This is an early WiP of my first serious map i´m doing for Doom 3. It´s a subway station. I guess i can´t export everything as one model but will have to export it as several different models.
And i hope i can normal map these inset thingies on the outer hull of the structure.

You can normal map the dome but you won't get near the same illusion of depth.

Actually you can export everything as one model, and for an open space like this there is no reason not to (although this is easier in Max than it is in Maya). I would redesign the vault so you have a combinaton of some realtime shadows, and normal maps. Make a few prominent beams etc. It should be fun to light. You may have issues with AI in such a space, but I'm sure it will look good!

The dome will kill the framerate if you let it like this (for casting dynamic shadows), so maybe you should try to double dome's squares...
For AI, you could "inline" your model, and create the basic structures with "clip "brushes and "AIclip" to let them walk where you allow them.

I really love to see really new architectures maps, especially "open" areas!

I would like to see an open area mapped out and run with a decent framerate.

Visually impressive, nice work!

Then comes texturing!, sounds, triggers, relays, script......
and ah yes lighting....an art in itself. have moving trains? see how it goes best luck with it looks great!

It is very impressive work, what modeling app did you use?

You really should make a lot of that in the map editor. Don't make everything static models when you can more easily make level geometry. It helps to know the strengths and weaknesses of both the level editor and a modelling program, so please do not become too dependent on making entire levels out of imported models.

Have "fun" texturing your model, because they will have pretty low-resolution textures if you export them as big models. (No megatexture technology for Doom 3 ) That's why it would be better to do the whole thing in the Doom 3 editor.

The design looks quite nice though, perhaps you should include a couple of round shapes (i.e. the handrails).

You can also use the Doom textures tiled into your 3D app, and for certain persons (like me ), it's easiest to UVWmap exactly like you want into Max (for example) rather than into Doomeditor...

The only advantage of using brushes is being able to generate visportals. Its always going to be easier and advantageous to make geometry in a dedicated modelling program.



The UV mapping abilities of any modelling program are far more sophisitcated than DoomEdit's. You cant unwrap in DoomEdit silly!

There are other advantages to using geometry. Game A.I. recognizes soild brushes as obstacles and don't need to be clipped.

You should also consider map performance. I don't have enough knowledge with the latter, so can anyone tell us if the map will be less or more hardware demanding if you mostly use static models?

It doesn't matter so *what* you use, it's *how* that counts. The map above will run like ass regardless - sorry about that Lumpy!

THe main issue here is the open-ness of the map, there's nothing there to do any blocking and using distance portals won't work becasue nothing to blocking the line of site to hide their effects.

The ceiling is better off as a normal map as rich_ said above, an object like that is drawing and wasting to many tris as a result of it being 'real', add light volumes to that and you've got problems with that many shadow volumes being cast at any one time.

The idea is great, but it's more of a visualistion than a game map so it'll need some serious 'compromising' to be done to idea (don't let that put you off btw Lumpengnom.

couldn't lumpy use that ceiling model as it is but with a noshadows casting material?

that would cut down on the number of shadow volumes being generated. also remember we can add in nodraw shadow casting parts of the models for when we want to fake the appearance of the visible model casting shadows.

Yeah that can be done but you always have to ask, is the feature worth the collosal polycount it's uses to be drawn?

Each light volume that hits it will draw polys (shadows or not) and the chances are it's going to need more than one light to fully illuminate it which will dublicate the tris for a feature that 'not important', i.e. it doesn't do anything other than look good.

So yes technically it can be done but it'll need a lot of work to make it FPS/runtime friendly.

I'm not even going to guess how many wholes there are that would need a shadowcasting texture btw!!

understood and good points kat

If you ask me (which I realize none of you actually did, but I can't shut up so here goes) nothing in that scene would really warrant the modeling of the scene.
The ceiling is too heavy to manage really, this is a really quite big and open scene and that's gonna be taxing for the engine to handle. And since it's fairly high up, a texture tiled on a patch mesh would do the trick just fine. As for all the ground stuff, nothing of it is very intricate or complex. It easily handled by brushes and some patch meshes.

ot: can the admins here ban Encetenpo45, all his posts seem to be totally ot and contain links to porn sites. its getting very annoying thinking there is an update to a thread I've been following, only to find its Encetenpo45 posting spam again

ratty i think its a spambot

This is an early WiP of my first serious map i´m doing for Doom 3. It´s a subway station. I guess i can´t export everything as one model but will have to export it as several different models.
And i hope i can normal map these inset thingies on the outer hull of the structure.

You can normal map the dome but you won't get near the same illusion of depth.

Actually you can export everything as one model, and for an open space like this there is no reason not to (although this is easier in Max than it is in Maya). I would redesign the vault so you have a combinaton of some realtime shadows, and normal maps. Make a few prominent beams etc. It should be fun to light. You may have issues with AI in such a space, but I'm sure it will look good!

The dome will kill the framerate if you let it like this (for casting dynamic shadows), so maybe you should try to double dome's squares...
For AI, you could "inline" your model, and create the basic structures with "clip "brushes and "AIclip" to let them walk where you allow them.

