Monitor Calibration and Colour Space - Particularly for those from the film world.

Alan Edward Klein said:

How does the brightness level of the monitor affect the final image that goes out as a file to the web? It doesn't change it.

Click to expand...

If your brightness level is too high on your monitor, when you look at a file and go to tweak it you will adjust the brightness level in the file lower.

When the file goes out to the web, it will have a built in lower level of brightness.
So if someone opens up the file on their computer, with a monitor brightness set to a more standard level, the image will look quite dark.

That is what calibration is about - adjusting the response of the monitor (and video system) to look "normal", when the file is "normal".

MattKing said:

If your brightness level is too high on your monitor, when you look at a file and go to tweak it you will adjust the brightness level in the file lower.

When the file goes out to the web, it will have a built in lower level of brightness.
So if someone opens up the file on their computer, with a monitor brightness set to a more standard level, the image will look quite dark.

That is what calibration is about - adjusting the response of the monitor (and video system) to look "normal", when the file is "normal".

Click to expand...

I understand that. That's why I said I use the histogram. Otherwise, if you make the CD level too low, you'll over-expose it and clip the whites.

But the results are what count. Did you look at my links?

Alan Edward Klein said:

But the results are what count. Did you look at my links?

Click to expand...

They look fine on my uncalibrated laptop monitor.
That may or may not be a good thing!
And as for the histogram, while it may tell you about clipping, it says nothing about the apparent brightness when something is actually viewed!

Alan Edward Klein said:

I am able to see the black and whites in that test although two or three of the darkest blacks seem very close. If my screne was darker, more of the black would blend. It is bright in my room right now.

My setting for luminance is 217Cd/m2. I didn't set the highest selection. I just left everything at default when I calibrated the monitor. Why does it matter? How does the brightness level of the monitor affect the final image that goes out as a file to the web? It doesn't change it.

Click to expand...

You are correct to say changing the brightness level of your monitor does not have any direct effect on the files. But (as others have said) if your monitor is set to a too-high luminace level, and if you adjust your images to "look right" on the high luminace monitor - say, maybe you reduce the exposure or dial down the highlights - then, when those files are viewed by others who are not viewing on boosted brightness monitors, your images may appear too dim.

Alan Edward Klein said:

I use the histogram to set levels to avoid clipping. So luminance should not affect that. Of course contrast, and some of the other sliders would be affected. I guess looking at the results would be the best way to tell if I'm getting good results. So maybe you can help me there.

Here are samples adjusted at this level. Do they look OK to you? These were shot as jpegs with a digital camera.

Red Rock Canyon, Las Vegas

Explore this photo album by Alan Klein on Flickr!

www.flickr.com

Here are samples of 4x5 Velvia 50 film scanned. A lot more adjustments had to be done so these are more critical. I used the same Cd luminance level as for the digitals

4x5 Color Large Format

Explore this photo album by Alan Klein on Flickr!

www.flickr.com

For both sets especially the film scans,
1. Does the brightness seem OK?
2. Contrast OK?
3. Shadows and highlight areas OK?
4. Do colors look normal? (they may look a little saturated as I saturated the digitals a little and the second set was Velvia 50s with high saturation in the film.)

Click to expand...

Looking at your images, they look fine on my calibrated iMac. Except for the top images in each set, which I assume have been purposefully darkened to make your title text more visible, right?

The first image after the title shot in the Red Rock series (you standing in front of the rock), the contrast is a little flat, but that is probably due to the strong backlight causing a little flair (glare?) in your lens - not a processing issue. Otherwise, colors, contrast, ect. all look good to me - so whatever you are doing it must be approximately correct.

For clarity, I've never thought that your shared files were likely to be way off, although that is always possible. Most of this thread is more about understanding where one can rely on certain shared information, and where reliance may be mis-placed.
That, and how and when one can meaningfully provide useful examples about things like colour crossover.

runswithsizzers said:

You are correct to say changing the brightness level of your monitor does not have any direct effect on the files. But (as others have said) if your monitor is set to a too-high luminace level, and if you adjust your images to "look right" on the high luminace monitor - say, maybe you reduce the exposure or dial down the highlights - then, when those files are viewed by others who are not viewing on boosted brightness monitors, your images may appear too dim.

Looking at your images, they look fine on my calibrated iMac. Except for the top images in each set, which I assume have been purposefully darkened to make your title text more visible, right?

The first image after the title shot in the Red Rock series (you standing in front of the rock), the contrast is a little flat, but that is probably due to the strong backlight causing a little flair (glare?) in your lens - not a processing issue. Otherwise, colors, contrast, ect. all look good to me - so whatever you are doing it must be approximately correct.

Click to expand...

I'm not seeing what you're seeing regarding the darker ones on top with the text. You should be able to click on the first picture in each group. It will open full size. Then click repetitively on the arrow on the right to go through the rest of the shots in full size. Try that. Do the pictures look better? (Note I don;t see flare in the second shot. I do see clouds on the left side of the sky however in a semicircular look.)

