EV Compensation -- I should know... but

cwgrizz

Senior Member
Challenge Team
Thanks Paul, yes you are stating about what I concluded as stated in my post with the photos. I guess I am too technically minded and not artsy or flexible enough in my thinking. It has to be all "Black and White" No Grey. Ha!

I have to get into my head that it is not a "Hard and Fast" rule of settings, but just general guide. I wasn't far off in my thinking, just had to get it confirmed.
 

cwgrizz

Senior Member
Challenge Team
As regards @cwgrizz's shot of the bird in an earlier post, not sure what type of metering was being used, but seems like that would be a good case to use spot or center-weighted metering (as opposed to using matrix metering and then applying EC).


Sent from my iPhone using Tapatalk

It was "Center Weighted Metering" with a +1 EV set.
 

gustafson

Senior Member
Just on general principals I want to say no, but I dont know how it behaves.

Edit: It might work. I just took my lens off the d7100. With auto iso, I pointed the body with no lens at the ceiling light. I got a nicely exposed blur at iso 100. I pointed at a darker area and took another shot and got a nicely exposed blur at iso 6400.

Thanks for testing this out. I'm going to try this with a non CPU lens on a D3300 and report back. If it works, it will be huge for me!


Sent from my iPhone using Tapatalk
 

nickt

Senior Member
As regards @cwgrizz's shot of the bird in an earlier post, not sure what type of metering was being used, but seems like that would be a good case to use spot or center-weighted metering (as opposed to using matrix metering and then applying EC).


Sent from my iPhone using Tapatalk
That's mostly what I do. On the d7100 I have my function button programmed for spot metering, so its real easy to spot meter as needed. Half the time that I have used compensation, I forget to turn it off and end up messing up some shots.
 

gustafson

Senior Member
Just on general principals I want to say no, but I dont know how it behaves.

Edit: It might work. I just took my lens off the d7100. With auto iso, I pointed the body with no lens at the ceiling light. I got a nicely exposed blur at iso 100. I pointed at a darker area and took another shot and got a nicely exposed blur at iso 6400.

OK, I tried auto ISO in M mode with a non CPU lens on the D3300 and no dice. This suggests metering is required for auto ISO to work. Interestingly, the camera allowed me to set exposure compensation values, although they don't seem to go into effect. On an unrelated note, I found I could use the scroll wheel to change ISO (with the Fn button pressed; the button needs to be assigned to do this) instead of using the info button and rocker switch. When life gives you lemons ... :)


Sent from my iPhone using Tapatalk
 

nickt

Senior Member
OK, I tried auto ISO in M mode with a non CPU lens on the D3300 and no dice. This suggests metering is required for auto ISO to work. Interestingly, the camera allowed me to set exposure compensation values, although they don't seem to go into effect. On an unrelated note, I found I could use the scroll wheel to change ISO (with the Fn button pressed; the button needs to be assigned to do this) instead of using the info button and rocker switch. When life gives you lemons ... :)
Ah...interesting. It looks like its a difference in body features. I checked my manual, it says I get center weighted and spot metering with a non-cpu lens. No matrix metering. So that's why I was able to take a well-exposed blur with no lens. Then I checked the d3300 manual, it says you get no metering at all with a non-cpu lens. Sorry.
 

Horoscope Fish

Senior Member
I'll just toss out a few thoughts on Spot and Center Weighted metering and a few other topics germane to our conversation...

Nothing wrong with either mode and both will do a good job at keeping your subject properly exposed, which is key of course. The downside (if there is one) is that both modes will do so at the expense of everything else, generally speaking. In the example of the bird on the pole, Spot or CW would have exposed the subject properly but I'm betting the sky would have been blown out. If it didn't blow out the sky, well, then... You're golden. Also, if it did blow out the sky, but you're okay with a blown out sky, and sometimes I am; then again we're golden. If you're NOT okay with it, and it IS happening, then you have a Dynamic Range issue to deal with and the best way I know of to deal with a DR issue in a single frame is to use the principle of ETTR: Expose To The Right.

Exposing to the right means increasing overall exposure as much as we can, without blowing out any pixels. That last part is key: absolutely NO blown out pixels. Exposure is then tweaked to perfection during post processing (shooting RAW has a huuuuge advantage here). And, since there is so much more dat0 contained in the highlights than there is in the shadows, it's far easier, and far cleaner (meaning there will be less noise) for the software to decrease a highlight than it is to increase a Shadow. This is one of the huge benefits to shooting RAW and also highlights (ha!) why I think a thorough understanding of how to read and use Histograms is so important.
......
 
Last edited:

hark

Administrator
Staff member
Super Mod
When it comes to scenes that are overly bright or overly dark, the adage is:

When it's bright, go brighter. When it's dark, go darker.

Since the camera will meter at 18% grey, a really bright scene will automatically be darkened by the camera. So you need to brighten the exposure. And when the scene is dark, the camera will make it lighter...so you need to darken the exposure.