I really love to see really new architectures maps, especially "open" areas!

I would like to see an open area mapped out and run with a decent framerate.

Visually impressive, nice work!

Then comes texturing!, sounds, triggers, relays, script......
and ah yes lighting....an art in itself. have moving trains? see how it goes best luck with it looks great!

It is very impressive work, what modeling app did you use?

You really should make a lot of that in the map editor. Don't make everything static models when you can more easily make level geometry. It helps to know the strengths and weaknesses of both the level editor and a modelling program, so please do not become too dependent on making entire levels out of imported models.

Have "fun" texturing your model, because they will have pretty low-resolution textures if you export them as big models. (No megatexture technology for Doom 3 ) That's why it would be better to do the whole thing in the Doom 3 editor.

The design looks quite nice though, perhaps you should include a couple of round shapes (i.e. the handrails).

You can also use the Doom textures tiled into your 3D app, and for certain persons (like me ), it's easiest to UVWmap exactly like you want into Max (for example) rather than into Doomeditor...

The only advantage of using brushes is being able to generate visportals. Its always going to be easier and advantageous to make geometry in a dedicated modelling program.



The UV mapping abilities of any modelling program are far more sophisitcated than DoomEdit's. You cant unwrap in DoomEdit silly!

There are other advantages to using geometry. Game A.I. recognizes soild brushes as obstacles and don't need to be clipped.

You should also consider map performance. I don't have enough knowledge with the latter, so can anyone tell us if the map will be less or more hardware demanding if you mostly use static models?

It doesn't matter so *what* you use, it's *how* that counts. The map above will run like ass regardless - sorry about that Lumpy!

THe main issue here is the open-ness of the map, there's nothing there to do any blocking and using distance portals won't work becasue nothing to blocking the line of site to hide their effects.

The ceiling is better off as a normal map as rich_ said above, an object like that is drawing and wasting to many tris as a result of it being 'real', add light volumes to that and you've got problems with that many shadow volumes being cast at any one time.

The idea is great, but it's more of a visualistion than a game map so it'll need some serious 'compromising' to be done to idea (don't let that put you off btw Lumpengnom.

couldn't lumpy use that ceiling model as it is but with a noshadows casting material?

that would cut down on the number of shadow volumes being generated. also remember we can add in nodraw shadow casting parts of the models for when we want to fake the appearance of the visible model casting shadows.

Yeah that can be done but you always have to ask, is the feature worth the collosal polycount it's uses to be drawn?

Each light volume that hits it will draw polys (shadows or not) and the chances are it's going to need more than one light to fully illuminate it which will dublicate the tris for a feature that 'not important', i.e. it doesn't do anything other than look good.

So yes technically it can be done but it'll need a lot of work to make it FPS/runtime friendly.

I'm not even going to guess how many wholes there are that would need a shadowcasting texture btw!!

understood and good points kat

If you ask me (which I realize none of you actually did, but I can't shut up so here goes) nothing in that scene would really warrant the modeling of the scene.
The ceiling is too heavy to manage really, this is a really quite big and open scene and that's gonna be taxing for the engine to handle. And since it's fairly high up, a texture tiled on a patch mesh would do the trick just fine. As for all the ground stuff, nothing of it is very intricate or complex. It easily handled by brushes and some patch meshes.

ot: can the admins here ban Encetenpo45, all his posts seem to be totally ot and contain links to porn sites. its getting very annoying thinking there is an update to a thread I've been following, only to find its Encetenpo45 posting spam again

ratty i think its a spambot

understood and thanks, but could it still be blocked or banned somehow?

This is an early WiP of my first serious map i´m doing for Doom 3. It´s a subway station. I guess i can´t export everything as one model but will have to export it as several different models.
And i hope i can normal map these inset thingies on the outer hull of the structure.

You can normal map the dome but you won't get near the same illusion of depth.

Actually you can export everything as one model, and for an open space like this there is no reason not to (although this is easier in Max than it is in Maya). I would redesign the vault so you have a combinaton of some realtime shadows, and normal maps. Make a few prominent beams etc. It should be fun to light. You may have issues with AI in such a space, but I'm sure it will look good!

The dome will kill the framerate if you let it like this (for casting dynamic shadows), so maybe you should try to double dome's squares...
For AI, you could "inline" your model, and create the basic structures with "clip "brushes and "AIclip" to let them walk where you allow them.

I really love to see really new architectures maps, especially "open" areas!

I would like to see an open area mapped out and run with a decent framerate.

Visually impressive, nice work!

Then comes texturing!, sounds, triggers, relays, script......
and ah yes lighting....an art in itself. have moving trains? see how it goes best luck with it looks great!

It is very impressive work, what modeling app did you use?

You really should make a lot of that in the map editor. Don't make everything static models when you can more easily make level geometry. It helps to know the strengths and weaknesses of both the level editor and a modelling program, so please do not become too dependent on making entire levels out of imported models.

Have "fun" texturing your model, because they will have pretty low-resolution textures if you export them as big models. (No megatexture technology for Doom 3 ) That's why it would be better to do the whole thing in the Doom 3 editor.

