Exposing for RAW .

Joined
Sep 17, 2006
Messages
12,072
Location
Hamilton , New Zealand
In another discussion the topic of how to expose for RAW came up and I got a little confused . I suggested "exposing to the right " to allow for more information to be adjusted later . Someone mentioned that this could lead to under-exposure because the jpeg may give misleading results .
My question is that if I did shoot in RAW and wanted to have the maximum information to recover later wouldn't exposing to the right , or over-exposing slightly but without blowing highlights , be the correct approach ?
That's assuming that what I understand to be "exposing to the right " is when you increase exposure to make sure the histogram touches the right hand side [ in extreme cases ] so that you have the maximum shadow detail without losing highlight detail .
ok , I understand that when the histogram touches the right it may just be a warning that detail may be lost but I don't understand how this can lead to an under-exposed RAW image .
Also , how far can you push the histogram to the right without being able to recover highlights ?
 
Joined
May 11, 2006
Messages
1,916
Location
Northern Italy, Piemonte
You've got the right idea; however, your typical camera doesn't give you all the information you need to absolutely guarantee you won't lose some highlight detail. The blinking warning is just that, a warning that the highlights might be over exposed. If you always adjust exposure to avoid the blinking highlight warning, chances are you are underexposing at least some of your shots.

If you use good software, like Adobe's Photoshop, you can recover highlight detail if only one of the three channels is blown out. In addition, it is not unusual in real constrasty lighting to go ahead and sacrifice the highlights, let them blow completely out, and live with the consequences. For example, light coming through a window on an indoor shot. You might just go ahead and forget about it in the interest of getting shadow detail in areas that mean something to the photo.

Half the information you're trying to capture is contained in the upper stop of exposure. IOW, under exposing by a stop tosses half the information available; not good. Plus you'll be increasing noise; also not good.

So use the histogram as a guide (but not a guarantee) and watch out for a tall spike up against the right hand side, especially a WIDE spike. Keep trying to avoid a big blank space on the right side of the histogram. On super important photos, bracket to be safe. All in all, this isn't a lot different from shooting film.
 
Joined
Sep 17, 2006
Messages
12,072
Location
Hamilton , New Zealand
Half the information you're trying to capture is contained in the upper stop of exposure. IOW, under exposing by a stop tosses half the information available; not good. Plus you'll be increasing noise; also not good.

So use the histogram as a guide (but not a guarantee) and watch out for a tall spike up against the right hand side, especially a WIDE spike. Keep trying to avoid a big blank space on the right side of the histogram. On super important photos, bracket to be safe. All in all, this isn't a lot different from shooting film.
Isn't shooting in RAW bracketing by 4 stops in a way ? , maybe by 2 stops in shadow details at least if you under expose the highlights .
Also , when you brighten shadows later is that effectively "boosting the iso " in the shadow detail ?
 
Joined
May 11, 2006
Messages
1,916
Location
Northern Italy, Piemonte
Isn't shooting in RAW bracketing by 4 stops in a way ? , maybe by 2 stops in shadow details at least if you under expose the highlights .
Also , when you brighten shadows later is that effectively "boosting the iso " in the shadow detail ?
Nope, Nope (NEVER UNDER EXPOSE THE HIGHLIGHTS), and well.........sort of.

The idea that somehow shooting raw is an excuse to get away with poor exposure or that somehow it doubles the dynamic range of your sensor is a myth.

Anytime you do much of anything in post processing you stress the image data and that introduces noise and posterization. A raw file is MUCH more resistant to this sort of degradation than a jpg file and that is the primary reason I shoot in raw. There are other advantages of course, but they are secondary.
 
Joined
Sep 17, 2006
Messages
12,072
Location
Hamilton , New Zealand
Thanks for the replies . Actually I didn't mean "under expose" the highlights , I meant 'if the highlights are correctly exposed and don't need adjustment we should at leats have another two stops of detail in the shadow areas .
Something interesting I have noted with pictures in National Geographic is that they always seem to be exposed for highlights and there are a lot of almost black shadows , and they look good !
I just got a good deal on a brand new D70S with the 1005 segment meter and am pleased with the results compared to my D80 that likes to blow highlights in favour of shadows .
 
