Incident light exposures compared

[Terrywoodenpic]

I am fully aware that a number of people are sick to death of my banging on about the advantages of Incident light exposure readings...
Those people should switch off now, before they suffer any more injury.

This morning I noticed the sun was shining through the French doors from one side of the room, and the widows on the opposite side were provide a similar but weaker side light to illuminate my black fire place. I closed the vertical blinds to soften the light. This led me to wonder how I would expose for this situation as the problems were fairly obvious.
I took three shots in raw, and saved them unaltered except for cropping and adding their histograms.

The first shot is with the camera set in aperture priority at Iso 200 (1/8 @F4)
As usual its automatic algorithms have reproduced the black fireplace as a mid grey and desaturated the vase and dried flower arrangement. But the Histogram looks good enough but a bit bright.

The second shot was taken with the incident reading taken towards the camera giving a manual exposure of 1/40 sec at F4 Iso 200. Saturation has greatly increased and the black is reproduced as its natural full black. The histogram favours the deep tones.

I took a third shot increasing the manual exposure to 1/20 sec at F4, as in the first shot, the colours are desaturated and the blacks have become greys but not quite as badly as in the first shot.

Using the Incident method has resolved two problems. Firstly it has coped with the extreme side light and secondly reproduce every thing to their original visual tones, and was not fooled by the large area of black in the shot. (had I pointed the meter at the stronger of the two light sources, the result would have been darker tones still, as that reading was close to 1/60 sec at F4)

]

 

[kestral]

Are those figures correct ? a 1/40 sec at F4 and 1/2 sec at f2

 

[Jayst84]

What's this supposed to show? That if you use a reflective meter incorrectly, you will get an incorrect exposure?

 

[Terrywoodenpic]

Are those figures correct ? a 1/40 sec at F4 and 1/2 sec at f2

Thanks

Have revise the OP first picture and exposure 1/8 @ F4
I pick up the wrong one.

 

[Terrywoodenpic]

What's this supposed to show? That if you use a reflective meter incorrectly, you will get an incorrect exposure?

It shows that by using an incident reading The tones will be reproduced correctly with out any adjustment first time.
In this situation a reflected reading is more difficult and needs compensation or the use of a grey card.

Digital cameras use algorithms to approximate exposure settings that usually are somewhat removed from the ideal.
That is to say tones in the original are unlikely to match tones in the reproduction.

Whilst a raw file can be adjusted by a stop at least. Lost saturation and colour shifts are difficult to correct accurately.

 

[Jayst84]

But if you use either meter correctly, you will arrive at the same exposure, with an identical file.

 

[Terrywoodenpic]

But if you use either meter correctly, you will arrive at the same exposure, with an identical file.

In the situation shown that is extremely unlikely. Even with as much professional experience as I have, I can think of no other way in such strong side light (there was no frontal light source).
Even a grey card can not be used accurately with fine oblique lighting. But why even bother when an incident meter can always give a correct result. And as you can see the histogram is no help.

 

[Alastair]

But the sort of person that knows how to incident meter (and possesses an incident meter) is very likely competent enough to know how to use their in-camera spot meter far better than any evaluative programme meter mode can process a scene. And can probably relate a histogram to the scene in front of them. Regardless of where the light is coming from, it's just another variation on the black dog/white cat problem.

I'll hazard a guess.. the black of the fireplace is about 1.33-1.5 stops under neutral grey (the sloping hood), the beige tone of the vase is within 0.33 stop either side of neutral. The carpet looks about 0.5 stops or thereabouts lighter than neutral. If spot readings of those three spots in the frame are roughly where I've guessed, then spot metering based of those should get pretty close to your incident reading.

 

[Jayst84]

In the situation shown that is extremely unlikely.

Using it correctly it is extremely likely. In fact it's practically guaranteed.

I'm not arguing against using an incident meter - people can do what they want. But any beginners reading this, shouldn't think that using an incident meter is some kind of magic bullet. If you learn how to use the tools you have, whatever they may be, you will get the results you want.

