Why you would use EV (exposure compensation) settings

[relate22]

I am confused why you would use EV settings. If the camera sets the proper exposure shouldn't the brightness or darkness of the photo be correct.?

Could someone please explain when you would use a + or - adjustment using the in camera EV?

Thanks

Robert
--
http://www.pbase.com/reelate2
http://www.pbase.com/relate2
My youtube channel====> [ b] http://www.youtube.com/user/relate2?feature=mhsn
What flying means to me.
http://vimeo.com/2598837
Flying highlights
http://www.youtube.com/watch?v=0lRu3P15BaY
My Youtube Channel http://www.youtube.com/user/relate2#p/u

 

[Deleted1929]

Two reasons :

Your idea of "proper exposure" is a human perception which the camera lacks. The camera makes a guess based on some simple measurements it takes of the scene. This automatic exposure may not match what you want. So you can use exposure compensation to get what you want.

That's one reason.

Also the simple estimate the camera makes can be fooled by particular circumstances which are obvious exceptions for a human, but which the camera cannot know of, as it lacks the ability to spot the special case. An example would be a snowy scene, which is very bright because the snow reflects light more strongly than uncovered ground. You would typically need to use exposure compensation to correct this.

That's the other reason.

--
StephenG

 

[John Deerfield]

Well, you need to understand how exposure works. So I might suggest this:

http://www.amazon.com/Understanding-Exposure-3rd-Photographs-Camera/dp/0817439390/ref=sr_1_1?ie=UTF8&qid=1325230445&sr=8-1

In a nutshell, your camera ( any camera) wants to meter the scene to return a value of 18% gray. Now you can change the area that the camera meters within the frame, but regardless of the metering method , the metering result is to always return a value of 18% gray. 18% gray is the "middle-of-the-road". Now this works fine when your light is middle-of-the-road, not too bright or not too dark (because your camera metering is a reflective meter, but lets keep this simply [buy the book]). So, bright scenes or dark scenes can fool the camera meter and return an incorrect exposure. So, you adjust for the camera not metering the scene correctly by using Exposure Compensation. You dial in positive compensation to record a scene brighter and negative compensation to record a scene darker, than the camera's metering system thinks it should be.

 

[PrasadIndi]

Hi Robert,
I'm no pro so let me put this in simple terms...

Do a little experiment:
1. Put your camera on auto.
2. Take a shot of a pure deep black object. be sure to fill the frame with it.
3. Now take a similar shot of a pure white paper/object.

You'll be surprised to see that both the pictures turn out to be almost the same. This happens because when the camera 'sees' the object, it by default, thinks that it should be mid-tone gray (that's the 18% gray) and makes the exposure accordingly.

In other words, it 'Overexposes' the blacks and 'underexposes' the whites (to make them look like mid-tone grays).

Now to correct this you use the EV.

On the -ve side for black bodies and on the +ve side for white bodies. Hence the camera 'understands' that your objects are 'black' or 'white' and not 'gray'.

Try this out.
This is very crude but hope it helps

I am confused why you would use EV settings. If the camera sets the proper exposure shouldn't the brightness or darkness of the photo be correct.?

Could someone please explain when you would use a + or - adjustment using the in camera EV?

Thanks

Robert
--
http://www.pbase.com/reelate2
http://www.pbase.com/relate2
My youtube channel====> [ b] http://www.youtube.com/user/relate2?feature=mhsn
What flying means to me.
http://vimeo.com/2598837
Flying highlights
http://www.youtube.com/watch?v=0lRu3P15BaY
My Youtube Channel http://www.youtube.com/user/relate2#p/u

 

[V3rm3]

Also, besides what everyone says, your gear might not just give the desired result, myTamron 70-300 for example is a little over-exposing, it's a common issue with these lens, i could get it replaced and hope the other one doesn't have the issue, but since besides that i am perfectly happy with the lens, a 0.3 EV compensation fixes the problem

 

[MaryGierth]

Could someone please explain when you would use a + or - adjustment using the in camera EV?

As others have said you will use it to make the image brighter or darker.

An example of when - you've taken a pic but the sky is blown out (looks white not blue) so you dial in negative ec and retake the picture. The sky will look bluer but the other parts of the image will be darker as well so you have to decide what is the better exposure. Generally a little underexposure saturates colors so its common on some cameras to always use a little negative ec.

Basically you just look at the picture you've taken and decide if the important parts of the image look the way you want. If they are too dark or too light then you adjust the exposure via ec.

EC is not a cure all tho for bad lighting. You may have to use flash or make other adjustments to change the lighting and get a better exposure.