Joined
Sep 17, 2006
Messages
12,072
Location
Hamilton , New Zealand
Hi Desmond!

This improvement in the pictures-source also produce less adjustments anf pp'ing at all. The 1005 segment meter, does a lot more as amount of pixels.
At next You should ty the d200, which will become pretty cheap this dasy, this will be another world to the d70s, which i hd for a week, to compare it to d5 and d80 results.

cheers tom
I tried a D200 and sold it after two weeks , a good camera , good results but soooo different to control compared to my other bodies - terrible battery performance too , compared to my D50 that did 3000 shots in one day and the battery level wasn't even showing 3/4 yet !
If all my cameras had those controls I would have kept it though .
 
Joined
Sep 17, 2006
Messages
12,072
Location
Hamilton , New Zealand
It depends on many things. Your not exposing for RAW, but exposing for the developer. Exposing to the right can be the wrong way. Human vision sees the most detail around zone 5. That is where I place my subjects. Most people don't know where a highlight begins, but they do know what the subject is. At least I hope so.

Here is some reading to get you started.

Zones and Digital: Two Methods of Exposing

Since we depend on the histogram for exposure judgement you might want to read this.
White Balance in Digital Cameras: Problems

Headroom in Highlights : Where is Zone V in The Digital World?

Peace in Lights

That should be enough to confuse everyone. :smile:
The first one was enough to confuse me . When I think of ''exposing to the right '' I imagined more in the line of adding exposure without blowing highlights to have more info to work with , but this article mentions that it is like dialing -2.5 EV ? Doesn't sound right .
The second article about wb makes sense , I suppose more so in lighting where the camera gets wb wrong - best to set manually there .
.... breezed through the last two articles - plenty to confuse anyone :eek:
Im pretty pleased so far with my D70S's metering and will be trying to get the right metering with my D80 to get similar results , I'm busy playing around with centre weighted on the D80 now .
 
Joined
May 11, 2006
Messages
1,916
Location
Northern Italy, Piemonte
The first one was enough to confuse me . When I think of ''exposing to the right '' I imagined more in the line of adding exposure without blowing highlights to have more info to work with , but this article mentions that it is like dialing -2.5 EV ? Doesn't sound right ........snip.........
No wonder you're confused by the first article. It's got so many errors that it's almost impossible to understand what the author had in mind.

I wonder if English is the author's second language and something got lost in the translation. The names and the bottom and the fact that there is a Russian version makes me think that it could be a misunderstanding by the author.

In any event, it is quite clear that whoever wrote the article didn't know what he was talking about and some statements are exactly backwards from what they should be. Or, perhaps he got something reversed in the translation to Russian. Who knows? ............. Confusing???:confused::confused::confused:

You bet.

Hint: Do not, as the author seems to suggest, under-expose your shots by 2.5 stops.:eek:
 
Joined
Sep 17, 2006
Messages
12,072
Location
Hamilton , New Zealand
No wonder you're confused by the first article. It's got so many errors that it's almost impossible to understand what the author had in mind.

I wonder if English is the author's second language and something got lost in the translation. The names and the bottom and the fact that there is a Russian version makes me think that it could be a misunderstanding by the author.

In any event, it is quite clear that whoever wrote the article didn't know what he was talking about and some statements are exactly backwards from what they should be. Or, perhaps he got something reversed in the translation to Russian. Who knows? ............. Confusing???:confused::confused::confused:

You bet.

Hint: Do not, as the author seems to suggest, under-expose your shots by 2.5 stops.:eek:
I think he was saying that you should not ettr because it would be like dialing in -2.5 in spot metering mode .
 
Joined
Apr 24, 2005
Messages
6,099
Location
Denver, CO
English is not his native language, but nothing is incorrect about his article Bob.

Just forget it. I have went through this before on the cafe and its not worth my time anymore.

Have a good day.
 
Joined
Jan 26, 2008
Messages
4,971
Location
Huntington Station, NY
This is a good thread. I've been doing a lot of experiments and testing with exposure, clipping highlights, and a variety of other things. There's no substitute for correct exposure but my observation is that underexposure, especially at higher ISO can be disastrous.

