At 06:54 �� 10/4/2002, you wrote: ><Goldfire bits snipped> > > > Being in the Aurora subject BTW, To switch to 512x256 mode 256 do you > > recommend setting both M0 on $18043 and $18063 (bits 3 on both) (together > > > of course with M1 (bit 4) on $18043) or just EMCR will suffice?... > >EMCR will suffice. The bits in the MCR that also appear in the EMCR are >really the same hardware bits. EMCR provides access to all functions, the >MCR was retained with only the standard bits because there was no guarantee >that the extra bits for the new functions would be written with sensible >values, every time existing programs wrote into it. >
That's what I figured because playing with both didn't give me any real difference and in any case any mode command resets them both. >Thanks, but it's actually even more ingenious (to my detriment...) than you >would think. First of all, there is a 'hidden' 1024 x 960 interlaced mode, >which is not supported in software. Hint: set display type at Multysinch >Interlaced (the manual says 'not' supported). 1024x768 becomes 1024x960 >interlaced. Hehehe I already have, although I am so against anything interlaced (well except TV) that I don't want to even try to do anything with it. The fact is that when accessing the Aurora screen memory, things FEEL a little faster (especially when accessing it via S*Basic) than the emulated screen. (Actually about twice as fast...) is it only my idea? Conceivably the emulation should take some speed out but half??? Or is it just me... Mind you I haven't written a benchmark or anything... just an overall feel :-) >Secondly, when nothing is mapped to FC000 (such as Qubide for instance), >the full 256k screen RAM is present rather than 240k like the manual says. >In theory, mode 4 on Aurora could go to 1024x1024 (odd aspect ratio, >though!), but more importantly, 16 color could go to 1024x512, 256 color to >512x512. The problem is that the MACH231 chip would have to be reprogrammed >for this, and a new Scan EPROM created, plus there would need to be a patch >wire on the board. Sadly, this 'extra' was taken out for future >compatibility with the GF and a possible LCD board. As it turned out, both >could have been done without getting rid of the last 16k and the last 32 >lines of the display in 16/256 color mode. >If I finally get my hands on the remaining 25 sets of Aurora parts, I may >even implement these changes, sort of a 'special edition' thing... Hehe better spend the time on making sure the GF becomes a reality :-) (How about that new CF processor? 233 MHz isn't it? - That's what the email that got lost was about.... I have to dig it out and send it again BTW :-)
