuring the light falling on (not reflected from) your subject gives the most consistent results.

All down to personal preference. I use more than just one of the above but do like my L358 :)
Click to expand...

i spot the area i want to be in 'grey' exposure...i watch for the intrusion of extra highlighted areas...and let the shadows take care of themselves
for me its a lottery anyway because i normally shoot the spot metering...then -1/3 and -2/3...do a chimp and reshoot if necessary
in the warmth of my kitchen i give myself the luxury of using that little X during the review period to rid myself of all the peepee poor grey card reflected light/ incident trash ups...and hope for at least 2-3 i can process...polish the turd as the saying goes
cheers
geof
 
Last edited:
mrcrow said:
i spot the area i want to be in 'grey' exposure...i watch for the intrusion of extra highlighted areas...and let the shadows take care of themselves
for me its a lottery anyway because i normally shoot the spot metering...then -1/3 and -2/3...do a chimp and reshoot if necessary
in the warmth of my kitchen i give myself the luxury of using that little X during the review period to rid myself of all the peepee poor grey card reflected light/ incident trash ups...and hope for at least 2-3 i can process...polish the turd as the saying goes
cheers
geof
Click to expand...

So how is spot metering more consistent?
 
EOS_JD said:
So how is spot metering more consistent?
Click to expand...

Spot metering will almost always deliver more consistent results, however they would only be consistently better if the person doing the metering knows what he is doing :lol: otherwise using the spot metering function will usually result in consistently poorer results.

To use the spot metering function correctly you need to be able to meter from a 'known tone' (that is, one that you can determine where it will fall within the tonal range). These are usually determined through experience but there are a few 'known' tones that can be used as starting points.

As we should all be aware, the meter is calibrated to a mid grey of between 12% and 18% depending on the manufacturer. This is a good starting point so we can start to look for tones (not colours) that equate to this, metering from these should give good results:

Green grass,
Blue sky at 90 degrees from the sun
Red Pillar Box
Normal Blue Denim Jeans
Asian Skin
Red Brick

Once you understand this you can then start to 'previsualise' your image and start to place 'non mid grey tones' so that they fall in the correct place in your final image.

Some suggestions to get you started, although it is best you learn from your own experiences as this will reinforce the message.

The following two are very important tones as often the dynamic age of a scene is too much for the sensor of the camera and you then have to decide which end of the tonal range you want to concentrate on:

The brightest point in the image that you wish to retain detail +2/+2.5 stops
The darkest point in the image that you wish to retain detail -2/-2.5 stops

Here are some others that may prove useful starting points:

Tarmac Road -2/3rds/-1 stop
Caucasian Skin +2/3rds/+1 stop
Black Skin -2/3rds/-1 & 2/3rds stop
Yellow Sand +1 stop
Concrete +1 stop

After trying a few and adjusting them slightly (to suit your meter) by using the histogram you will build up the experience required to know instinctively where various tones need to fall and can adjust accordingly.

This is a good read on how to use the Zone System in this digital age.
 
You must log in or register to reply here.
Back
Top