Archived

This topic is now archived and is closed to further replies.

What is the "Sub pixel accuracy" ?

This topic is 6644 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Recommended Posts

__Basically it's a fancy term for how well rendered polygons are anti-aliased (smooth edged.) "Sub pixel" refers to the rendering resolution (as opposed to the display resolution) of graphics.
__The accuracy of is often stated as a ratio of the two resolutions (rendered:displayed.) For example, a scene is rendered at 640x400 and is downsampled (squished) to fit on a 320x200 display. The accuracy would be 2:1 (sometimes stated as 2x.) In this example, the rendering time would be quadrupled! =(
__However, we're talking about polygons. The rendering time doesn't need to be that long. All that needs to be done is a sub-sampling of the lines constructing the polygons. Basically, instead of taking the interger of the calculated result (of y=x*slope), you split the value in to two parts: a bottom value (botValue = value - int(value)) and a top value (topValue = value - botValue). All that's left is adding (the new values multiplyed by the color of the line) to (the value of the current pixel * (100%-value)) and voila! Anti-aliased lines (of course someone else, probably could explain this better :} )
__I hope this helped!

Share this post


Link to post
Share on other sites
Hi everybody !

I'm not sure that I'm in the correct place, but I can't find better place to ask it.
So I've read somewhere about the "Sub pixel accuracy". The article said,that with this, smooth movement can be implemented in a polygon based word. There wasn't any explanation ,or theory or code in the article.
So I would like to ask what is the Sub pixel accuracy, how can it be implemented, what is its theory.

Thanks !

Share this post


Link to post
Share on other sites