To LUGNET HomepageTo LUGNET News HomepageTo LUGNET Guide Homepage
 Help on Searching
 
Post new message to lugnet.cad.devOpen lugnet.cad.dev in your NNTP NewsreaderTo LUGNET News Traffic PageSign In (Members)
 CAD / Development / 10534
10533  |  10535
Subject: 
Re: Finaly: I completed my new LDraw render engine
Newsgroups: 
lugnet.cad.dev
Date: 
Sat, 27 Jan 2007 22:57:45 GMT
Viewed: 
2909 times
  
Niels Bugge wrote:
This is just the kind of program I've been longing for since before I decided to
MLCad all my mocs - it was actually one of the main motivations. Man, I'm
looking forward to see it completed!!

If everything goes well, I'll expect the first beta in a couple of months.


A couple of questions:

- Assuming that the bot is an "actor" is there a limit to how many you can
  include?

Not as far I'm concerned. An actor describes the connections between an
group of parts. You can then use any actor in an 'character'. And there
is no coded limit on how many actors and derived characters there can
be. So in theory the only limiting factors will be memory and cpu.

To give an idea, the current renderer draws the wasp model in the
preview pictures at over 400 FPS on my Athlon 64 3200+ with an Radeon
9800 Pro. So in theory on this system one could make an animation with
about 15 of them running around and still keep 25 fps. But if the system
is going on it knees, you can always turn on 'skeleton' rendering for
one or more of the characters. Plus I will be doing frustum culling,
which means only stuff that's visible will be send to the VGA card, that
way you could have like a 100 characters in real time as long there's
never more than 15 of them on screen at once.

This is based on rendering only. The fully animating version will, off
course, need time to calculate all the angles etc. But my goal is to at
least be able to get 10 models at once one my system.

Also the speed is not an issue at all for the final povray export, it
will just take longer to render a single frame :)


- What about background and lego landscapes? Will it be possible to use
  background images (like starfields made via Photoshop) and what about virtual
  brick-build environments (like datsville)?


Background models, as far the software concerns, will be just
actors/characters but with no or very little joints.

I have not really thought about Background bitmaps yet, but once the
basic application is finished, these kinds of things are next on the to
do list.

- Will mpd's be supported so that you don't need to split complex models up in
  ldr's in order to assemble them again via the project parts window? - all my
  submodels are already positioned precisely in the joint coordinates, so if the
  program was able to import the mpd's directly I would only need to define the
  range (and direction) of motion in said parts.


The file loader supports MPD, but currently the whole assembled model
will be seen as a single project part for use in the actor editor.

But I was planning on adding some wizards (especially for generating
minifig actors). So an wizard for MPD based actor import can be very
useful indeed, thanks for the idea.

- Will it be possible to rotate the {base part} (dogfighting aircrafts and tanks
  moving over rugged terrain don't have immovable parts)?


Every actor has a main part, and in that main part you define a center.
This center's position and angle can be animated just like joints. The
only concern here is, animation of a 'dog fight' will be a bit tricky.
Because the user interface basicly works with XYZ rotation angles only.
This can give all sorts of problems, like the 'gimbal lock'. The
animation engine it self uses matrices but I am unsure on how to let
users directly manipulate the rotation matrices in a user friendly way.
Or perhaps these kind of movements demand specific animation plugins
whom can do whatever they want on a matrix level.


- Are you using L3p to make the POV-files, or can you include extra pov-tweaks
  to the output files (like special ambience and diffusion values, skyspheres
  and special lighting)?


I'm planning to use L3P to get the basic povray object definition for
the used parts. After feeding stuff to L3P it will be stripping the
result files to use the objects in a new 'combined frame or scene'
script. These scripts can than be edited before you feed them to povray
by hand. Or maybe I'll setup some custom postprocessing facilty, I don't
know jet.

Cheers NB

Roland



Message has 1 Reply:
  Re: Finaly: I completed my new LDraw render engine
 
(...) Thank {you} if you implement it ;-) - I'd hate to do all that work over again. BTW if you're able to incorporate that function, it would be nice if you can merge individual submodels since more than 90% of mine are usually just angled sections (...) (17 years ago, 28-Jan-07, to lugnet.cad.dev)

Message is in Reply To:
  Re: Finaly: I completed my new LDraw render engine
 
(...) This is just the kind of program I've been longing for since before I decided to MLCad all my mocs - it was actually one of the main motivations. Man, I'm looking forward to see it completed!! A couple of questions: Assuming that the bot is an (...) (17 years ago, 27-Jan-07, to lugnet.cad.dev, FTX)

7 Messages in This Thread:

Entire Thread on One Page:
Nested:  All | Brief | Compact | Dots
Linear:  All | Brief | Compact

This Message and its Replies on One Page:
Nested:  All | Brief | Compact | Dots
Linear:  All | Brief | Compact
    

Custom Search

©2005 LUGNET. All rights reserved. - hosted by steinbruch.info GbR