Do you adjust your camera lcd brightness?

I realized the other night that images viewed on the camera seem to be properly

exposed, but when on the computer a bit dark. I believe my monitor is properly

calibrated and was wondering if my camera lcd is just too bright when shooting

at night. Do you adjust brightness on your lcd?

Yup. The camera LCD is just for convenience, not for evaluating exposure accuracy. I adjust it as necessary to suit the ambient light level.

Great question. Until recently, I had the LCD up fairly bright on my D200. I realized my shots were nothing like what I saw on the LCD so I turned it down for daylight. My exposures are much more accurate.

I think every nikon DSLR I have owned (not 100% sure about the D70 and D100 but the D2H and D2X for sure) have had the LCD brightness turned down substantially to be more accurate. You sacrifice bright sun viewing, but thats not usually when I am shooting.

Never evaluate exposure based on how it looks on the LCD. Use the histogram and if the green channel isn't close to the right than it's underexposed.

Yes. Always as low as possible out of personal preference. I have also always assumed that this makes my batteries last longer, ad it certainly seems to (as does shooting jpeg, turning off review and not chimping). You can't judge anything but composition/focus off of the preview, as there are too many variables, and they are processed images. Check the histogram. It is your equivalent of an in-camera densitometer (albeit a rather rough one).

 

Keith

Yes, I had to adjust the brightness to -2 on my D40 for the underexposed images to look underexposed.

Yes, I keep it low. Joe Smith

I figured it was too bright. The real issue is understanding my histogram better, thanks everyone.

since it seems this issue is done I will hijack this thread.

How DO you read the histogram properly. I never really knew...maybe I should...

Using a D80 and D40

"How DO you read the histogram properly."

 

> That depends on whether you are shooting RAW or JPEG.

 

For JPEG you basically want to get a good spread from top to bottom. Accomplishing this means paying attention to your Image Optimization settings such as contrast and saturation -- since you ask about the D40 and D80 we'll skip Active D-lighting.

 

For RAW you want to put the histogram as far to the right as possible. With the D40 and D80 this means using low contrast and saturation and watching the green channel (except under tungsten lighting, when the red channel can be problematic, but don't confuse this with other lighting when the red channel clipping is caused by WB gain and is the most common reason people end up underexposing their images).

The histogram you want depends on the print you want. The closer your exposure is to what you envision in the print, the less you have to alter the file, and thus the better the quality and the easier the editing.

 

If you don't know what you want your final print to look like, the rule of thumb is to expose the histogram as far to the right as you can without clipping anything (except maybe some speculars that will look fine as pure white in the print).

 

If you have a camera that allows you to look at the histogram for each channel, double check the red channel as well, especially when shooting in tungsten light. It usually blows out before the others.

 

Keith

"If you have a camera that allows you to look at the histogram for each channel, double check the red channel as well, especially when shooting in tungsten light."

 

> You are half right. In other than tungsten lighting you should ignore the red channel if you are shooting RAW.

...I do not shoot raw anymore. No time, no space, and I do not shoot digital for that which I intend to be high-quality personal work. What I do is shoot digital when I want something quick, or something that I know I will likely never print, but only share with people digitally. If I want to fiddle around and take a lot of time and be perfect, I shoot film.

 

Could you explain why to ignore the red channel when shooting raw? Even when I have shot raw, sometimes the red channel blows out a little even on a perfect looking combined histogram. It looks bad and cannot be recovered.

 

Keith

I don't mess with my lcd settings at all. All I check for is my histograms. I check my channels & for blinkies - - - lcd brightness matters not to me. But I do care that I can see what's on the lcd.

 

JMHO

 

Lil

Hi Keith,

 

You asked about the red channel and shooting in RAW. It is possible for the red channel very infrequently to clip before any other channel does, but aside from tungsten lighting the only way for that to happen is if you are shooting a very saturated red with no other competing highlights. Of course, I am referring to the actual RAW data without any WB gain applied. The same can happen to the blue channel if you are shooting a very saturated blue (and this does happen occasionally on sunny days shooting a bright blue sky). I cannot find an image of mine right now where the red channel clipped before the green channel, but I often see heavily saturated reds going out of gamut for commercial printers that use sRGB colorspace. Perhaps you have an image you could share with me because I'm sincerely interested in seeing it.

 

Below is a section of a colorchecker chart shot at +1.3 stops of EC using flash and center weighted metering; which is when the green channel clipped in the yellow patch. The insert is a close-up shot of the red patch alone taken at +2.7 stops of EC and as you can see the color is completely recoverable. I test using flash at various distances, and the earliest I could get the red patch to clip in the red channel was +2.3 stops of EC.<div></div>

Happens sometimes with flash and on certain folks' skin...and with the bright red objects you mentioned. Yes; it is entirely dependent on the color of the object and/or the color of the light hitting the object. Most often on "white" people and in warm afternoon light, but worthy of mentioning given my experiences. It doesn't happen often, but when it does happen, the combined histogram won't let you know you are clipping; only the histogram for the individual channel (be it red, green, or blue) will. Just yesterday, on two separate shots, I got a review image with no "blinkies" and a histogram that looked just right. Upon review, certain areas of red were rendered as 255 red in the computer. I don't think recovery looks right, or that it is ethically sound for many of the type of pictures I take...not all, just a lot of them. Thanks for sharing that test. I can prep some pix if you want, but I know why the channel blew out on the ones that did.

