CORE Copperhead +1
Moderators: MR.D, Moderators
Re: CORE Copperhead +1
polies are important..
a 4 legged beast, that there will be up to 16 of with 1k pollies in each foot alone, will own spring terribly..
a 4 legged beast, that there will be up to 16 of with 1k pollies in each foot alone, will own spring terribly..
Re: CORE Copperhead +1
The correct answer is to use LODs, only 2 levels would be perfectly fine in my opinion for the purpose of RTS. Then you can have an uberdetailed model when zoomed close and/or high settings or a lowpoly unit for the normal RTS perspective, with icons at the "strategic zoom" level.
Re: CORE Copperhead +1
The reason people say "polies matter" and "pollies don't matter" are because they are both oversimplifications.
Polycount obviously matters. It's more stuff the engine has to render, which is always going to take a longer time. HOWEVER, and this is where the "it doesn't matter" crowd comes from, there are other bottlenecks in spring that make the tiny change is speed from polycount disappear in the noise. Animation, pathfinding, los calculations, how efficient your unit and effect scripts are written, how well your lua is written, all matter so much more to overall run speed that debating whether to use 1,300 or 1,500 polys is ridiculous when you could use 10,000 without noticing any difference.
This was demonstrated experimentally several version ago. Whether it's true for all computers or for all versions, I don't know.
Polycount obviously matters. It's more stuff the engine has to render, which is always going to take a longer time. HOWEVER, and this is where the "it doesn't matter" crowd comes from, there are other bottlenecks in spring that make the tiny change is speed from polycount disappear in the noise. Animation, pathfinding, los calculations, how efficient your unit and effect scripts are written, how well your lua is written, all matter so much more to overall run speed that debating whether to use 1,300 or 1,500 polys is ridiculous when you could use 10,000 without noticing any difference.
This was demonstrated experimentally several version ago. Whether it's true for all computers or for all versions, I don't know.
Re: CORE Copperhead +1
Because half the people think it is and half dont... obviously.
Spherebot (CA peewee) is 1k polies and people spam hundreds of them. Make up your own mind...
Spherebot (CA peewee) is 1k polies and people spam hundreds of them. Make up your own mind...
Re: CORE Copperhead +1
I think polycount is always important, but managed well, you can have units with high poly.
A good rule of thumb being this, if a unit is going to be built many many times (more than 20 on screen at any time) be very limited with its polycount.
A model doesn't have to be 200 triangles to be a "good model", but a unit that will be very common and of which there are many of them on screen, the fewer polies = less strain on your video card.
If a unit is going to be seen less frequently or are later game units, why not splurge a bit on polies for the sake of details.
All of my LvL-1 units are under 800 triangles because I knew that there would be many of them used in an average game.
This Set of LvL-2 units I'm making I'm bumping up the poly limit to around 1400-1500 per unit, as there will be less numbers of them in comparison to LvL-1.
These people who say that Polycount isn't important, either don't know, or they personally have a fast enough computer to handle just about anything.
Its also not about any individual unit, but the sum of everything that will likely be in a single game being rendered.
And Btw, Spherebot in CA does cause a pretty substantial FPS hit when people spam them, its probably not as noticeable because many of the LUA implementations cause similar FPS lag, first that comes to mind is the ARM nanolathe particles which do some nasty performance dropping.
Always keep units which will be massed as the lower poly units, and you're golden.
A good rule of thumb being this, if a unit is going to be built many many times (more than 20 on screen at any time) be very limited with its polycount.
A model doesn't have to be 200 triangles to be a "good model", but a unit that will be very common and of which there are many of them on screen, the fewer polies = less strain on your video card.
If a unit is going to be seen less frequently or are later game units, why not splurge a bit on polies for the sake of details.
All of my LvL-1 units are under 800 triangles because I knew that there would be many of them used in an average game.