I try to push my exposure as much 'to the right' as I can without clipping highlights, but I have noticed that the blinkies seem to be more of a 'yellow light' than a red one. All I mean is that blinkies seem to tell you what's about to come rather than what is actually happening.

I've taken many images that I've seen blinkies, opened them in Photoshop and checked the histogram and channel values, only to find that there is no clipping or just barely a hint of clipping. When there's only a hint of highlight clipping I don't find there to be any loss of detail.
 
Joined
Jan 26, 2005
Messages
15,253
Location
Marysville, WA
I've taken many images that I've seen blinkies, opened them in Photoshop and checked the histogram and channel values, only to find that there is no clipping or just barely a hint of clipping. When there's only a hint of highlight clipping I don't find there to be any loss of detail.
I have noticed exactly the same, both with NX2 and CS3. I think I remember reading somewhere or hearing from someone, so please consider this at best 4th hand and quite possibly totally false, that the "newer cameras" do start the Blinkies at a lower level. Now, if I have a Bald Eagle whose entire head is blinking, I adjust exposure, but certainly for a little "blink" here and there, especially on a specular highlight, no rpblem at all.

Which always brings up the point when evaluating a historgram, that you need to understand the scene as well. A white bird across a dark backround will show a historgram far to the left of the same bird against a bright blue sky, and yet both are correct.
 
Joined
Jan 26, 2008
Messages
4,971
Location
Huntington Station, NY
I have noticed exactly the same, both with NX2 and CS3. I think I remember reading somewhere or hearing from someone, so please consider this at best 4th hand and quite possibly totally false, that the "newer cameras" do start the Blinkies at a lower level. Now, if I have a Bald Eagle whose entire head is blinking, I adjust exposure, but certainly for a little "blink" here and there, especially on a specular highlight, no rpblem at all.

Which always brings up the point when evaluating a historgram, that you need to understand the scene as well. A white bird across a dark backround will show a historgram far to the left of the same bird against a bright blue sky, and yet both are correct.
Absolutely Bill. I have always said that the histogram is useful, it's a tool, and a rough representation. BUT... the histogram rarely tells the whole story and that is something we still have to evaluate for ourselves.
 
Joined
Jan 26, 2005
Messages
15,253
Location
Marysville, WA
Absolutely Bill. I have always said that the histogram is useful, it's a tool, and a rough representation. BUT... the histogram rarely tells the whole story and that is something we still have to evaluate for ourselves.
I remember this "really smart guy" a number of years ago, who really believed the "ETTR" mantra, to the point where said "really smart guy" would make sure that the Histogram always showed good coverage in the upper quadrant. After looking at many images that were, to be kind to this "really smart person" just a bit ( :wink::wink: ) overexposed, I finally got smart and learned how to read the historgram and the scene :redface::redface:

And, by the way, I really am a "really smart guy", who at times gets hung up on the latest "this is the RIGHT way" to do "it", whatever "it" is.

I still don't have a clue, but I get far fewer over-exposed images these days :wink:

The links from Charles that were referenced are much the same. The information is there, to be used along with the other information and tools we have at our disposal.
 
T

TonyBeach

Guest
I've taken many images that I've seen blinkies, opened them in Photoshop and checked the histogram and channel values, only to find that there is no clipping or just barely a hint of clipping. When there's only a hint of highlight clipping I don't find there to be any loss of detail.
Several settings adjustments in the camera contribute to a false indication of overexposure -- and most of them are the default settings. Start with color space, if you open a file in Photoshop you may be opening it in ProPhoto RGB, and that gamut is much wider than sRGB or even Adobe RGB (both of which are smaller than the camera's gamut). Contrast and saturation are also primary causes of misleading information from the camera about exposure latitude, and even sharpening can change the results (those halos appear as clipped highlights).

This is why I constantly point out that shooting JPEG and RAW simultaneously is not optimal. If you adjust your camera's settings so that your JPEG more accurately reflects the RAW data, then the JPEG is a mess; if you set your camera's settings so that your JPEG looks good, then you are underexposing your RAW data -- unless you blew it, then you might just have a perfect RAW file by accident.
 