 

[Yv]

Using it correctly it is extremely likely. In fact it's practically guaranteed.

I'm not arguing against using an incident meter - people can do what they want. But any beginners reading this, shouldn't think that using an incident meter is some kind of magic bullet. If you learn how to use the tools you have, whatever they may be, you will get the results you want.

...is likely to be scared witless! I know what it all means and it's still frightening - moved to Talk Photography, we [the staff] don't think this is a 'basics' subject, we really are trying to keep that section for those questions that are only one step up from 'how do I turn the camera on?' However, if you prefer, it could also be moved to Talk Equipment I guess, just holler via the staffroom if you want that.

 

[sk66]

So, of the three images, which is better? IMO, not one is "right" but the last is the most "usable."
I can tell you I would rather pull the darks down in post as opposed to try to lighten the rest of the image.
I can see no reason for a difference between the first and third images other than possibly an AWB change or a methodology issue.

 

[Pookeyhead]

In the situation shown that is extremely unlikely. Even with as much professional experience as I have, I can think of no other way in such strong side light (there was no frontal light source).

With a reflective reading, yes there is... I'd have seen that great big black fireplace in the centre of the frame and then dialled in -1/3 exp comp... more if it was centre weighted metering.


Something else strikes me as odd here. You say shot 1 is 1/2 sec @ f2.5 but the reflective shot is 1/40th @ f4. That's a massive difference, yet the images look no more than 1 to 1.5 stops apart... you SURE you've not edited these images? Because if not, there's something very wrong with your metering technique.

@f2.5...

1/2
1/4 +1 stop
1/8 +2 stops
1/15 +3stops
1/30 +4stops
1/40 +4 1/3 stops

That's 4.3 stops difference... BEFORE you even take into account the 1.3 stops of aperture change that also took place, yet clearly those two images are probably no more than one and a half stops apart.


You are correct in assuming incident meters give more accurate results without having to compensate for anything (when used correctly), but what you posted makes little sense... so I suspect it wasn't being used correctly.

Your first incident metered shot looks under exposed by around 1/3 stop to me on a well calibrated screen, and it's histogram seems to corroborate this.

Either you've written that data down wrong, or there's something very odd with your metering... or they've been edited. Nothing else makes sense. That last shot appears to be what I'd think of as correct in that situation, yet you said you had to amend the results the incident meter gave you. Correctly used, an incident meter shouldn't require you to amend anything... it's an incident meter.

I think your technique needs some work.

 

[kestral]

With a reflective reading, yes there is... I'd have seen that great big black fireplace in the centre of the frame and then dialled in -1/3 exp comp... more if it was centre weighted metering.


Something else strikes me as odd here. You say shot 1 is 1/2 sec @ f2.5 but the reflective shot is 1/40th @ f4. That's a massive difference, yet the images look no more than 1 to 1.5 stops apart... you SURE you've not edited these images? Because if not, there's something very wrong with your metering technique.

@f2.5...

1/2
1/4 +1 stop
1/8 +2 stops
1/15 +3stops
1/30 +4stops
1/40 +4 1/3 stops

That's 4.3 stops difference... BEFORE you even take into account the 1.3 stops of aperture change that also took place, yet clearly those two images are probably no more than one and a half stops apart.


You are correct in assuming incident meters give more accurate results without having to compensate for anything (when used correctly), but what you posted makes little sense... so I suspect it wasn't being used correctly.

Your first incident metered shot looks under exposed by around 1/3 stop to me on a well calibrated screen, and it's histogram seems to corroborate this.

Either you've written that data down wrong, or there's something very odd with your metering... or they've been edited. Nothing else makes sense. That last shot appears to be what I'd think of as correct in that situation, yet you said you had to amend the results the incident meter gave you. Correctly used, an incident meter shouldn't require you to amend anything... it's an incident meter.

I think your technique needs some work.

That was the reason I asked if the readings given were correct.They cannot have been.There was far too big a difference in exposure.I noticed it before you! Na Na.

