Hi(gh)!

To get rid of pixel artifacts when rendering Earth surface reliefs 
(mesh2 objects, not heightfields!) from topographic data, I tried smooth 
triangles... but as Melody's normal vector generator run abysmally slow 
(200 days for just one 3601 by 3601 data points tile!), I'm looking for 
a good introduction into smooth triangle normal calculation (the 
official 3.7 tutorials says nothing about it) - or at least a tool which 
does the math for me!

See you in Khyberspace!

Yadgar

Post a reply to this message

=?UTF-8?Q?J=c3=b6rg_=22Yadgar=22_Bleimann?= <yaz### [at] gmxde> wrote:
> Hi(gh)!
>
> To get rid of pixel artifacts when rendering Earth surface reliefs
> (mesh2 objects, not heightfields!) from topographic data, I tried smooth
> triangles... but as Melody's normal vector generator run abysmally slow
> (200 days for just one 3601 by 3601 data points tile!), I'm looking for
> a good introduction into smooth triangle normal calculation (the
> official 3.7 tutorials says nothing about it) - or at least a tool which
> does the math for me!
>
> See you in Khyberspace!
>
> Yadgar

From:  http://www.povray.org/documentation/view/3.6.1/295/

"...the corners are defined as in regular triangles and <Normal_n> is a vector
describing the direction of the surface normal at each corner.

These normal vectors are prohibitively difficult to compute by hand. Therefore
smooth triangles are almost always generated by utility programs. "

So, my first intuition is to just calculate the _surface_ / face normal vector
of the regular flat triangles, and average those normals to assign to the common
vertex.

Second, I'm not saying that this has _anything_ to do with what we're talking
about here, but it "rang a bell" and seemed superficially related:

https://www.youtube.com/watch?v=bohL918kXQk

So maybe there's a matrix way to approach this...

Perhaps some feelers could be sent out to StackExchange, StackOverflow, people
like Dan Schiffman, etc to see if this has already been approached from a
user-perspective.

Post a reply to this message