LDraw.org Discussion Forums

Full Version: LDraw Parts Tracker now uses LDView
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
Chris Wrote:I am pleased to announce that the LDraw Parts Tracker has been updated to render its images using LDView, instead of ldglite.
Thanks to Travis Cobbs for building a version of LDView on the server and to Don Heyse for all his previous support for ldglite.
Is it expected that silver color (80) looks so desperately black here http://www.ldraw.org/cgi-bin/ptdetail.cg...336p36.dat or here http://www.ldraw.org/cgi-bin/ptdetail.cg...336p35.dat ?

It appears not to be configured to use ldconfig.ldr. It also appears not to be using subdued lighting (which may be intentional). I'll talk to Chris about fixing the configuration.

(When ldconfig.ldr support was first added to LDView, it didn't work well with metallic and chrome colors, so I made the default off. At this point in time, that's probably not a good default. I'll probably change it in the next release.)
The antialiasing looks really good on the edge lines. Is there a way to get ldview to antialias the edges of the printed polys as well? Alternatively, is FSAA available with the osmesa renderer? Or as a last resort, render at 2x and resample back down with a post processor like imagemagick to smooth out the edges of the printed bits.
No, antialiasing the edges of printed polygons isn't supported. It on the list of potential future enhancements. FSAA also isn't supported by my OSMesa renderer. I strongly suspect that OSMesa itself does support it, since it seems to be pretty full-featured. I'm not sure how hard it would be to integrate support for it into my OSMesa LDView. I could probably also implement fake FSAA in LDView in the snapshot saving code. Multiplying the line widths by two and then doing a 2-to-1 downsample by averaging each group of 4 pixels would be pretty straightforward.
I suspect "fake FSAA" snapshots would really improve the legibility of the printed parts on the tracker. So it'd be really cool if you could pull that off.

I believe the postprocessing FSAA technique was used on these instructions and you can see the printed parts look much better than they would without it.
The configuration problem has been fixed and all images re-generated.