 

[Terrywoodenpic]

With a reflective reading, yes there is... I'd have seen that great big black fireplace in the centre of the frame and then dialled in -1/3 exp comp... more if it was centre weighted metering.


Something else strikes me as odd here. You say shot 1 is 1/2 sec @ f2.5 but the reflective shot is 1/40th @ f4. That's a massive difference, yet the images look no more than 1 to 1.5 stops apart... you SURE you've not edited these images? Because if not, there's something very wrong with your metering technique.

@f2.5...

1/2
1/4 +1 stop
1/8 +2 stops
1/15 +3stops
1/30 +4stops
1/40 +4 1/3 stops

That's 4.3 stops difference... BEFORE you even take into account the 1.3 stops of aperture change that also took place, yet clearly those two images are probably no more than one and a half stops apart.


You are correct in assuming incident meters give more accurate results without having to compensate for anything (when used correctly), but what you posted makes little sense... so I suspect it wasn't being used correctly.

Your first incident metered shot looks under exposed by around 1/3 stop to me on a well calibrated screen, and it's histogram seems to corroborate this.

Either you've written that data down wrong, or there's something very odd with your metering... or they've been edited. Nothing else makes sense. That last shot appears to be what I'd think of as correct in that situation, yet you said you had to amend the results the incident meter gave you. Correctly used, an incident meter shouldn't require you to amend anything... it's an incident meter.

I think your technique needs some work.

These are the Exif data from the three shots.


Revised

Firstly these were taken as described.....
The first was an in camera reflection reading taken in aperture priority, pointed at the centre of the shot... as expected it over exposed the blacks. this was a demonstration to show just this effect so no compensation was made.

The second shot was made using the the incident meter as recommended, with the dome pointing to the camera. Sekonic meters have a claimed accuracy of 1/3 stop it would seem to be with in this expectation, and produced a result very close to the original subject luminance. Though I agree that an increase of about 1/4 to 1/3 stop might improve the rendition. however this can not be derived from the histogram. and I wanted the shot as read.

The third shot was using the same incident reading but I increased by a full stop as I suspected this would not be far off the first shot, though this prove a slight under estimate.

There has been no editing except for a slight crop.

I have been using Incident, reflection and comparative photometers professionally for some 60 years and my technique in this is not in question.

 

[ernesto]

however your maths is a little odd...
1/2 sec at F2.5 is the same as 1/3sec at F2

F4 @ 1/40 is equal to F2.8 @1/20 and equal to F2 @1/10
so the difference in exposure between the two shots is between 1/3 second and 1/10 second which is a factor of 3.3333 times = close to 1.5 stops which looks about right.

I think your shutter speeds are going the wrong way there. You need a faster shutter speed as you open up the aperture...

 

[Terrywoodenpic]

So, of the three images, which is better? IMO, not one is "right" but the last is the most "usable."
I can tell you I would rather pull the darks down in post as opposed to try to lighten the rest of the image.
I can see no reason for a difference between the first and third images other than possibly an AWB change or a methodology issue.

By far the closest to the original in both colour and tone is the second one. ( on my calibrated screen and in camera and in reality)
However many might prefer it slightly lighter which could easily be done in PP. (I happen to like well full tone well saturated results.)
The best result might be obtained by a little dodging on the second one, to just bring out a suggestion of detail in the deepest blacks..

There is indeed very little between the first and third as that was the intention, so as to show that a similar over expose poor result would be obtained by compensation by even one stop from the actual Incident reading.

 

[Terrywoodenpic]

I think your shutter speeds are going the wrong way there. You need a faster shutter speed as you open up the aperture...

You are right. I messed up there.

I have now put in the correct photo and exif data and histogram. The one I uploaded had been taken two or so hours later. In different light.
It now makes sense

 

[Terrywoodenpic]

I have found the error I put up the wrong first picture. ( the correct one is a good deal lighter.)
I will correct things this afternoon. I put up a shot I had taken later in error It was not till I checked the time that it fell in place.

The 2nd and 3rd shots are correct.