 

Keith

"Happens sometimes with flash and on certain folks' skin..."

 

> I would like to see the RAW file. I am close to 100% certain that the data is not clipped.

 

"I don't think recovery looks right, or that it is ethically sound for many of the type of pictures I take...not all, just a lot of them."

 

> The only reason I can imagine that recovery would be wrong is if the green channel also was clipped. If you have saturation turned up on your camera than the embedded JPEG could be hiding that.

 

As for ethics, I find that consideration as it regards optimizing exposure baffling. Why is it ethical to underexpose an image? Is it unethical to use a different exposure speed, aperture or ISO?

It is unethical to ADD information to a file in the world of journalism. You're not "optimizing exposure" in any computer. The only place to do that is in the camera. Besides, optimum exposure is an aesthetic choice. It sounds a little crazy to your everyday photographer, but it is very theoretically important in journalism, and different than burning in, in which you are using what is there instead of simulation.

 

I know how to read a histogram. The data is clipped.

 

Let's just forget about it. We are coming from different places and won't agree no matter how much we write about it.

 

Thanks for the help.

 

Keith

"It is unethical to ADD information to a file in the world of journalism."

 

> It would be no more unethical to adjust your exposure then it would be to add contrast, resize, crop or sharpen for example. Your ethical rules are peculiar if you are saying that EC is outside of what is ethical.

 

"Besides, optimum exposure is an aesthetic choice."

 

>Not really. Determining where the highlights and shadows fall is an aesthetic choice, but there is only one optimum EV, which is to saturate the photosites as fully as possible without clipping any important data.

 

"I know how to read a histogram. The data is clipped."

 

> I'm going to continue to dispute that. Do you know how much gain is typically added to the red channel? Often your camera's settings distort the data that is coming to the camera's histogram from the embedded JPEG, that is a near certainty in your case. Finally, I am not sure how you are handling the files so that you cannot adjust EC to recover highlights -- perhaps it is a problem with the converter you are using.

 

"Let's just forget about it. We are coming from different places and won't agree no matter how much we write about it."

 

> Considering that you do not want to be bothered by shooting RAW, then I agree that you should not be using ETTR techniques, in that regard I agree that we are coming from different places. My main issue here is that telling others to avoid clipping the red channel when shooting RAW will often lead to underexposure and that channel is rarely indicative of how saturated the red channel photosites have become.

I am talking about histograms on the computer, and the red channel is, indeed clipped, when the in-camera combined histogram was fine. I don't have separate histograms on any of my cameras...which is the problem. I shoot jpegs totally flat.

 

As for the "ethics" issue. I am not suggesting that anyone else hold themselves to this except for journalists. I sure don't in the personal work that I consider to be "non-journalism". All I am saying is that there is a fundamental difference between "recovery" and an exposure adjustment. Some recovery programs may not actually add information to the file, but others do. It is not a visual issue, but a purely theoretical one, and I agree that it is a little silly from the standpoint of an "artist" (or whatever you might call it). The idea is that recovery of highlights by adding information to the file that was not captured by the exposure if the same things as adding a tree or pasting someone's face onto someone else. Ethical considerations were far less stringent in the film days. As an example, people used to flip negatives all the time. Now, with digital, manipulation is so easy and rampant that ethical standards have been tightened. People born and bred on digital seem to think that "can" equals "should". For the same reason, journalists do not edit images on layers. For one thing, it's simply practical; no time. For another, editing on layers in the equivalent of combining multiple images. All this is a conceptual ethical argument; not arguing that it is the best way to make the best pix.

 

The idea that there is an optimum exposure is a good idea in general circumstances when you don't know what you want your final product to look like. The optimum exposure of which you speak gives you the most options and versatility later on. However, I prefer to commit to my final image in camera as much as possible, instead of playing it safe. If I want a dark print, I'll shoot a dark exposure. If I want a high contrast, "pushed" look, I'll underexpose like I did with film, and "overdevelop" to bring the high end up.

 

As for being "bothered" to shoot raw...I understand the benefits of shooting raw, and I don't pretend to try and preach that everyone shoot jpegs, or even that they are superior overall. The choice between the two really depends on what you are going to do with the files, IMO. Journalistsic editors aren't going to sit there editing a raw, doing layers, etc. If the pic isn't FTPd to the people who want it ASAP, you're done. They've bought it from someone else. Many magazine journalists shoot raw, but not wire service or newspaper folks.

 

Keith

"I am talking about histograms on the computer, and the red channel is, indeed clipped, when the in-camera combined histogram was fine. I don't have separate histograms on any of my cameras...which is the problem."

