btw, Bitmap class seems to be using similar threshold on its smoothing property, when you change scaleX to something small. could be interesting to find out this native logic.
On Nov 24, 12:23 am, "Rob Bateman" <[EMAIL PROTECTED]> wrote: > ah i see! > > your right - it's different in away. The drawtriangle primitive always > arranges it's screenvertices in a clockwise order, regardless of whether > it's rendering the front or back of a face. This comes in handy for > post-render operations, such as hit-detection... > > Rob > > > > > > On Sun, Nov 23, 2008 at 10:12 PM, Makc <[EMAIL PROTECTED]> wrote: > > > Hey Rob, > > > I do understand what it is there for. But my point was that if you > > flip the order of vertices, the value will change its sign, and so the > > code, as is, will force no smoothing whatever the matrix is. Maybe > > this does not happen in away, but in sandy this does happen when the > > back side of face is rendered. > > > On Mon, Nov 24, 2008 at 12:09 AM, Rob Bateman <[EMAIL PROTECTED]> > > wrote: > > > Hey Makc > > > > the smooth formula is to stop smoothed materials from causing processor > > > burnout when the texture matrix has a hight level of transform. > > compressing > > > a bitmapfill bitmapData object into a line causes the flashplayer to > > throw a > > > fit as it tries to smooth hundreds of pixels on top of each other. This > > line > > > of code switches smoothing off if the material transform gets anywhere > > close > > > to thoes levels - usually as a result of a face being viewed end-on. > > > > this has been checked in teh past, and it seems to work for the purpose > > it > > > was designed. I'm not sure off the top of my head whether negative values > > > are allowed - really makes no difference as long as it stops the > > smoothing > > > bug! > > > > Rob > > > > On Sun, Nov 23, 2008 at 9:47 PM, Makc <[EMAIL PROTECTED]> wrote: > > > >> In AbstractRenderSession you have this construct: > > > >> .beginBitmapFill(bitmap, m, repeat, smooth && (v0x*(d2 - b2) - v1x*d2 > > >> + v2x*b2 > 400)); > > > >> while it is very cool hack, I think this value is negative in 50% of > > >> cases and so it's its abs() must be > 400 (why this value, btw? 100 > > >> worked for me, for example) > > > > -- > > > Rob Bateman > > > Flash Development & Consultancy > > > > [EMAIL PROTECTED] > > >www.infiniteturtles.co.uk > > >www.away3d.com > > -- > Rob Bateman > Flash Development & Consultancy > > [EMAIL PROTECTED] Hide quoted text - > > - Show quoted text -