It is now all corrected

 

[Alastair]

The first and third look quite similar though they seem to have very different exposures

Not really, there's a lot more fill in the shadows of the first compared to the third. It's most noticeable on the shaded side of the brickwork near the vase.

Which just goes to show that you can't do a scientific test unless you're in control of all the parameters. My guess is that the sun was coming in and out of cloud and because you're not controlling and maintaining your lighting at a constant level the whole exercise becomes a bit vague.

 

[Terrywoodenpic]

But the sort of person that knows how to incident meter (and possesses an incident meter) is very likely competent enough to know how to use their in-camera spot meter far better than any evaluative programme meter mode can process a scene. And can probably relate a histogram to the scene in front of them. Regardless of where the light is coming from, it's just another variation on the black dog/white cat problem.

I'll hazard a guess.. the black of the fireplace is about 1.33-1.5 stops under neutral grey (the sloping hood), the beige tone of the vase is within 0.33 stop either side of neutral. The carpet looks about 0.5 stops or thereabouts lighter than neutral. If spot readings of those three spots in the frame are roughly where I've guessed, then spot metering based of those should get pretty close to your incident reading.

It is rather simpler to take a single incident meter and you will peg all the tones first time. Spot readings in most cameras use a very wide 5 degrees or so and unless you have a known sample grey, are not that useful.

 

[Terrywoodenpic]

...is likely to be scared witless! I know what it all means and it's still frightening - moved to Talk Photography, we [the staff] don't think this is a 'basics' subject, we really are trying to keep that section for those questions that are only one step up from 'how do I turn the camera on?' However, if you prefer, it could also be moved to Talk Equipment I guess, just holler via the staffroom if you want that.

I do not understand why any one is scared by incident light meters. it make it all so simple and compared to reflection meters, they are pretty much foolproof. Even a simple Weston meter with invercone works first time, almost every time.
My first Norwood director, that I bought in 1956 is still nearly as accurate as my modern Sekonic L398m.

It only gets complicated if you want to set ratios with flash and ambient... which neither of mine can do as they are ambient only. But a Minolta 1VF or VF even makes that easy. But I no longer have a need for that.

 

[Mr Will]

I do not understand why any one is scared by incident light meters. it make it all so simple and compared to reflection meters, they are pretty much foolproof. Even a simple Weston meter with invercone works first time, almost every time.
My first Norwood director, that I bought in 1956 is still nearly as accurate as my modern Sekonic L398m.

It only gets complicated if you want to set ratios with flash and ambient... which neither of mine can do as they are ambient only. But a Minolta 1VF or VF even makes that easy. But I no longer have a need for that.

I don't think that anyone is scared of them, they are just an unnecessary hassle for most situations and utterly impractical if your subject is at a distance or moving.

Normal scenes - evaluative is fine.
High contrast where exposure is critical, switch to spot and point at something mid-toned (or the border between black and white if no midtones exist). Does not need to be grey.
Camera without a spot meter? Hold your hand in front of the subject and take a meter reading off your hand.
Really want to be a perfectionist? Use a grey-card. This will get your white-balance correct as well as your exposure (which in incident meter won't).

You seem to be looking for a problem that doesn't exist.

 

[Terrywoodenpic]

I don't think that anyone is scared of them, they are just an unnecessary hassle for most situations and utterly impractical if your subject is at a distance or moving.

Normal scenes - evaluative is fine.
High contrast where exposure is critical, switch to spot and point at something mid-toned (or the border between black and white if no midtones exist). Does not need to be grey.
Camera without a spot meter? Hold your hand in front of the subject and take a meter reading off your hand.
Really want to be a perfectionist? Use a grey-card. This will get your white-balance correct as well as your exposure (which in incident meter won't).

You seem to be looking for a problem that doesn't exist.

That is my point
They are less of a hassle and more accurate than any other method.

They are no problem at all for distant or moving subjects as they are measuring the incident light not the moving subject.
The light is likely to be the same near the camera position as it is on a distant subject. They are also equally good for aerial photography. Norwood made his first one for exactly this use, when in the US army airforce.

 

[Mr Will]

That is my point
They are less of a hassle and more accurate than any other method.

