Adobe Gamma and similar - how accurate?

<p>In my day-job I design training, and I'm working on a photography-related piece right now. I'm trying to determine the best way of controlling what a learner sees on his/her computer screen regarding the brightness of an image.<br>

Overall, it seems as though the only practical way to address this (within the context of a course, that is) is to have the learners adjust their monitors using something like Adobe Gamma, either the real Gamma or some in-house, home-grown version thereof.<br>

My question is, how accurate is gamma? Will it hit within, say, half a stop across the learner population, assuming that they perform the process properly? Are there other, similar solutions (i.e., solutions which don't require anything more than the monitor and the user) which perform better?</p>

<p>Adjusting the gamma with tools like that should get the correct shade of gray to be halfway between the brightness of black and white. The match can be as good as the user’s vision is capable of distinguishing, though I do not know how many stops that is likely to be.</p>

<p>The gamma only determines the relative brightness with respect to white or black. The absolute brightness in cd ∕ m² is independent of gamma.</p>

<p>That method will not check the brightness of the other 253 shades of gray, so you have to hope that the display actually follows the gamma curve that has been set. Overall they should be fairly close.</p>

<p>Keep in mind that on most monitors, the dark grays near black are displayed much brighter than they really should be. RGB 0,0,0 should really be completely black and almost never is, and 1,1,1 should still be darker than almost any monitor’s 0,0,0 really is. For 1,1,1 I would expect most monitors to be multiple stops too bright. If you only need consistency, this at least becomes less of an issue since most monitors are like that.</p>

<p>I am not aware of any better solutions requiring no additional hardware. In theory you could match more colors with extra steps, halfway between the 50% gray and white, halfway to black, etc., but it would be time consuming, more error prone, and I have never seen such a tool.</p>

<p>Could you give more detail on why you are looking to control gamma? It is common to just assume either a gamma of 2.2 or the similar sRGB tone curve.</p>

<p>From what you've described, I'd just throw a stepwedge at the beginning of the course materials and have the learner adjust their brightness and/or contrast until everything's distinguishable.</p>

<p>Adobe Gamma is no longer supported, and with either Adobe Gamma or a home-grown eyeball solution you still end up needing the user to install yet another piece of software. I don't know what the market is for your training, but if any of it's being delivered to public computers (e.g., library, school lab) that may not be an option for the user. They can almost always fiddle with a monitor if they need to, on the other hand.</p>

<p><strong>Edit:</strong> Wrote my post before your latest one. Considering the rest of the information, yeah, you probably need something better than a stepwedge.</p>

<p>Thanks for the feedback.</p>

<p>It's not necessarily gamma that I need to address per se, but relative brightness, with an eye toward achieving something like an absolute brightness. The purpose is to get the learner to capture a virtual image that will be displayed on the screen, evaluate that image exposure, and adjust the camera controls before capturing another (and perhaps several more) image(s).</p>

<p>So, in other words, if a user's monitor is two stops too dark, that user is not going to be able to assess the exposure of the virtual image he/she has just captured and then make the correct adjustments.</p>

<p>Colin,</p>

<p>That sounds like a simple enough solution. Are you (or anyone else, for that matter) aware of anything like that which I could take a look at?</p>

<p>You would need to get his permission to use this particular chart, but the one about halfway down this page might do.<br>

<a href="http://www.normankoren.com/makingfineprints1A.html">http://www.normankoren.com/makingfineprints1A.html</a></p>

<blockquote>

<p>My question is, how accurate is gamma? </p>

</blockquote>

<p>Not very considering the crude controls on modern LCDs and how easily our visual system can be fooled in terms of adjusting for a single, solid color set (see: http://web.mit.edu/persci/people/adelson/checkershadow_illusion.html). Then there is the fact that what’s critical for proper color previews are ICC profiles and ICC aware applications that rely on the quality of a profile being well built and consistent. The RGB numbers you see today should look identical today and in a year. There’s a reason Adobe Gamma is unsupported and has been for years! The task requires instrumentation! </p>

<p>I was personally able to use Adobe Gamma on CRTs and get a decent result...but on LCDs it was absolutely impossible to get anything reasonable. Like Andrew says...it definitely needs instumentation.</p>

<p>Adobe Gamma is useless. Even Adobe has stopped putting this on their installation discs. You need to buy or borrow a kit which has a tool (photometer) to measure colors on you monitor and software to use those measurements to create a profile.</p>

<p>Joe C - Thanks for the site. I'll check that out. If we do end up needing something like this we'll create from scratch, but seeing what other people do is always valuable.</p>

<p>Thomas - that's been my experience as well. I was wondering if I just felt that it was innacurate because I was comparing it to my Spyder.</p>

<p>Thanks again for all the feedback.</p>