Joined
Jan 26, 2005
Messages
15,253
Location
Marysville, WA
if you set your camera's settings so that your JPEG looks good, then you are underexposing your RAW data -- unless you blew it, then you might just have a perfect RAW file by accident.
Or you do what many of us have learned to do, which is to set your in-camera settings to reflect enough to judge on the LCD while understanding how the histogram and RAW are affected. What I have learned mostly from this is evaluation of the scene, which is often difficult with the subjects I tend to shoot as they are often moving rapidly from shadows to well lit areas. This is also why the very first step in my NX2 processing is to turn off all the in-camera settings. Where I entirely agree with you is that many factors are involved, and you certainly can't just blindly rely what you see on the LCD. I mean, how the heck did we ever live without them???? :wink:
 
T

TonyBeach

Guest
Or you do what many of us have learned to do, which is to set your in-camera settings to reflect enough to judge on the LCD while understanding how the histogram and RAW are affected.
Right, that would explain all the "my reds blow out" threads I see. The vast majority of photographers, many of them seasoned, do not really understand what's going on in the RAW data and rarely see it outside of color spaces and with gamma and contrast applied. The result of this widespread misunderstanding is lots of underexposed images and demands for better cameras to make up for the photographer's deficiencies.

I mean, how the heck did we ever live without them [histogram displays]?
http://www.bythom.com/filmtodigital.htm

Many photographers want to use their in-camera histogram as a substitute for the light meter they used to carry around when shooting film. The transition from slide film to digital though usually misses a crucial factor and the one that most effects these in-camera histograms, and that is WB gain. Film had a native WB, and so do digital sensors; it's just that when we adjust film we do it with a filter over the lens, but when we adjust digital it is done by digital multiplication of the red and blue channels and by defining green somewhere between absolute magenta and absolute green.
 
Joined
Sep 17, 2006
Messages
12,072
Location
Hamilton , New Zealand
; if you set your camera's settings so that your JPEG looks good, then you are underexposing your RAW data .
But since there are 2 stops of adjustment either way with RAW would that not still mean that you have at least two stops to add to the image later ? [understanding of course possibilities of noise issues ] .
I understand that a lot of it has to do with "reading the scene " as well as the histogram and that there are times when blown highlights are unavoidable . I have also played around with manual and taking an exposure off the back of my hand which has yielded better results than my D80's 'miserable matrix metering' [ m3 from now on ! ]
 
T

TonyBeach

Guest
But since there are 2 stops of adjustment either way with RAW would that not still mean that you have at least two stops to add to the image later ? [understanding of course possibilities of noise issues.
Who told you there are two stops on either side of exposure latitude? There are not. The difference between RAW and JPEG is about one stop of headroom, so you are always underexposing shooting JPEG. That means that most photographers are already close to one stop under optimally exposed base ISO. If you underexpose by another two stops, you are three stops below base ISO; that's like going from ISO 100 to ISO 800, and if you start at a higher ISO you will end up at an even higher equivalent ISO.
 
Joined
Sep 17, 2006
Messages
12,072
Location
Hamilton , New Zealand
Who told you there are two stops on either side of exposure latitude? There are not. The difference between RAW and JPEG is about one stop of headroom, so you are always underexposing shooting JPEG. That means that most photographers are already close to one stop under optimally exposed base ISO. If you underexpose by another two stops, you are three stops below base ISO; that's like going from ISO 100 to ISO 800, and if you start at a higher ISO you will end up at an even higher equivalent ISO.
I had read before that jpegs have around1/3 of a stop lattitude and RAW has two stops [ hmmmm , maybe not either way ] but when I open a RAW in NX the base adjustments allow for two stops adjustment either way ?
Am I making the wrong inferences from this ?
 
Links on this page may be to our affiliates. Sales through affiliate links may benefit this site.
Nikon Cafe is a fan site and not associated with Nikon Corporation.
Forum post reactions by Twemoji: https://github.com/twitter/twemoji
Copyright © 2005-2019 Amin Forums, LLC
Top Bottom