They are no problem at all for distant or moving subjects as they are measuring the incident light not the moving subject.
The light is likely to be the same near the camera position as it is on a distant subject. They are also equally good for aerial photography. Norwood made his first one for exactly this use, when in the US army airforce.

Taking an extra meter reading rather than using the camera supplied one - that's extra hassle.

No problems at all for distant subjects? What if you are in shade and the subject in sunlight (or vice versa)?
No problems with moving subjects? What happens if they move away before you've figured out where the light is coming from, taken your meter reading, set the camera and taken the shot?

You seem blind to the fact that of your original three shots; the reflective shot was overexposed, the incident one was underexposed and the only one that was correct was the one you compensated for. How does that prove it is any more accurate?

(I can see the benefit for aerial photography. By it's very nature you are taking photos of large expanses of things of varying reflectivity that are in the same lighting conditions as you are. What relevance this has to the current discussion I am not sure. It's like me pointing out that you can't use an incident meter to photograph the moon.)

 

[Terrywoodenpic]

Taking an extra meter reading rather than using the camera supplied one - that's extra hassle.

No problems at all for distant subjects? What if you are in shade and the subject in sunlight (or vice versa)?
No problems with moving subjects? What happens if they move away before you've figured out where the light is coming from, taken your meter reading, set the camera and taken the shot?

You seem blind to the fact that of your original three shots; the reflective shot was overexposed, the incident one was underexposed and the only one that was correct was the one you compensated for. How does that prove it is any more accurate?

(I can see the benefit for aerial photography. By it's very nature you are taking photos of large expanses of things of varying reflectivity that are in the same lighting conditions as you are. What relevance this has to the current discussion I am not sure. It's like me pointing out that you can't use an incident meter to photograph the moon.)

Of the three shots the middle one is very like the original indeed. (The light only measured under 80 foot candles) it is almost perfectly exposed.
However you might prefer the over exposed version even if many areas are desaturated. I have noticed a tendency for many digital photographers to prefer their shots with very open shadows, even if this loses texture and definition. However if the compensation had been about a 1/3 stop, more people might prefer it.( as previously discussed)

Those who regularly use incident meters know that the sunlight, and over cast light, changes very little over an hour or so. They take readings to predetermine their settings. If you are in shadow and have not measured the sunlight ... move.
Only with scurrying clouds is there a problem for all hand meters. but even that can be over come.

It is true you can not take a photograph of the moon with an incident light meter.... however using a camera might just help.

You may or may not be aware that virtually every cinema film you have ever seen, relied on incident light readings. As does a large majority of studio portraiture.

 

[Pookeyhead]

That is my point
They are less of a hassle and more accurate than any other method.

Sorry.. but sometimes they are a pain. If you're shooting fast.. impromptu/street/candid portraits in changing light... any action or sports in changing light.. in fact anything where you have to work fast, is not a great time to be using an incident meter.

I use them all the time for anything static, and in the studio... anything else, I'll use the camera's matrix metering and experience... it's faster.

The fact is though, I've never experienced any of the problems you demonstrate in your original post. It doesn't even look like a particularly tricky situation to meter. Incidentally (forgive the pun)... where were you pointing the meter?

I think it's a bit unfair to say that people are "scared" of them just because they don't use them. They're just not necessary a great deal of the time.

Plus.. yes I can incident meter on a sunny day... and that will be valid for some time... until I turn around of course... then I have to meter again... etc etc.. sorry... they're a pain if you want to work fast and unobtrusively. They have their place.. which is when you have the time to use it. Anything that has the potential to prevent you getting the image you set out to take, is something that has no place in your work flow for that job.

Right tools for the job and all that.

I have been using Incident, reflection and comparative photometers professionally for some 60 years and my technique in this is not in question.

Sorry for doubting you, but you surely must realise you posted an impossibility the first time around, and hence why you were being called into question

 

[Terrywoodenpic]

Sorry.. but sometimes they are a pain. If you're shooting fast.. impromptu/street/candid portraits in changing light... any action or sports in changing light.. in fact anything where you have to work fast, is not a great time to be using an incident meter.