 

[relate22]

Thanks sjgcit, John Deerfield & PrasadIndi all very helpful and I am finally getting a hand on EV adjustment.
--
http://www.pbase.com/reelate2
http://www.pbase.com/relate2
My youtube channel====> [ b] http://www.youtube.com/user/relate2?feature=mhsn
What flying means to me.
http://vimeo.com/2598837
Flying highlights
http://www.youtube.com/watch?v=0lRu3P15BaY
My Youtube Channel http://www.youtube.com/user/relate2#p/u

 

[relate22]

Thanks V3rm3 and MaryGierth.
--
http://www.pbase.com/reelate2
http://www.pbase.com/relate2
My youtube channel====> [ b] http://www.youtube.com/user/relate2?feature=mhsn
What flying means to me.
http://vimeo.com/2598837
Flying highlights
http://www.youtube.com/watch?v=0lRu3P15BaY
My Youtube Channel http://www.youtube.com/user/relate2#p/u

 

[GarageBoy]

But intentionally under exposing for deeper skies kills shadow detail
Is it better to fix that in post?

 

[Leonard Migliore]

But intentionally under exposing for deeper skies kills shadow detail
Is it better to fix that in post?

Depends on the camera but, at least on newer DSLR's, you can pull up the shadows a lot without getting horrible noise. So I would, in general, not be afraid of underexposing, at least if I were shooting RAW.
--
Leonard Migliore

 

[John Deerfield]

So I would, in general, not be afraid of underexposing, at least if I were shooting RAW

If I were shooting RAW, I would intentionally overexpose to retain shadow detail.

http://www.digitalphotopro.com/technique/camera-technique/exposing-for-raw.html

 

[scorrpio]

So I would, in general, not be afraid of underexposing, at least if I were shooting RAW

If I were shooting RAW, I would intentionally overexpose to retain shadow detail.

http://www.digitalphotopro.com/technique/camera-technique/exposing-for-raw.html

It really depends on the scene. If your highlights have interesting details, you would want to underexpose in order to prevent blowing them out. If you want shadow detail and don't care if highlights blow out a bit, you overexpose.

 

[Doug Pardee]

The following is off-topic, but this is a forum for beginners, and there's no point in giving them bad information.

In a nutshell, your camera ( any camera) wants to meter the scene to return a value of 18% gray.

This is one of the great myths of photography, and dang it's hard to beat it down. In-camera metering and automatic exposure is far more complicated than all that, and "18%" doesn't figure into it at all (except in a rare case).

To start with, there are four pieces that need to fit together.

1. The in-camera metering system that measures the light.
2. The auto-exposure system that sets the exposure settings.
3. The ISO setting that the auto-exposure system uses.
4. The ISO setting that controls the sensitivity of the sensor.

In-camera metering systems used to be very simple. They measured the light from the whole screen and reported the Bv (brightness value) level. The Bv level calculation included a factor called "K" which was set by the manufacturer. Typically, K was chosen so that an exposure directly based on the reported Bv level would give an overall brightness level of 12-13% (not 18%).

Today, almost all cameras use some form of multi-zone metering — it might be called evaluative metering, matrix metering, pattern metering, or whatever. Multi-zone metering attempts to guess what kind of scene is being photographed, then make allowances for that by adjusting the reported Bv level up or down. Crude example: if the apparent subject seems a lot darker than the rest of the frame it's probably a back-lit subject, so the Bv level will be reported on the low side, closer to what the apparent subject is.

The auto-exposure system then takes the Bv measurement and attempts to set an appropriate exposure depending on the mode (P, A/Av, S/Tv, various scene modes, etc.) and — for some modern models — auto-ISO capabilities. In doing so, it uses the ISO value as an Exposure Index, and it also factors in the Exposure Compensation that is the topic of this whole thread.

After the image has been captured, the sensor data is read out using the ISO value to determine the amplification needed for the signals. The lower the ISO value, the lower the amplification.

So here's the thing...

There is no "18%" target for the image brightness. The manufacturer has total control over (1) how the multi-zone metering operates, (2) what K factor is chosen for the metering system, and (3) what the relationship is between the Exposure Index in auto-exposure and the sensor amplification at a given ISO setting. The manufacturer's goal is to try to produce what its customers will consider to be well-exposed images in as wide a variety of conditions as possible.

Very few auto-exposed images from a digital camera will be anywhere close to 18%. Most people find that to be way too dark for color images, and generally prefer something closer to 25%. So that's what the manufacturers aim for.