You also have the option to change your method of metering. Sometimes matrix metering is way off depending on the scene. Try out center weighted metering and spot metering when you are in high contrast situations to see if one yields more accurate results. You can also take a meter reading of the grass--sometimes that will yield a more accurate exposure.
 

gustafson

Senior Member
I'll just toss out a few thoughts on Spot and Center Weighted metering and a few other topics germane to our conversation...

Nothing wrong with either mode and both will do a good job at keeping your subject properly exposed, which is key of course. The downside (if there is one) is that both modes will do so at the expense of everything else, generally speaking. In the example of the bird on the pole, Spot or CW would have exposed the subject properly but I'm betting the sky would have been blown out. If it didn't blow out the sky, well, then... You're golden. Also, if it did blow out the sky, but you're okay with a blown out sky, and sometimes I am; then again we're golden. If you're NOT okay with it, and it IS happening, then you have a Dynamic Range issue to deal with and the best way I know of to deal with a DR issue in a single frame is to use the principle of ETTR: Expose To The Right.

Exposing to the right means increasing overall exposure as much as we can, without blowing out any pixels. That last part is key: absolutely NO blown out pixels. Exposure is then tweaked to perfection during post processing (shooting RAW has a huuuuge advantage here). And, since there is so much more dat0 contained in the highlights than there is in the shadows, it's far easier, and far cleaner (meaning there will be less noise) for the software to decrease a highlight than it is to increase a Shadow. This is one of the huge benefits to shooting RAW and also highlights (ha!) why I think a thorough understanding of how to read and use Histograms is so important.
......

Thanks for the insights, very helpful.

A couple of follow up questions:

Is Active D-Lighting relevant to this discussion? I have mine turned off, but wondering if it's an effective and idiot-proof option for high contrast situations.

Second, can you point me to a primer on interpreting and leveraging the in-camera histogram to optimize exposure? Or provide a quick-and-dirty guide on how to do it?

My understanding of the histogram is rather poor. All I know is you want to avoid having the bell curves way over to the right (underexposed?) or left (overexposed?) or clipped (blown highlights?). However, in practice, I find acceptable images that break these rules. Case in point: I took some moon shots the other day, and noticed that the curves were all smushed to the left on the histogram, and clipped. When I tweaked exposure to get what I thought were acceptable shots, I saw that the histograms hadn't changed very much at all.

To your other point about avoiding blown pixels, I've seen whenever I have a light source or a strong reflection thereof in the image, I get the blinking indicator in the "Highlights" screen in playback mode. I'd imagine if I dialed down exposure to avoid blown pixels, then the rest of the image would underexposed. Is that correct or am I missing something?

Thanks in advance!


Sent from my iPhone using Tapatalk
 
Last edited:

Horoscope Fish

Senior Member
Thanks for the insights, very helpful.

A couple of follow up questions:

Is Active D-Lighting relevant to this discussion? I have mine turned of, but wondering if it's an effective and idiot-proof option for high contrast situations.
We can certainly make it relevant! :D

Active D Lighting is applied to JPG's only, or I suppose if you use Nikon's post-processing software (NX-D, et al) the effect can be applied to RAW files. I'm a little unclear on the details of that since, 1. I shoot RAW exclusively and, 2. I don't use Nikon's post processing software; not that there's anything wrong with it. Fact is, I don't like ADL and that's because in my limited experience with it, I saw it introduce a lot of noise in the shadows; the more strongly it was applied, the more noise I saw in the shadows. Used sparingly, I guess it could help but in my estimation, and this is just me talking here, ADL is solution in search of a problem. If you've decided you want to control your highlights, shadows and midtones, shoot RAW and go nuts; and I say that without taking anything away from those that shoot JPG. I like JPG's just fine. But we also have to be honest and the simple fact is, you gain a huge degree of control over your exposure when you shoot RAW and that's why most of us turn to it. If you shoot JPG I can suggest you try ADL and see what you think. I never want anyone to take what I say for granted.


Second, can you point me to a primer on interpreting and leveraging the in-camera histogram to optimize exposure? Or provide a quick-and-dirty guide on how to do it?
I shore can! One of my favorite tutorials on Histograms comes from Luminious Landscape, a great site full of superb tutorials: Understanding Histograms. Once you've read that tutorial, and reeeally understand it, move on to: A Practical Guide to Understanding RGB Histograms by Steve Hoffman.


My understanding of the histogram is rather poor. All I know is you want to avoid having the bell curves way over to the right (underexposed?) or left (overexposed?) or clipped (blown highlights?). However, in practice, I find acceptable images that break these rules. Case in point: I took some moon shots the other day, and noticed that the curves were all smushed to the left on the histogram, and clipped. When I tweaked exposure to get what I thought were acceptable shots, I saw that the histograms hadn't changed very much at all.
The above tutorials should explain all of this. What you are describing is perfectly normal, you just don't know why yet. Once you do, you're going to wonder how you ever got along without understanding histograms. They're freaking awesome.


