Author Topic: the human models thread  (Read 4612 times)

Shoukaiser/NeoXS

  • Posts: 6
  • Turrets: +0/-0
the human models thread
« on: August 15, 2001, 02:03:00 pm »
hey all, i guess i just sort of appeared hehe, ive been watching trem for a while

on a sidenote, just to mention one person atm, norf i respect your usefullness and creativity a lot more now, well actually ive only really seen it on this forum, but your a lot all around cooler here, or whatever youd like to call it hehe...while i doubt that youd care, just thought it was something to serve up on the platter

anyway, getting back to my original purpose, i remember a while back, people were wondering how human models were gonna be done, has that been decided since?  As a modeller/animator, the animation part is no problem to me, especially since its easier for q3 with its tags system (well for me anyway), but i would say to just keep to upper body lower body, head backpack and gun instead of having arms/legs in there, and anything else unless its crucial...gun head backpack are easy since they require little to no animation

as far as memory goes, if you have a bunch of pieces of armor, with it all adding up to a poly count of say, 1000-2000 polies, is that memory intensive for q3? I still havent gotten into animating/modelling much for it, but ive got grasps of things, i guess i should start working on ppm's at some point, maybe start coding/modelling/animating my own mod or something

so uh, wats da deal?

Norfenstein

  • Posts: 628
  • Turrets: +81/-78
the human models thread
« Reply #1 on: August 15, 2001, 05:22:00 pm »
Good people bring out the best in me...

From http://tremulous.sourceforge.net/phpBB/viewtopic.php?topic=24&forum=1&19

Hellbringer:  "I am pretty sure we have decided to use alpha channels to show/hide the extra parts of the models. This should keep the poly count as low as we posibly can and look the best we can do. The other two choices we were looking at was different skins and the use of tags but we found that each would suck."

Though I'd assume everything's still being considered...

Shoukaiser/NeoXS

  • Posts: 6
  • Turrets: +0/-0
the human models thread
« Reply #2 on: August 16, 2001, 02:14:00 am »
ah ok, thats a good idea, thats down to the skinner, correct?  Alpha channels used in certain types of grphical format right?  In q3 its for invis and some other things isnt it?  I dont really remember, but i do in a vague way

DarkStorm

  • Posts: 104
  • Turrets: +0/-0
    • http://tremulous.sourceforge.net
the human models thread
« Reply #3 on: August 16, 2001, 02:17:00 am »
Alpha Channels decide the opacity of an image.  Usually a black and white file of 256 colors, alpha channels are the difference between 24-bit and 32-bit color.
 have seen the light...
...and it burns.

Veda

  • Posts: 98
  • Turrets: +13/-5
the human models thread
« Reply #4 on: August 16, 2001, 06:22:00 am »
/me pokes his head around the corner

32 bit targa files are 1 24bit color image with a 256color(8bit) alpha map attached.

DarkStorm

  • Posts: 104
  • Turrets: +0/-0
    • http://tremulous.sourceforge.net
the human models thread
« Reply #5 on: August 17, 2001, 10:53:00 pm »
...yeah, and colors are generated by similar channels of Red Green and Blue, a good imaging program can seperate a full color image into its red green and blue component images, so that you can edit each of these individually.  That's why full color is called 24 bit.

8 bits per channel
3 channels
24 bit image

add an 8 bit alpha channel

32 bit image

[ This Message was edited by: DarkStorm on 2001-08-17 14:54 ]
 have seen the light...
...and it burns.