I mentioned a "rare case" where 18% does come in. The current ISO standard for digital camera sensitivity (ISO 12232:2006) has an option called Standard Output Sensitivity (SOS). A camera manufacturer can choose to rate the ISO sensitivity using SOS provided that the camera doesn't use multi-zone metering and the output format is sRGB JPEG. In that situation, the ISO SOS specification basically says that the output of the sRGB JPEG must be 18% plus or minus 1/3 stop.

SOS is rarely used by camera manufacturers, for a number of reasons. Obviously, the multi-zone metering and output formats other than sRGB JPEG get in the way. More significantly, almost everyone thinks that 18% outputs are seriously underexposed (at least for color photos). The Canon EOS Rebel XTi (400D) never officially used SOS, but the images it produces strongly suggest that there was an attempt to reduce the sensor amplification at a given ISO setting in order to produce the 18% figure. And buyers howled at how dark the pictures were that it produced. The earlier and later models in the Rebel line produce considerably brighter pictures.

The bottom line:

The camera manufacturer decides how bright auto-exposed images should be, and "18%" doesn't figure into it at all (with the rare exception of SOS ISO ratings).

Some references:

A layman's introduction from Thom Hogan:
http://bythom.com/graycards.htm

A bunch of gory detail, unfortunately written prior to the current ISO sensitivity standards:
http://dougkerr.net/pumpkin/articles/Scene_Reflectance.pdf

The DC-004 specification on which the current ISO sensitivity standards are based:
http://www.cipa.jp/english/hyoujunka/kikaku/pdf/DC-004_EN.pdf

 

[Leonard Migliore]

So I would, in general, not be afraid of underexposing, at least if I were shooting RAW

If I were shooting RAW, I would intentionally overexpose to retain shadow detail.

http://www.digitalphotopro.com/technique/camera-technique/exposing-for-raw.html

It really depends on the scene. If your highlights have interesting details, you would want to underexpose in order to prevent blowing them out. If you want shadow detail and don't care if highlights blow out a bit, you overexpose.

I define "overexposure" as that which causes highlights (at least, highlights you were interested in) to blow out. Once an area goes to its maximum level in any channel, you lose that detail permanently (like in a lot of pictures of red flowers). But shadows don't go black as quickly. So if you want to record both highlight and shadow detail, you have to limit your exposure to something that retains data in the brightest areas. Then you can bring up the shadows quite freely with something like Lightroom's Fill Light. The limit on that is when the noise starts to become objectionable.
--
Leonard Migliore

 

[Graystar]

The following is off-topic, but this is a forum for beginners, and there's no point in giving them bad information.

In a nutshell, your camera ( any camera) wants to meter the scene to return a value of 18% gray.

This is one of the great myths of photography

My Nikon D90 uses the REI method for ISO (as stated in the specs.) A picture of a gray card (or any neutral surface,) when converted with no processing into sRGB, gives RGB values of 100, 100, 100 (12.7% gray) for Spot and CW metering, and 119, 119, 119 (18% gray) for Matrix metering. So my Nikon see both 12.7% gray and 18% gray...depending on the metering mode.

Nikon's default processing tends to increase the brightness on JPEGs out of the camera, as well as RAW images processed with the bundled processor (ViewNX) are brighter (which explains why some people keep -1/3 EV dialed into their Nikons all the time.) But images process with tools such as Raw Therapee don't do that and the image tones are more accurate.

.

 

[John Deerfield]

It really depends on the scene. If your highlights have interesting details, you would want to underexpose in order to prevent blowing them out. If you want shadow detail and don't care if highlights blow out a bit, you overexpose.

I define "overexposure" as that which causes highlights (at least, highlights you were interested in) to blow out. Once an area goes to its maximum level in any channel, you lose that detail permanently (like in a lot of pictures of red flowers). But shadows don't go black as quickly. So if you want to record both highlight and shadow detail, you have to limit your exposure to something that retains data in the brightest areas. Then you can bring up the shadows quite freely with something like Lightroom's Fill Light. The limit on that is when the noise starts to become objectionable.

I might suggest reading the article. I wouldn't suggest overexposing the highlights to the point that all detail is lost in the Raw file. But the histogram is from the processed file and that information has had a curve applied to it. The Raw data is linear information. It is possible to "overexpose" the Raw file, the histogram will still be that of an overexposed JPEG, but the linear data is NOT overexposed. As to how much information exists within the file (and the highlights) will vary from camera to camera. And the article suggests testing your camera. But in any event, I would rather overexpose a Raw file and underexpose a JPEG. I overexpose the Raw file to bring out the best in the shadows while still retaining detail in the highlights. I underexpose a JPEG (or rather try to nail exposure) to protect the highlights in the JPEG.