I use them all the time for anything static, and in the studio... anything else, I'll use the camera's matrix metering and experience... it's faster.

Of course I agree with that. I use what ever is appropriate at the time.

The fact is though, I've never experienced any of the problems you demonstrate in your original post. It doesn't even look like a particularly tricky situation to meter. Incidentally (forgive the pun)... where were you pointing the meter?

Not actually difficult at all... the first shot I took like a beginner, using the auto matrix with aperture priority, framing strait ahead, and of course it could not cope with all that strong black and recorded out a mid grey as expected.
Normally one would look for a neutral grey and use the exposure lock... and bobs your uncle.
The light was unusual in that it only came from the sides with nothing from the front. So would probably have given a rather strange reading from a grey card as even a slight directional change would have magnified any exposure difference.

For the second shot I pointed the dome at the camera as per the accepted method... though I did check what would happen if I pointed to the stronger sidelight, as there has been a lot of argument about that on the web recently... of course it indicated a slightly shorter exposure which would have further darkened the shot (definitely not what was needed)

I think it's a bit unfair to say that people are "scared" of them just because they don't use them. They're just not necessary a great deal of the time.

I was only remarking on what someone else had said about themselves. I can see nothing scary about any sort of meter. But a SEI Photometer would probably have me stretching my memory rather more than somewhat.

Plus.. yes I can incident meter on a sunny day... and that will be valid for some time... until I turn around of course... then I have to meter again... etc etc.. sorry... they're a pain if you want to work fast and unobtrusively. They have their place.. which is when you have the time to use it. Anything that has the potential to prevent you getting the image you set out to take, is something that has no place in your work flow for that job.

Right tools for the job and all that.

People do seem to limit them selves to what their particular camera offers in the way of metering, which I find somewhat sad. There are always workarounds, but people new to photography who want to learn all the techniques, should not dismiss hand held meters and in particular Incident light meters from their armoury. They are still the best way to balance light and set ratios, or peg tones to match the original. But you already know this. Unfortunately many do not believe it.

My ancient Norwood super Director is actually extremely fast to use as it reads directly in F stops and at any pre-set shutter speed or Iso... but very few were ever made in the two years that Waltz held the patent licence. And now the helio valve (like a lens diaphragm) is worn and difficult to set.... so it is retired like me.

Sorry for doubting you, but you surely must realise you posted an impossibility the first time around, and hence why you were being called into question

No worries... It had me puzzled for some time... I had forgotten I had played around after having a sandwich lunch, and had taken another shot in the A priority mode. Same subject quite different light level. And of course such things fall butter side down.

 

[AESamuel]

I'm sorry but I really don't see a point in this thread at all

You've got three photos, the first is overexposed, the second is underexposed and the third is roughly right. You're preaching incident light meters but the best exposed picture (as shown by the histogram too) is one where you took a reading and then adjusted it. Why take an incident reading and then adjust it when you can use the camera's meter and then adjust it?

 

[Terrywoodenpic]

I'm sorry but I really don't see a point in this thread at all

You've got three photos, the first is overexposed, the second is underexposed and the third is roughly right. You're preaching incident light meters but the best exposed picture (as shown by the histogram too) is one where you took a reading and then adjusted it. Why take an incident reading and then adjust it when you can use the camera's meter and then adjust it?

Then it is not for you.....
and it is clear you have not read or perhaps understood what I have written.

For your Information the middle shot is the correct exposure, with the tones and colours matching the original.

The one you like is washed out and has lost detail and saturation in all the tones, as explained in post#16

A modern trend has become for some to prefer somewhat over exposed shots as they can see more in the dark tones, but at the expense of the other tones.

 

[Nod]

Purely because nobody else has said anything...

What do we expect from Cannon (sic) shooters?

 

[ding76uk]

Is there any point to this thread?

 

[Terrywoodenpic]

Is there any point to this thread?

I can't see a point to some of the replies.

