Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!usc!apple!radius!pierce From: pierce@radius.UUCP (Pierce T. Wetter III) Newsgroups: comp.sys.mac.hardware Subject: Re: Mac IIci onboard video and gamma correction? Keywords: IIci video gamma Message-ID: <1225@radius.UUCP> Date: 6 Dec 89 19:00:34 GMT References: <5200@okstate.UUCP> <37037@apple.Apple.COM> <37038@apple.Apple.COM> Lines: 99 The hopefully last word on gamma correction: Monitor phosphors are not linear devices. If you shoot twice as many electrons at a phosphor you do not get twice as much light. You actually get less light or more or less the gamma-th root of the number of electrons. You can see this for yourself by putting a progression of gray bars on the screen: 5% 10% 15% etc. The values you put in for RG&B will seem to have little relation to the resulting brightness. The low numbers will be too dark, and there may be a big jump between the high numbers. In theory, all you would need to do to correct this fact is do create a table that indexes the values you send to the display card vs. the screen brightness. This is called a gamma correction table. All Mac II video cards support gamma correction tables. Unfortunately, its not quite that simple. Monitors possess something called burn-in. The electrons that bombard the phosphors come off a plate inside the picture tube. As a monitor gets older, there are less and less free electrons. (This is half simplification/ half analogy. It is not strictly accurate, but is mostly meant as an illustration.) Because of this the gamma response of a monitor changes with time. Most picture tubes start with a gamma of 2.8. Apple has chosen to use a table which will correct this to 1.8. (A comprimise between T.V.(2.2-2.8) and scanned images (1.0). Most programs assume that gamma is 1.0, or linear (twice the red is a red that is twice as bright.). There are a few problems with this approach. First of all, a gamma of 1.4 means that everything looks bad. For example, scanned images will look too dark and video images will look too washed out. Also there may be offsets in the monitor, 0 is supposed to be black but 0 may be 1% brightness or black may range from 0-15. (Circuitry is never exact. Turn up your brightness and contrast all the way, and look at a black square then turn them down all the way. Black is a relative term.) Finally, the different colors may age at different rates or have different gammas. The facility for changing the gamma table to "Uncorrected" in the Monitors CDEV prevents the gamma table from loading giving you the base gamma of the monitor, appoximately 2.8. This is probably a hint that Apple plans to have some video products soon. The NTSC (USA TV) signal is precorrected to a gamma of 2.2. Most TV sets have a gamma of 2.8, so if you set your monitor to Uncorrected and look at some video production it would look like standard TV. For anything else its pretty useless. ****** Plug for the project I am working on ****** Radius, Inc.'s solution is a product called PrecisionColor. (695 Sugg. Retail) This is a little ADB device and an associated CDEV. The CDEV talks to the ADB device, running through all 256 shades of each color and recording the brightness. Once this is done, you can set your gamma to a CALIBRATED gamma of (1.0, 1.2, 1.4, 1.8 (Apple), 2.2 and 2.8). You can also change the color temperature (white balence) to: 9300-Weird standard monitor bluish white 7500-Europeon daylight 6500-Old standard graphic arts daylight 5500- Yet Another Choice 5000- Standard Daylight - Paper White -Most useful 4100- Cool white flourescent - Paper under office lights. Also included is my portion of the project, the PrecisionPantones. PANTONE Colors* were originally matched on a Mitsubishi monitor with a different phosphor set from the currently popular Sony Trinitrons. Because of this, PANTONE Colors on the screen do not match the Pantone Colors in the standard booklet. What you See is Not What You Print. Because the PrecisionColor Calibrator removes the gamma, and offset variables from the monitor, Pantone and Radius have teamed up to improve the display of PANTONE Colors on Macintoshes. With the calibrator you get a Quark Xtension for Xpress 2.12 which will display Pantone Colors on-screen more accurately than ever before, a similar thing for Freehand 2.03, and a set of exact values for use in the "Monitor Calibration" dialog in PageMaker and Illustrator. The calibrator comes with a site license for the CDEV for 5 different machines (it is assumed that you only buy one calibrator and use that to calibrate every five machines in the office), and will ship 1Q '90. Note: it only currently works with our montors and the Apple RGB 13". Pierce. Disclaimers: Radius, Inc., has no knowledge and takes no responsibility for my postings whatsoever. I personally make no gauruntee as to their accuracy since anything made with a unix editor is automatically suspect. Details: The standard Apple gamma is not actually 1.8 its 1.8 that has been t tweaked until they liked how it looked. Note the discrepancy between the TV signal as broadcast (gamma =2.2) and the TV signal as received (gamma = 2.8) Since PrecisionColor is not yet shipping some details may change. Assuming any gamma for any uncalibrated system may be better than nothing, but make sure that users can turn off that assumption or change it to something else if they have a calibrated system.