The fact they look good generally is important to me. At least I'm on the right track.

I do know how to navigate Flickr, but maybe I don't use it enough to understand it's visual language. What is the "first picture in each group"? The very top photo, the big one, with your white Red Rock title text on it? Or the next one after that (you in front of the rock)? At first, I thought your title shot was part of the group of photos - and if I clicked on it, it might, clear up, or launch the filmstrip, or whatever Flickr calls it. Something about the darkened photo made me think it was clickable, but I see now, it is more of a banner or a graphic for your title and the photo was meant to stay dark in the background. The straight shot of that scene is actually the last image in the group.

I was hesitant to use the word "flair" but I don't know what else to call it. Look in the upper left-hand corner of the rock you are standing in front of. Look at the park logo on the rock - the black logo is not black. The face of the rock looks a bit washed out - like there is lot of dusty atmosphere between you and the rock. Was it very dusty / foggy? If not, I think the low contrast is due to the bright backlight causing some minor reflections inside your lens.

The first picture is the one with me standing in front of the rock. The Flickr darkened photo at the top with the name of the folder is the banner. It uses one of the actual shots. So apparently we were discussing the wrong pictures. You're right that there's reflection off the rock and my head and also washing out the black lettering. Nothing I would bother trying to correct when I edited. The sun appears very bright coming from the left accounting for that "glare". I handed my camera to a stranger nearby and asked her to shoot me. But the girl was young and cute and I don't want to comment on her photographic skills.

Thanks for reviewing the folder. The previous comments that the colors looked pretty acceptable make me think the settings for my monitor are OK even with the very high CD for luminance.

I've been reading about color science and want to check my understanding with folks here, so here's how I understand it;

Since our eyes can only see color with three different types of cone cells, we only need three numbers to describe colors. To take that even further, imagine a universe where our ears could only hear three frequencies, then any piano would only need three keys and any sheet music would only need to describe how soft or hard to play each of those keys relative to each other. Then it becomes a matter of how hard to you actually hit the keys relative to each other and exactly what 3 tones you tune each of your keys to.

Say that sheet music used a scale of 1-10 on how hard you should hit the piano key. You would have to know how hard to actually hit the key if it said "5", or "2" or any other number. Well the sheet music would have to come with some sort of note on the front that said "A 10 on the scale refers to hitting the key such that you measure 10 decibels from 1 foot away". It would also have to say "Note A should be tuned to 440hz". With those two pieces of information, you could get a loudness and pitch meter and you could calibrate your finger and the piano so that you were producing the music the same way the composer intended. Now replace the sheet music with the image file, and the notes about loudness and frequency with a color space, then you have color science.

Of course there are more details like gamma which means that the loudness get louder as you go up the scale (a 1 to a 2 are closer together than a 9 to a 10), and you can't actually change the spectrum of the channels in your monitor to "tune" them like you could a piano, so the color profile has to potentially mix a little bit of red or green in when its asked to show pure red.

So then the difference between sRGB and AdobeRGB would be the exact color of light in terms of wavelength for each of the three primary colors, and the intensity of the light when the pixel asks for a maximum value.

So choosing a color space for your image will be like deciding which information to put on the front page of your sheet music (what tuning and loudness the player should use), and calibrating your monitor is like actually pulling out your loudness and pitch meter and making sure your piano actually follows those standards.

A color space is the agreed upon standard, a color profile is a file with data that tells the computer how to convert any image data encoded in one specific color space into actual signals to the monitor so that that specific monitor displays the pixels correctly according to the color space standard the image data was encoded with.

Please tear me a new one if this isn't a good analogy but it's how I currently think about color space

jonmon6691 said:

I've been reading about color science...

Click to expand...

Hi, I think their are significant differences that rock the boat, so to speak, regarding the analogy with sound. The ear, for example, CAN distinguish between different frequencies. Whereas the human eye cannot. It's true that the eye has three color-sensing functions, but there can be multiple makeups of frequencies that appear identical to the human eye.

If you wanted to use the ear analogy, you would ideally find someone with three ears, each ear being specialized to a frequency range. Say one ear largely sensitive to low frequencies, yet not able to distinguish between them. It would only detect the relative strength of the combined lowish frequencies. Another ear is mostly sensitive to middle frequencies, but may have significant overlap into the lower frequency range. Then another ear for higher frequencies. This hypothetical person would hear sound as a relative strength comparison between the three "hearing functions." Yet still not able to distinguish between specific frequencies. So I would just say that actual human hearing is quite different from color vision.

Something that you might find interesting to read is the "Colorshop Color Primer" by Fred Bunting (I wish he had written more). It was included with an earlier, late 1990s?, hand-held instrument called the Digital Swatchbook. It looks like the primer can still be found online.

A key thing in modern color science is the so-called color-matching experiment (p.17). (Look up Wright and Guild for more detailed info.) This allows construction of the so-called color-matching functions that everything is based on.

Hope you find it readable. Best of luck.