Re: Antialiasing Problem on iR

New Message Reply Date view Thread view Subject view Author view

Angus Dorbie (dorbie++at++bitch.reading.sgi.com)
Fri, 9 Aug 1996 17:10:13 +0100

Here ye go, note the 15, 1 bit z & stencil

Pixel format:
                      32 bits for pixel tag word
                    + 30 bits for RGB10 front buffer
                    + 30 bits for RGB10 back buffer
                    + 4 samples *
                        ( 24 bits for RGB8 color
                         + 16 bits for (Z & stencil) )
                = 252 bits

So you can get 2 * 1280x1024 with 256 pixel depth from 1RM6.

Warning, the 252 bits is dangerously close to the limit but in this case
it works.

Rgds,
Angus.

On Aug 9, 3:52pm, Mark Baranowski wrote:
> Subject: Re: Antialiasing Problem on iR
> On Fri, 9 Aug 1996 09:55:58 +0100, you wrote:
>
> >No, Shawn is correct.
> >
> >This is an Infinite Reality, 2RMs gives you enough screed real-estate
> >for 4++at++1280x1024, If you want to drive these at 60 Hz then you won't
> >have the bandwidth on the DG4 so you'll have to settle for
> >3++at++1280x1024_60+another_small_format.
> >
> >Again, small pixel depth supports 4 sample anti-aliasing, so it isn't
> >possible to configure an iR managed area which doesn't support some
> >anti-aliased visual.
>
> Are you sure small pixels support 4 sample anti-aliasing? I had a
> similar understanding to Brian Hill, in that a minimal multisample
> feature set:
>
> 8 bits RGBA 24 bits
> double buffering 24 bits
> Window IDs etc 64 bits
> 4 multisamples RGBA 96 bits
> 4 multisamples 23 bit Z 92 bits
>
> requires 300 b/p and is therefore too large for small pixels (256
> b/p).
>
> We only managed to get 3 (identical format ++at++ 60HZ) multisampled
> channels with the 1024x768_60. This was also the only combination
> which ircombine told us supported medium pixels.
>
> What have I got wrong/misunderstood?
>
> Please let me know ASAP, as we are about converge our projection
> system on Tuesday for 1024x768_60 on the basis that it was the only
> one to let us multisample on 3 channels.
>
> Regards,
>
> Mark.
> --
> Mark Baranowski at work (baranowski++at++marklynn.demon.co.uk)
> Pera, VR Division. Melton Mowbray, Leicestershire. UK.
> Tel: +44 1664 501501, Fax: +44 1664 501553
>-- End of excerpt from Mark Baranowski

=======================================================================
List Archives, FAQ, FTP: http://www.sgi.com/Technology/Performer/
            Submissions: info-performer++at++sgi.com
        Admin. requests: info-performer-request++at++sgi.com

New Message Reply Date view Thread view Subject view Author view

This archive was generated by hypermail 2.0b2 on Mon Aug 10 1998 - 17:53:20 PDT

This message has been cleansed for anti-spam protection. Replace '++at++' in any mail addresses with the '@' symbol.