3d surface/terrain Matid`s & "polycount"
by Archies · in Artist Corner · 03/10/2005 (3:32 pm) · 7 replies
Hello GG Community
I'm considering a RacingType genre game first and foremost, im curious as to "Polycount/Facecount" limits in Torque ? Per-Vehicle-Model. This would possible be a mixed 1st person and Racing if all goes well and im very fearfull of the limits within TGE for vehicles and maybe vehicle count ? 4 players - 8 players 128 players ?. Als objects to replace the entire terrain element of TGE with an object"RACETRACK MESH = 12-15000 polygons ?
I have recently tested many static objects with 9k plus on the count and all looked ok, I was hoping from lowest 4000-7000-12000 polygons ? and i hope that TGE could cater for this but for the vehicles rather than static? if not, could you elaborate a little more on the Specs and Max limits what i could use and with "MULTPLAYER" in mind, collision on and off maybe ? which brings me to a Netcode/Network libraries question ?, i was reading somewhere that this was not included in the TGE indie ? is this correct and would have to use external sources for multiplayer networking or will it come out the box with Mp support.
Terrain/Surface Id's ?
Whilst testing with the demo, i exported a few models with varied success for both Vehicle and Track, the "Question" which then sprung to me was "Surface Id's" ?
For instance the terrain surface values for grass,tarmac,mud etc for varied values of Friction,Elasticty, and other Tge variables. Can these be used by [b]models[b] dts ? to store surface data for a racetrack. If so would this be the case of assigning Material Id's according to the surface parameters, this would be great and certainly put my mind at ease.
.
Looking forward to any replies and hope i can make the move and purchase what looks to be a stunning piece of kit.
I'm considering a RacingType genre game first and foremost, im curious as to "Polycount/Facecount" limits in Torque ? Per-Vehicle-Model. This would possible be a mixed 1st person and Racing if all goes well and im very fearfull of the limits within TGE for vehicles and maybe vehicle count ? 4 players - 8 players 128 players ?. Als objects to replace the entire terrain element of TGE with an object"RACETRACK MESH = 12-15000 polygons ?
I have recently tested many static objects with 9k plus on the count and all looked ok, I was hoping from lowest 4000-7000-12000 polygons ? and i hope that TGE could cater for this but for the vehicles rather than static? if not, could you elaborate a little more on the Specs and Max limits what i could use and with "MULTPLAYER" in mind, collision on and off maybe ? which brings me to a Netcode/Network libraries question ?, i was reading somewhere that this was not included in the TGE indie ? is this correct and would have to use external sources for multiplayer networking or will it come out the box with Mp support.
Terrain/Surface Id's ?
Whilst testing with the demo, i exported a few models with varied success for both Vehicle and Track, the "Question" which then sprung to me was "Surface Id's" ?
For instance the terrain surface values for grass,tarmac,mud etc for varied values of Friction,Elasticty, and other Tge variables. Can these be used by [b]models[b] dts ? to store surface data for a racetrack. If so would this be the case of assigning Material Id's according to the surface parameters, this would be great and certainly put my mind at ease.
.
Looking forward to any replies and hope i can make the move and purchase what looks to be a stunning piece of kit.
About the author
#2
You would not want to make your racetrack as a regular DTS model, though that would look great. Since you are having a race track, (formula1, nascar or what?) you would be best off using a .dif shape. The issues you would have involving collision on a DTS object for the track would be huge. Besides, a .dif would allow cool skidding effects etc.
You really should use .dif for the track. You can set different surface properties easily that way.
BTW, depending on how many cars you will have on the screen, poly counts can be quite high. If you have a high poly car, you can have a few LODs to make the race car at a distance render fewer polygons.
I am really suprised no one has made a car game for TGE yet. There is such cool support for it built right in.
03/11/2005 (5:42 pm)
TGE does come out of the box with network support. Even the indie version. The difference between indie and commercial versions is the with the indie you must publish through GG.You would not want to make your racetrack as a regular DTS model, though that would look great. Since you are having a race track, (formula1, nascar or what?) you would be best off using a .dif shape. The issues you would have involving collision on a DTS object for the track would be huge. Besides, a .dif would allow cool skidding effects etc.
You really should use .dif for the track. You can set different surface properties easily that way.
BTW, depending on how many cars you will have on the screen, poly counts can be quite high. If you have a high poly car, you can have a few LODs to make the race car at a distance render fewer polygons.
I am really suprised no one has made a car game for TGE yet. There is such cool support for it built right in.
#3
Ive been trying several engines and TGE seems very nice indeed, ive been browsing and testing with the demo fr a week or so now. Could i be so cheeky as for you to post maybe a few links for *.dif ?, programs ? tools ? converters maybe from max,blender ?. Would be great if i could create my work in a common 3d tool then convert to *.dif, although im not familiar with this format of *.dif and im sure anything you can offer will give me a better outlook on some tests.