 

[tijuana taxi]

Well I was interested thanks Terry, have been experimenting myself with incident light readings using an old Minolta Autometer III

Not sure about current trends, but trying to see details in everything is nigh on impossible. Why anyone wants to try and get detail into something like sunlit water is beyond me, human eye doesn't detect it so why worry
Same goes for minute detail in very distant objects, again normal vision cannot see it so what does it matter

I am also more interested in getting the correct tone and colour, if incident metering does that i'm all for it
(obviously this method can only be used under certain conditions)

Cheers
Rich

 

[Terrywoodenpic]

Well I was interested thanks Terry, have been experimenting myself with incident light readings using an old Minolta Autometer III

Not sure about current trends, but trying to see details in everything is nigh on impossible. Why anyone wants to try and get detail into something like sunlit water is beyond me, human eye doesn't detect it so why worry
Same goes for minute detail in very distant objects, again normal vision cannot see it so what does it matter

I am also more interested in getting the correct tone and colour, if incident metering does that i'm all for it
(obviously this method can only be used under certain conditions)

Cheers
Rich

The Minolta Autometer lll was an excellent meter and I expect most still will be.

The limitations of an incident meter are being able to get in to position, to measure the light falling on the subject. though this can often be done from another position... Outside, the sunlight falling in your garden will be the same as the sunlit hill in the distance so there is no need to go for a long walk.
A printed photograph or a photograph on a computer screen can only cope with a limited brightness range., while a camera can capture a far higher one. However when we process the file we have to make a choice of what part of that range we want to see. We can do this by compressing the tonal range ( making it less contrasty) which changes the relative tones or letting some tones either burn out or get lost in the shadow region. Or a combination of the two.

The Incident reading guarantees to reproduce the tones as you see them. But that is not the end of the story, as this still leaves you with the same optional choices as listed above.
Reflected readings often show more detail in the deep shadows, but lighter middle tones and highlight areas. This over exposure leads to desaturation of colours. It is the over exposure that causes this, not the meter method used.
The actual colour is determined by the camera and the colour balance that is set.

It is good to know that what the meter reads, will give the least compromised tones and saturation, and this is the case with Incident readings.

 

[Vince1551]

I'll blast some bracketing shots. Yes yes I'm lazy

 

[LCPete]

Very interesting Terry
I've not used an incident light meter before
For what I do, mainly wildlife stuff probably too fiddly to set up but still looks like a useful tool for many situations
I would say that the middle shot is correct but with raw I guess you could correct the bottom one to be the same
That's what I do anyway expose to the right
Would be interesting to see if there was any difference in the final result between correcting an ETTR raw and an incident light metered shot like the middle one

 

[Terrywoodenpic]

I will have a try with the bottom raw later.
I did try to "Improve the middle one in Photoshop yesterday. even opening it up 1/3 of a stop lost tonality and saturation. though a little burning did help to bring it back, but it was hardly worth the effort. If you look closely at the histogram of no 2 it contains tones to the right, it is Just that there are not many of them in the subject, but what there are in the grasses and vase and on the wheel are important so exposing further to the right just tends to lose detail or burn out these areas. which is what has happened in No3.
But as i say I will have a shot working the raw file from 3.

 

[Terrywoodenpic]

I'll blast some bracketing shots. Yes yes I'm lazy

The shot gun principal certainly works but give you more work later.
It will give the variation of exposure you need, but also give three Different pictures.

I hate it when the first has the best exposure and the last the best subject.

 

[Terrywoodenpic]

This is a Rework of number three from its raw file...
I had to reduce exposure by 1.35 increase vibrancy by +24, saturation by +8 and create a large s curve.

In photoshop cs3 I had to increase red by +8 and reduce blue by -8

All in all it was quite a fiddle to try to Get anywhere near Number 2.... which was arrived at with out doing any thing at all
even then it is not the same, just another version, that is a little too warm in comparison. and not as close to the real thing which is rather subdued. but I know some would prefer this greater saturation... which could be added to number 2 very simply. Not sure I would bother if I had a number to alter.


View attachment 7458