To your other point about avoiding blown pixels, I've seen whenever I have a light source or a strong reflection thereof in the image, I get the blinking indicator in the "Highlights" screen in playback mode. I'd imagine if I dialed down exposure to avoid blown pixels, then the rest of the image would underexposed. Is that correct or am I missing something?
You're perfectly on track here. To protect the highlights the shadow regions of a shot will frequently be underexposed to some degree. To fix this the shadows are lifted in post-processing. Now if the dynamic range is way high, this is going to be tough... Shooting a black cat sitting on a pristine white snow bank under a clear sky with full sun is going to challenge just about any DSLR currently on the market which is where learning to do HDR photography comes into play, but... Doing HDR requires multiple shots and I'm limiting this discussion to single-frame exposures. So, back to your question... Yes, to protect the highlights you shoot them as close to the edge of your histogram as you can, without blowing out any pixels (ETTR). This will bring the shadows up as much as possible, while at the same time keeping the highlights as high as possible (no blowouts allowd) and the final balancing is performed in post-processing where it's almost always better to pull something down (subtract from what is already there) than it is to try and add something that is not already there. I'm not sure that last little bit makes sense but hopefully you're catching my drift.
.....
 

gustafson

Senior Member
... and the final balancing is performed in post-processing where it's almost always better to pull something down (subtract from what is already there) than it is to try and add something that is not already there. I'm not sure that last little bit makes sense but hopefully you're catching my drift.
.....

Paul, thanks for the detailed response, looks like I have some great reading ahead of me this weekend But sincerely, much appreciated!

That last bit made sense to me. I'm guessing the rationale is our cameras can tell us if we're blowing pixels, but not if our shadows are too dark to recover from, so we're better off working backward from inside the edge of blown highlights. Also seems logical that we would use as much of the light available to us as possible (as it is information about the image) as opposed to dialing it down and losing image data. Ha! I'm starting to sound like a pro already





Sent from my iPhone using Tapatalk
 

Horoscope Fish

Senior Member
Paul, thanks for the detailed response, looks like I have some great reading ahead of me this weekend But sincerely, much appreciated!
You're very welcome.


That last bit made sense to me. I'm guessing the rationale is our cameras can tell us if we're blowing pixels, but not if our shadows are too dark to recover from, so we're better off working backward from inside the edge of blown highlights. Also seems logical that we would use as much of the light available to us as possible (as it is information about the image) as opposed to dialing it down and losing image data.
Well the Histogram will tell you if you've blown out a tonal range in the shadows just the same as if you've blown out a range in the highlights -- and neither is good (your histogram tutorials will explain all of this); because once you've blown out a tonal range it means you've lost detail in that range, detail that can not be recovered. A tonal range that has blown out can be played with but not very well so that situation is simply best avoided.
.....
 

gustafson

Senior Member
You're very welcome.



Well the Histogram will tell you if you've blown out a tonal range in the shadows just the same as if you've blown out a range in the highlights -- and neither is good (your histogram tutorials will explain all of this); because once you've blown out a tonal range it means you've lost detail in that range, detail that can not be recovered. A tonal range that has blown out can be played with but not very well so that situation is simply best avoided.
.....

Thanks for clarifying! I've printed and bagged your reading assignments and can't wait to uncover the secrets they hold.


Sent from my iPhone using Tapatalk
 

gustafson

Senior Member
I shore can! One of my favorite tutorials on Histograms comes from Luminious Landscape, a great site full of superb tutorials: Understanding Histograms. Once you've read that tutorial, and reeeally understand it, move on to: A Practical Guide to Understanding RGB Histograms by Steve Hoffman.

Just reporting back on my weekend reading. Highly edifying. Will have to read Steve Hoffman's guide a second time (or a third) to fully grasp the material. It really gets into the weeds.

On a related note, stumbled across Ken Rockwell's article on Adaptive D-Lighting, for what its worth. Doesn't cover whether it applies to RAW or JPEG, but he makes a distinction between D-Lighting in-camera vs. during post-processing, which was news to me. I plan to tinker with it on my D3300 and report back here.

Nikon ADR: Adaptive Dynamic Range
 

nickt

Senior Member
On a related note, stumbled across Ken Rockwell's article on Adaptive D-Lighting, for what its worth. Doesn't cover whether it applies to RAW or JPEG, but he makes a distinction between D-Lighting in-camera vs. during post-processing, which was news to me.
Active d lighting is mainly a jpg thing. If you shoot raw, you should not use it, but if you use only Nikon processing software, then you might benefit from it in raw too. There are details though. Used at lower levels, its just a jpg thing. If you push it to the higher levels, it also alters the exposure a bit and that will affect your raw file by under exposing it a bit. I don't have a specific source, but if you google "does active-d lighting affect exposure" you will find some forum talk and articles.
Bottom line is that most people here would say turn it off if you are shooting raw.
 
Top