Date: Sat, 29 Mar 1997 18:59:32 GMT From: Richard Tobin <richard@cogsci.ed.ac.uk> To: Amancio Hasty <hasty@rah.star-gate.com>, Richard Tobin <richard@cogsci.ed.ac.uk> Cc: multimedia@freebsd.org Subject: Re: newest bt848 driver Message-ID: <28003.199703291859@pitcairn.cogsci.ed.ac.uk> In-Reply-To: Amancio Hasty's message of Fri, 28 Mar 1997 16:25:40 -0800
next in thread | raw e-mail | index | archive | help
> > Something else that needs to be changed is in METEORSETGEO. The magic > Well thats what you are supposed to do 8) Yes, I know that :) > > By the way, what is the magic number 1.21875 in METEORSETGEO. Is it > > Good Question! The formula in the Bt848 databook is wrong so I came up with > my own formula 8) Right: the formulae in the databook are inconsistent. I'd like to know why rather than just plugging in numbers that seem to work. In particular, it says that the active portion is (for NTSC) 754 out of 910 unscaled pixels 640 out of 780 square pixels 720 out of 858 CCIR601 pixels These are all different proportions: 82.86%, 80.15% and 83.92% respectively. I also came across a description on NTSC with the figures given in uS instead of pixels, and that said the active portion is 52.4 out of 63.5 microseconds which is 82.52%, suggesting that the unscaled pixel version is closest to reality. Your calculation in the driver for calculating HSCALE corresponds to the second ratio. I wonder of you are missing a small piece of image off the end of each row as a result? Could you (or someone else with NTSC) change 1.21875 (ie 780/640) to 1.2069 (ie 910/754) in the horizontal scale calculation and see whether there is junk displayed at the end of each row? -- Richard
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?28003.199703291859>