Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!bryan From: bryan@cs.utexas.edu (Bryan Bayerdorffer @ Wit's End) Newsgroups: comp.sys.amiga Subject: Re: Amiga 3000 flickers? Message-ID: <204@mohawk.cs.utexas.edu> Date: 17 Jul 90 23:57:48 GMT References: <1990Jul14.043408.20134@cunixf.cc.columbia.edu> <5144@ethz.UUCP> <13260@cbmvax.commodore.com> Reply-To: bryan@cs.utexas.edu Organization: Spam Detection & Removal Squad, Austin, TX Lines: 28 Spam-Content: Negligible In article <13260@cbmvax.commodore.com> hood@cbmvax (Scott Hood) writes: [about flickering first line on A3000 interlaced display] =- =-This is due to how the current Amber chip does the de-interlace =-function. On 1950 monitors which allow the user to put the black =-borders behind the bezal of the monitor, the user will never see this =-video artifact. On other multiscan monitors that don't allow the user =-to adjust the size of the screen this top line is visible. It is only =-present when de-interlacing the Amiga's interlaced display modes. =- There are two minor problems with this. First, an overscan screen can begin on this flickering scan line, so if you want to hide the flicker you have to lose at least one line of overscan. Big deal. However, when the screen is stretched vertically on the 1950 to hide the flicker, one has to choose between losing the corners of the display behind the rounded CRT mask and using less horizontal overscan. Since the 1950 has no (external, at least) analog size control, a display using full horizontal overscan can't be squeezed just that little bit needed to clear the corners. Essentially it boils down to a choice between using full overscan and putting up with that maddening flicker. It's not such a big deal, though, since one can just buy a different monitor with a rectangular mask. Given that the 1950's colors and dot pitch are so, well, mediocre, that's not a bad idea anyway. What was the technical reason for not eliminating this flicker? Presumably it has nothing to do with the requirements of multifrequency monitors, since the flickerFixer does not exhibit this behavior.