 

> Okay, under artificial light like incandescent for instance, the red channel can be problematic, but usually the green channel is very close though. The combined histogram is your problem, it is neither tied to the red nor the green channel, it just measures the average luminance. As for flash, that is more balanced light (about the same as typical daylight) and the red channel practically never clips before the green channel under that lighting.

 

Sorry for the miscommunication on my part, I don't know what camera you are using and presumed one of the newer models.

 

Regarding ethics, well maybe you have an editor who is a bit peculiar. Adjusting exposure compensation adds no data to the image, that's like saying demosaicing RAW data adds data because the pixels originally only had one color each.

 

Of course, if you are time constrained and have to deliver a JPEG, expose the image the way you want it to come out.

Perhaps we have different ideas of what the word "recovery" means. I think what you are describing as recovery is raw exposure compensation/adjustment. All that does is darken the clipped areas. It looks ugly past a point, but doesn't add anything that wasn't there. (It adds "digital things", meaning it alters the information that the pixels hold, but not "actual things", meaning the creation of visual detail where there was nothing but 255 equivalent luminance.) This isn't what I'm talking about in regards to ethical considerations. That is not "true" recover; only the lowering of values. What I am describing as recovery is what you get from a specific "recovery program". They do indeed "look around" at nearby pixels and make up what they think would have been in the blown out areas, as opposed to just darkening what is already there.

 

I thought I had made clear that I knew the problem was coming from the fact that the combined histogram was averaging the three channels, so even if one of them slightly clipping, the in-camera combined histogram would look fine.

 

"Even when I have shot raw, sometimes the red channel blows out a little even on a perfect looking combined histogram."

 

"It doesn't happen often, but when it does happen, the combined histogram won't let you know you are clipping; only the histogram for the individual channel (be it red, green, or blue) will."

 

Funny thing is, in my experience, on a combined histogram that is on the upper edge, but is not clipped, I tend to notice the reds going first once I get it into the computer, but in your experience, you notice the greens going first.

 

At any rate, I am simply training myself to watch out for highly pure colors, and when I see them, to give a little less exposure than I think is right.

 

...and saving up for a more modern camera!

 

Keith

"I think what you are describing as recovery is raw exposure compensation/adjustment. All that does is darken the clipped areas."

 

> I think we have both been wrong. I should not have been writing "recovery" as I meant exposure compensation/adjustment. However, you are mistaken when you write that the RAW data in the red channel is often clipped before the green channel is. I am not talking about recovering an overexposed image, I am talking about exposing an image properly and then adjusting it to where you want the tones to fall prior to conversion from RAW. You continue to suggest that the red channel is often clipped, it is not and the clipping you are observing is almost always a consequence of WB gain (amplification) being applied to the RAW data.

 

"Funny thing is, in my experience, on a combined histogram that is on the upper edge, but is not clipped, I tend to notice the reds going first once I get it into the computer, but in your experience, you notice the greens going first."

 

> I often use uni-WB on my D200, and I can tell you from looking at thousands of images (perhaps tens of thousand of images) that the green channel never requires any amplification, so when it is clipped it is clipped; but the red channel always requires amplification, so what looks like clipping (even on your computer) is more often the result of that amplification. Using ACR and Nikon's software you get no indication of this as they both make presumptions about the red channel amplification and apply that before you even see it, then you end up adjusting data that has already been adjusted. Capture One and Raw Magick Lite appear to allow WB to be reset, and using a gray reference point you can adjust NX or Nikon Capture to get the right WB -- ACR seems hopeless but I could be wrong about that and may be confusing poor color profiling with poor implementation of WB.

Thanks. But in practice, how do you keep it from happening (the WB gain from excessively boosting one channel)? You need that WB to be applied, right? So it is just part of the equation you have to deal with. Maybe with AWB, I can see what you are describing happening. But in my experience (just as many pix as you; I shot over 20,000 digital pix on assignments - meaning non-personal work - last year, and personally edited most of them) Shooting jpeg (80 to 90 percent of the time) or raw, I use three preset white balance settings (daylight, tungsten, and sometimes fluorescent, when I don't have the opportunity to shoot a grey card in the fluorescents, which cycle through a variety of color temps that I prefer to average with a long exposure on a grey card. I shoot grey cards in foul mixed lighting (usually fluorescents in basketball/volleyball arenas) whenever I have the time.

 

What I mean is, obviously I WANT the WB to be applied, so whether the channel is being clipped due to exposure or due to WB gain, there is no practical difference: it is still clipped.

 

I don't use AWB because I want a constant white balance of my choosing (just like with film); not shot-to-shot variation. This seems to make color balancing insanely more easy for me; especially when I am selecting and prepping on average 20 pix from each event, including captions, keywords, etc., within the space of a few hours (preferably less, but I'm not the fastest machine around).

 

But, yes. I can entirely see the problem going away with the use of AWB, which I imagine matches the three channels to each other.

 

Keith