This Set of LvL-2 units I'm making I'm bumping up the poly limit to around 1400-1500 per unit, as there will be less numbers of them in comparison to LvL-1.
These people who say that Polycount isn't important, either don't know, or they personally have a fast enough computer to handle just about anything.
Its also not about any individual unit, but the sum of everything that will likely be in a single game being rendered.
And Btw, Spherebot in CA does cause a pretty substantial FPS hit when people spam them, its probably not as noticeable because many of the LUA implementations cause similar FPS lag, first that comes to mind is the ARM nanolathe particles which do some nasty performance dropping.
Always keep units which will be massed as the lower poly units, and you're golden.
Re: CORE Copperhead +1
Fair enough. Copperhead is a beauty anyway, like the new ridges and stuff (although the smooth cockpit was really nice too).
I think my computer could handle the polies, it'll play Crysis fine which must have millions more. Obviously can't speak for anyone elses though.
I think my computer could handle the polies, it'll play Crysis fine which must have millions more. Obviously can't speak for anyone elses though.
Re: CORE Copperhead +1
spherebot is also more complex both in objects and their hierarchy ALONG with the animation involved. It is not a one to one comparison.
Re: CORE Copperhead +1
lups nano spray is TWICE as fast as engine one ...MR.D wrote:And Btw, Spherebot in CA does cause a pretty substantial FPS hit when people spam them, its probably not as noticeable because many of the LUA implementations cause similar FPS lag, first that comes to mind is the ARM nanolathe particles which do some nasty performance dropping.
(and core nanolasers are SLOWER than arm nano particles) ⌐_⌐
Re: CORE Copperhead +1
Lua Particle System
Re: CORE Copperhead +1
So thats why I get fps drops when I'm viewing over ARM nano towers, and not when I look at the core beam nano??jK wrote: lups nano spray is TWICE as fast as engine one ...
(and core nanolasers are SLOWER than arm nano particles) ⌐_⌐
Gotta be an Nvidia vs ATI thing..
Because those ARM nano towers lag the crap outta me..
- Guessmyname
- Posts: 3301
- Joined: 28 Apr 2005, 21:07
Re: CORE Copperhead +1
Whether polycount matters or not is heavily dependent on your settings. Icon distance is the biggest for obvious reasons, but also reflectivity and that other whose name I forget (specularity): the more polygons, the more calculations need to be done for the reflections etc. Shadows probably also contribute to lag in a fashion directly proportional to polygons-on-screen, but seeing as I never have shadows on (they don't work with my graphics card), I can't say.
Re: CORE Copperhead +1
oh no doubt they can but the key thing here is to not waste polies on uneeded details or poorly executed geometry. I consider myself pretty perfectionistic but to be honest, there is only soo much you can do.
anyway, I do see things that I do not agrea about with the geomitry mr d uses some times but I think by and large he does a pretty good job. Hell, I thought I did pretty good with the morty model then rattle found a few issues. shit, the first 2 releases had the turret on the wrong side
anyway, I do see things that I do not agrea about with the geomitry mr d uses some times but I think by and large he does a pretty good job. Hell, I thought I did pretty good with the morty model then rattle found a few issues. shit, the first 2 releases had the turret on the wrong side
Re: CORE Copperhead +1
What i dont understad is why models are being totaly redone in the same way as the originals??i think the convehicle could have been made to be much better looking than just making the original with a facelift.
Mr.D you can change the original models if u feel u can make a better job youtself.make more detailed units.seriously.
Except units like peewees aks gators flash and so on you can make very detailed and high poly stuff.
Why make a box con vehicle f u can make it to be an awsome unit with complex opening animation and complex chasis.
Mr.D you can change the original models if u feel u can make a better job youtself.make more detailed units.seriously.
Except units like peewees aks gators flash and so on you can make very detailed and high poly stuff.
Why make a box con vehicle f u can make it to be an awsome unit with complex opening animation and complex chasis.