 

[Leonard Migliore]

It really depends on the scene. If your highlights have interesting details, you would want to underexpose in order to prevent blowing them out. If you want shadow detail and don't care if highlights blow out a bit, you overexpose.

I define "overexposure" as that which causes highlights (at least, highlights you were interested in) to blow out. Once an area goes to its maximum level in any channel, you lose that detail permanently (like in a lot of pictures of red flowers). But shadows don't go black as quickly. So if you want to record both highlight and shadow detail, you have to limit your exposure to something that retains data in the brightest areas. Then you can bring up the shadows quite freely with something like Lightroom's Fill Light. The limit on that is when the noise starts to become objectionable.

I might suggest reading the article. I wouldn't suggest overexposing the highlights to the point that all detail is lost in the Raw file. But the histogram is from the processed file and that information has had a curve applied to it. The Raw data is linear information. It is possible to "overexpose" the Raw file, the histogram will still be that of an overexposed JPEG, but the linear data is NOT overexposed. As to how much information exists within the file (and the highlights) will vary from camera to camera. And the article suggests testing your camera. But in any event, I would rather overexpose a Raw file and underexpose a JPEG. I overexpose the Raw file to bring out the best in the shadows while still retaining detail in the highlights. I underexpose a JPEG (or rather try to nail exposure) to protect the highlights in the JPEG.

I think we're all trying to do the same thing (get the most highlight and shadow detail) and are differing in terminology. I don't consider ETTR overexposure but perhaps you do; in any case, I bet we both end up with similar exposures.
--
Leonard Migliore

 

[John Deerfield]

The bottom line:

The camera manufacturer decides how bright auto-exposed images should be, and "18%" doesn't figure into it at all (with the rare exception of SOS ISO ratings).

Bottom line: the camera can measure the EV value incorrectly. Whether using the 18% gray, the 12% gray, or some other camera specific measurement. The whole idea is to measure the light intensity within the frame within the scope of the metering system. And that such a reflective reading can be incorrect . And hence you use exposure compensation to make an adjustment.

For the record, I have found that in my classes, when I use an 18% gray target, a great majority of the cameras give me a spike dead center in the histogram. Perhaps not all . If memory serves, the Nikon D70 comes in under 18% gray. In any case, using the 18% gray as a teaching reference works as well as any other method I am aware off. It has been awhile since I read Thom's article, but I believe he even points out that there isn't any way of knowing the exact calibration of every camera. Without that knowledge, all of our discussions are academic. And the 18% gray reference works as well as any other.

 

[Graystar]

For the record, I have found that in my classes, when I use an 18% gray target, a great majority of the cameras give me a spike dead center in the histogram.

Aside from the fact that the histogram's center is not 18% gray, any uniform surface is going to produce a spike in the same exact location, regardless of how dark or bright it really is.

.

 

[John Deerfield]

Aside from the fact that the histogram's center is not 18% gray,

Really?

Here is what I know. I take a picture of my 18% gray target ( http://www.bhphotovideo.com/c/product/671141-REG/Cameron_1203261_32_Reference_18_Gray.html ). Is it really 18% reflective gray? I don't know, I just trust the manufacturer. When I do this under even lighting, I get a spike right dead center in my histogram. When I bring that image into Photoshop, the RGB values are 128, 128, 128 (or near enough). All leading me to conclude I have "middle" gray (values ranging from 0-255, making 128 the middle. Or maybe you would rather simply create a layer in PS and have PS fill it with 50% gray. In which case you get a solid layer of gray measuring 128, 128, 128 RGB no matter where you put the eyedropper... and this layer will also have a single line right in the middle of the levels.... hhhmmmmm) and since it was labeled 18% gray, I assume (uh-oh) that 18% gray is 128, 128, 128 RGB. I can do this with every camera I own (I currently have five). So every camera gives me a spike, right in the middle, from what is called an 18% gray target. The RGB of this image measures 128, 128, 128. Not only can I do it with every camera I own. I can repeat it with almost every camera I have had in class (again, there are exceptions as it has been pointed out that not every camera meters for exactly 18% gray... but we can't know this without either a target or information from the engineers). Now if you want to call the middle of my histogram something else, go for it.

Seriously, how does this help the discussion at hand? EV compensation exists to compensate for metering inaccuracies regardless of how the meter is calibrated to measure the existing EV. By using the 18% gray as a starting point, we can begin to understand the issue. Does the OP's camera meter for 18% gray? Who knows? Better yet, what possible difference can it make? Use a reflective meter in the camera, taking a picture of a gray target should produce a "correct" exposure. Taking one of a black target should produce an overexposed image and a white target an underexposed image.