Very much appreciate it if you could help on this as id like to get a little test setup in the demo before i hit that purchase button which im sure i will be doing.
btw, about the cars, no problems there, all my preliminary tests late last night went very well indeed, i was more impressed than ever once it render things quite smoothly with only 1 lod..
Thx
Arch
03/11/2005 (5:48 pm)
Hello Brian, thx very much for your reply,Ive been trying several engines and TGE seems very nice indeed, ive been browsing and testing with the demo fr a week or so now. Could i be so cheeky as for you to post maybe a few links for *.dif ?, programs ? tools ? converters maybe from max,blender ?. Would be great if i could create my work in a common 3d tool then convert to *.dif, although im not familiar with this format of *.dif and im sure anything you can offer will give me a better outlook on some tests.
Very much appreciate it if you could help on this as id like to get a little test setup in the demo before i hit that purchase button which im sure i will be doing.
btw, about the cars, no problems there, all my preliminary tests late last night went very well indeed, i was more impressed than ever once it render things quite smoothly with only 1 lod..
Thx
Arch
#4
Hopefully i can post some updates on this and maybe some things that ive been testing. Thank you to you all for your help and input during my thought periods of this great TGE demo.
03/13/2005 (8:50 am)
++ Im about to make the purchase of TGE with intent on additional updates of TSE if/when that time is needed.Hopefully i can post some updates on this and maybe some things that ive been testing. Thank you to you all for your help and input during my thought periods of this great TGE demo.
#5
Actually, that is not true at all. Their is NO requirement anywhere that you publish through GG (although if they are willing to publish your game, they give an extremely agreeable contract). The difference between "Indy" and "Commercial" licensing requirements is basically that if your company makes more thatn $250k a year US, you need to go back and purchase commercial licenses.
@Archies: I'm surprised that no one has brought this up yet, but your poly counts for models are extremely high. Torque can certainly handle models with that many polys, don't get me wrong, but I would suggest that you could trim out a very large portion of those polys with very few lost details, especially if you have a strong texturer. If you are starting to get 100's of those 4k-7k poly models on screen, you will most certainly get performance issues.
*.dif files are an internal representation that is converted from .map files generated by Quark (Quake Army Knife I think it is), a free editor that is available on the web. GG is also hard at work creating the Torque Constructor, which will be an organic modelling application that creates .dif files directly (and can also import .map files), removing the need for converters. The current converter that most people use is "map2dif", which can be searched for in the forums here, or the main documentation page.
03/13/2005 (9:20 am)
Quote:TGE does come out of the box with network support. Even the indie version. The difference between indie and commercial versions is the with the indie you must publish through GG.
Actually, that is not true at all. Their is NO requirement anywhere that you publish through GG (although if they are willing to publish your game, they give an extremely agreeable contract). The difference between "Indy" and "Commercial" licensing requirements is basically that if your company makes more thatn $250k a year US, you need to go back and purchase commercial licenses.
@Archies: I'm surprised that no one has brought this up yet, but your poly counts for models are extremely high. Torque can certainly handle models with that many polys, don't get me wrong, but I would suggest that you could trim out a very large portion of those polys with very few lost details, especially if you have a strong texturer. If you are starting to get 100's of those 4k-7k poly models on screen, you will most certainly get performance issues.
*.dif files are an internal representation that is converted from .map files generated by Quark (Quake Army Knife I think it is), a free editor that is available on the web. GG is also hard at work creating the Torque Constructor, which will be an organic modelling application that creates .dif files directly (and can also import .map files), removing the need for converters. The current converter that most people use is "map2dif", which can be searched for in the forums here, or the main documentation page.
#6
I'm yet to try anything with *.dif or dif2map yet but this sounds encouraging to me as ive still much work from previous years and projects in the map format which i honestly never would see the light of day again until now and TGE. :) Fingers crossed.
03/13/2005 (10:40 am)
Thx Stephen, sounds like its a winner in all directions i think, i certainly understand what your saying on the polycount and thx for the notes on that as whatever i do will be multiplayer based so these are some things i wanted to get clued up on.I'm yet to try anything with *.dif or dif2map yet but this sounds encouraging to me as ive still much work from previous years and projects in the map format which i honestly never would see the light of day again until now and TGE. :) Fingers crossed.
#7
Thank you for clarifying that. I consider myself educated. Should I ever make 250k on a game, I will gladly buy a commercial license ;)
03/14/2005 (8:05 pm)
@ StephenThank you for clarifying that. I consider myself educated. Should I ever make 250k on a game, I will gladly buy a commercial license ;)
Torque Owner Archies
Most tests went well for "Visual" aspects of meshwork both vehicle and track based formats i am working with, now think this is where the problem will be. Collision Surfaces ?? mmm i presume the "bounds" for object meshs dts export is the limit hey as once i changed that for mesh data it crashes bigtime.
...