Re: Real-time Depth-of-Field Image processing

New Message Reply Date view Thread view Subject view Author view

Angus Henderson (angus++at++death.reading.sgi.com)
Fri, 5 Jan 1996 12:27:19 +0000


> I have two questions for you folks:
>
> - Have you heard of a different approach than the one described
> here for simulating the focus effect?
>

You could try blurring the scene by rendering it in a tiny window and
"rectzoom"ing it up to VGA. You could have multiple fields of focus by drawing
the scene once for each field and slicing them up with near/far planes or 6
axis clipplanes. You would need to layer the slices somehow, that would work if
rectzoom zoomed up the alpha values as well as RGB as each slice would have
transparent sections to see down to the next one. I just made this up so it has
no bearing on real GL but it might work....

Oh yes I have just thought - rather than all that rectzooming - you could
render your slices into texture maps usin fbsubtexload ( or some dodgey openGL
equivalent! ). Then I know you will maintain the alpha. My that would be a fibe
demo if it worked.

> We've also noted that a third graphics pipe would be necessary to
> handle the three separate CCTV views at 30 Hz. But even that will not
> give us enough time to render other channels on the pipe.
>
>

I wonder if Sirius video could help you, then you can render on one pipe and
pur the pixels into a texture map on another. That would improve your
performance but increase your latency.

mad,bad&glad

ANgus


New Message Reply Date view Thread view Subject view Author view

This archive was generated by hypermail 2.0b2 on Mon Aug 10 1998 - 17:52:13 PDT

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