MtnMan Loc: ID

Excellent summary.

The only thing I quibble with is the parenthetical included in #4. If you left that out I'm in 100% agreement with points 1-4.

There is simply no way your camera white balance setting can affect what the sensor records: the RAW image. None, Nada.

White balance and picture controls are only used by your camera's computer when making a jpeg image from the RAW file.

The only possible exception, and I hesitate to mention it, is if you use image review or live view and change your exposure because of it. As you correctly note white balance and picture controls do affect the jpeg LCD image.

I don't know about the last two points.

Capn_Dave

Bottom line your camera exposure settings do matter. The key word EXPOSURE. The Focus does matter. These 2 things effect the sensor and how much light it gets and how sharp the image is. That is what is limited to about + - 3 stops in processing. Everything else is just what the little computer adds. Believe it or not Photoshop Lightroom Gimp can tweak the shot to bring it to what your minds eye saw. These can't make up for poor composition, out of focus, piss poor light.
All this stuff combined can make the difference between a snapshot and a great shot

MtnMan Loc: ID

I did a little research on this. The only point that I find agreement on is that you want a properly exposed image.

Often the choice of metering mode is most important. Most cameras have at least three choices for metering (not to be confused with the choices for autofocus which might look similar): point, center weighted, and matrix. If using the first two a slight movement of the composition can greatly affect metering.

Bryan Peterson, the guru of exposure, opines, "If you are within 1/3 or even 2/3 of a stop, your exposure will in all likelihood turn out just fine...as long as you set the exposure on the minus side...you want to avoid 'blown highlights'".

So he is on the underexpose team.

The slightly overexpose team argues that there is more information in the top quarter of the histogram than the lower 2/3. An interesting observation but I don't get what it has to do with your exposure setting.

Generally you do get a broader dynamic range with RAW images because the image has more bit depth. Most jpeg images are 8 bit. Most RAW images 14 or more. That means you might have more leeway for correcting afterwards. But you can't correct for blown-out or blocked-up areas of the image...there is simply no information there.

naturepics43 Loc: Hocking Co. Ohio - USA

Thanks for the reply and researching this subject. I too have done a lot of research and reading. It seems that the sources I've found lean toward ETTR. I'm not saying it's the right way, just what I've found. Since shooting conditions vary so greatly, I'm thinking that experience and trial & error will be the best teacher. AND shoot in RAW so hopefully I can rescue a slight error in exposure.

MtnMan Loc: ID

Yes, that is a good strategy.

I have to confess that there are few times I have had an issue with exposure where it would have been recoverable had I shot it in RAW. It usually happens when I am using spot metering and get the spot in the wrong place. It is so far off there is no hope.

I shoot in jpeg most of the time. I save using RAW for images I think will be really important. My reason is that the RAW file sizes are ginormous with my D800. I'd rather use the file space for more images; e.g. bracketing or shooting in continuous for moving birds, etc.

So here is a recent bird since this is a photography site..

Capn_Dave

Great capture Mtnman. you caught the moment. birds in flight are particularly hard

naturepics43 Loc: Hocking Co. Ohio - USA

Well done!
:thumbup: :thumbup: :thumbup:

My biggest problems with exposure and excessive noise come from BIF shots in a bright blue sky.

MtnMan Loc: ID

Thanks. Wally was easy because he lives at the National Center for Birds of Prey and they were flying him from one person to another. He works for food. I knew right where to stand and even had multiple tries at it.

Getting wild birds like that is much more challenging.

Capn_Dave

You got that right. It's still a great capture. I doesn't look unnatural

Dave

Bram boy Loc: Vancouver Island B.C. Canada

such a hero

Bram boy Loc: Vancouver Island B.C. Canada

mt man you have that right . I have taken all kinds of shots , with the white bal. set all over the map deliberately to see what happens , in raw mode , Nada , none .

Bram boy Loc: Vancouver Island B.C. Canada

well I'm 72. this has been my hobby since I turned 25. and your getting to
wrapped up in the in the settings of the camera and such. the two cameras I have now are four years old . and there is 15000 actuations on each one . and I never take the white ball. out of auto. except when I'm checking the differences in me picking it for a photo or auto picking it . or doing a white bal setting. and it does no good getting it all right in camera , what ever that means . that saying is as bad as ( nice capture ) . any way if your composition is lousy . and the picture is not interesting enough to hold your attention for a minute . then no post proessing on earth is going to save it . it's just more to delete . and try again .

IJC1962 Loc: Motown

I shoot with a D800/D200/D70 (converted for IR) and an F100 for film. I have always believed that you should take the best possible image at the time and avoid cropping and post processing otherwise it is not fun.
This is just an opinion but it does help one learn the skill of composing and shooting
IJC

Gauss Loc: Earth

Over/under exposure and the amount of data in the historgram...

Here's my understanding. The histograms in-camera and in most software are based on 8 bits or 256 levels. Underexposed images have fewer levels of luminance/brightness, thus require fewer bits to represent and therefore contain less data. A histogram for such an image would be shifted to the left with little or no bars showing in the higher values. Say you underexpose so much that the highest value on the histogram is 64. Such an image only requires 6 bits and conveys only 1/4 the information of an image properly exposed to make full use of the histogram (64/256 = 1/4). And this is why it's said that the right side of the histogram contains more information; as you move to the right, using more bits/levels, it's a doubling trend in terms of the amount of information. This directly relates to your exposure setting.

Obviously, too much under or overexposure loses data. So, underexposing to save highlights must be done with delicate touch due to the halving rate of loss of data with increasing underexposure.

Regarding RAW, overexposure, and the recovery of highlight...

Today, most cameras are 12 or 14 bits. This means the sensor can capture a lot more information than can be expressed in a jpeg using 8 bits. Fourteen bits can distinguish 16,384 levels of luminance; that's 64 times what the jpeg can contain. Pure white is represented by the highest value a given number of bits can represent, 16,383 and 255 for 14 and 8 bits respectively. Thus, in converting the RAW data to jpeg, the top 64 values of the 14 bit data are set to 255. Yet 63 of those values are not pure white. In addition, any in-camera setting like vibrancy and saturation may push additional lower values up to the white level. Thus, it's common for a jpeg image to have blown highlights even though at the sensor was not saturated, i.e. the RAW, 14 bit image did not have blown highlights. But remember, per what I said earlier, the in-camera histogram is based on 8 bits as is the LCD display. So what appears overexposed there, may not be in the RAW data. This is folks are talking about when they suggest slight overexposure for digital using RAW. They're referring to what you see on the histogram, not that you truly overexpose the sensor. The software can expand those higher values of the 14 bit image so they fall into values below 255 in the jpeg, recovering some highlight detail. From this perspective, you would not want to underexpose digital images based on the in-camera histogram or what you see on the LCD because you'd be needlessly losing information. And indeed, it takes practice to correctly judge when an image is really overexposed versus just the jpeg.