Re: CORE Copperhead +1
Because the originals are easily recognizable, and they work..
There is nothing from stopping any of you from doing this modeling and stuff too, if you require something a little different, go pick up wings or blender and get started.
I'm little over halfways done with the vehicles on CORE, and I'm not about to pull a 180 and start making units which are completely different from the OTA and Cinema models available from TA.
I'm going to keep going the way I have, and try to improve a little on each one as I go, which is all I can really do or wish to.
There is nothing from stopping any of you from doing this modeling and stuff too, if you require something a little different, go pick up wings or blender and get started.
I'm little over halfways done with the vehicles on CORE, and I'm not about to pull a 180 and start making units which are completely different from the OTA and Cinema models available from TA.
I'm going to keep going the way I have, and try to improve a little on each one as I go, which is all I can really do or wish to.
Re: CORE Copperhead +1
And doing original work takes at least 5times as long as working straight from references. You have to think about what to do, try different things until you get it right, and get creativity blocks all the time. While with references and especially if your good and have the techniques down you can just mechanically munch trough a lot of stuff steadily.
Of course this is why concept art is done...
Of course this is why concept art is done...
Re: CORE Copperhead +1
For what it's worth, I totally support this view.MR.D wrote:Because the originals are easily recognizable, and they work..
There is nothing from stopping any of you from doing this modeling and stuff too, if you require something a little different, go pick up wings or blender and get started.
I'm little over halfways done with the vehicles on CORE, and I'm not about to pull a 180 and start making units which are completely different from the OTA and Cinema models available from TA.
I'm going to keep going the way I have, and try to improve a little on each one as I go, which is all I can really do or wish to.
I love the thought of new models and higher poly counts, for the obvious reason of bring Spring/*A mods more up to date, but I also hate some of the horrible hashing that's happened to some of the remodels(e.g. the CA peewee. *Shudder*).
So keeping the OTA style/look while increasing detail is awesome.
Re: CORE Copperhead +1
Ok.Makes sense.
But i would be very disapointed if the building werent as detailed as supcoms.
They have some very nice and detailed buildings like mexes,shield generators and other buildings.
Imo before its too late buildings should have more changes in them and im not talking about storage buildings :) im talking labs mexes solars.
some fresh look on things.Im not saying make totaly new buildings but expand heavily onwhat exists. like 50%-50% old stuff and ur own cool ideas.
labs can be very very complex you dont have many of them.
I really didnt like the new vehicle lab and notcause it doesnt look better than the original,it does but i think a more drastic change is required.
Thats just me though its a single persons opinion but i reallythink some buildings have a totaly outdated and lame design.
But i would be very disapointed if the building werent as detailed as supcoms.
They have some very nice and detailed buildings like mexes,shield generators and other buildings.
Imo before its too late buildings should have more changes in them and im not talking about storage buildings :) im talking labs mexes solars.
some fresh look on things.Im not saying make totaly new buildings but expand heavily onwhat exists. like 50%-50% old stuff and ur own cool ideas.
labs can be very very complex you dont have many of them.
I really didnt like the new vehicle lab and notcause it doesnt look better than the original,it does but i think a more drastic change is required.
Thats just me though its a single persons opinion but i reallythink some buildings have a totaly outdated and lame design.
Re: CORE Copperhead +1
because when you put your spin on something people make idiotic claims about it.Gota wrote:What i dont understad is why models are being totaly redone in the same way as the originals??i think the convehicle could have been made to be much better looking than just making the original with a facelift.
Look at my morty, which is supposedly "lol" gundam influenced when in reality I pulled more from macross then anything.
goddamn smoth for thinking that an angular torso would be better then a flat cube when it comes to deflecting shots!
- Guessmyname
- Posts: 3301
- Joined: 28 Apr 2005, 21:07
Re: CORE Copperhead +1
Actually I figured that was more to do with the big Zeon-style red eye you gave it...