The design looks quite nice though, perhaps you should include a couple of round shapes (i.e. the handrails).

You can also use the Doom textures tiled into your 3D app, and for certain persons (like me ), it's easiest to UVWmap exactly like you want into Max (for example) rather than into Doomeditor...

The only advantage of using brushes is being able to generate visportals. Its always going to be easier and advantageous to make geometry in a dedicated modelling program.



The UV mapping abilities of any modelling program are far more sophisitcated than DoomEdit's. You cant unwrap in DoomEdit silly!

There are other advantages to using geometry. Game A.I. recognizes soild brushes as obstacles and don't need to be clipped.

You should also consider map performance. I don't have enough knowledge with the latter, so can anyone tell us if the map will be less or more hardware demanding if you mostly use static models?

It doesn't matter so *what* you use, it's *how* that counts. The map above will run like ass regardless - sorry about that Lumpy!

THe main issue here is the open-ness of the map, there's nothing there to do any blocking and using distance portals won't work becasue nothing to blocking the line of site to hide their effects.

The ceiling is better off as a normal map as rich_ said above, an object like that is drawing and wasting to many tris as a result of it being 'real', add light volumes to that and you've got problems with that many shadow volumes being cast at any one time.

The idea is great, but it's more of a visualistion than a game map so it'll need some serious 'compromising' to be done to idea (don't let that put you off btw Lumpengnom.

couldn't lumpy use that ceiling model as it is but with a noshadows casting material?

that would cut down on the number of shadow volumes being generated. also remember we can add in nodraw shadow casting parts of the models for when we want to fake the appearance of the visible model casting shadows.

Yeah that can be done but you always have to ask, is the feature worth the collosal polycount it's uses to be drawn?

Each light volume that hits it will draw polys (shadows or not) and the chances are it's going to need more than one light to fully illuminate it which will dublicate the tris for a feature that 'not important', i.e. it doesn't do anything other than look good.

So yes technically it can be done but it'll need a lot of work to make it FPS/runtime friendly.

I'm not even going to guess how many wholes there are that would need a shadowcasting texture btw!!

understood and good points kat

If you ask me (which I realize none of you actually did, but I can't shut up so here goes) nothing in that scene would really warrant the modeling of the scene.
The ceiling is too heavy to manage really, this is a really quite big and open scene and that's gonna be taxing for the engine to handle. And since it's fairly high up, a texture tiled on a patch mesh would do the trick just fine. As for all the ground stuff, nothing of it is very intricate or complex. It easily handled by brushes and some patch meshes.

ot: can the admins here ban Encetenpo45, all his posts seem to be totally ot and contain links to porn sites. its getting very annoying thinking there is an update to a thread I've been following, only to find its Encetenpo45 posting spam again

ratty i think its a spambot

understood and thanks, but could it still be blocked or banned somehow?

This is an early WiP of my first serious map i´m doing for Doom 3. It´s a subway station. I guess i can´t export everything as one model but will have to export it as several different models.
And i hope i can normal map these inset thingies on the outer hull of the structure.

You can normal map the dome but you won't get near the same illusion of depth.

Actually you can export everything as one model, and for an open space like this there is no reason not to (although this is easier in Max than it is in Maya). I would redesign the vault so you have a combinaton of some realtime shadows, and normal maps. Make a few prominent beams etc. It should be fun to light. You may have issues with AI in such a space, but I'm sure it will look good!

The dome will kill the framerate if you let it like this (for casting dynamic shadows), so maybe you should try to double dome's squares...
For AI, you could "inline" your model, and create the basic structures with "clip "brushes and "AIclip" to let them walk where you allow them.

I really love to see really new architectures maps, especially "open" areas!

I would like to see an open area mapped out and run with a decent framerate.

Visually impressive, nice work!

Then comes texturing!, sounds, triggers, relays, script......
and ah yes lighting....an art in itself. have moving trains? see how it goes best luck with it looks great!

It is very impressive work, what modeling app did you use?

You really should make a lot of that in the map editor. Don't make everything static models when you can more easily make level geometry. It helps to know the strengths and weaknesses of both the level editor and a modelling program, so please do not become too dependent on making entire levels out of imported models.

Have "fun" texturing your model, because they will have pretty low-resolution textures if you export them as big models. (No megatexture technology for Doom 3 ) That's why it would be better to do the whole thing in the Doom 3 editor.

The design looks quite nice though, perhaps you should include a couple of round shapes (i.e. the handrails).

You can also use the Doom textures tiled into your 3D app, and for certain persons (like me ), it's easiest to UVWmap exactly like you want into Max (for example) rather than into Doomeditor...

The only advantage of using brushes is being able to generate visportals. Its always going to be easier and advantageous to make geometry in a dedicated modelling program.



The UV mapping abilities of any modelling program are far more sophisitcated than DoomEdit's. You cant unwrap in DoomEdit silly!

There are other advantages to using geometry. Game A.I. recognizes soild brushes as obstacles and don't need to be clipped.

You should also consider map performance. I don't have enough knowledge with the latter, so can anyone tell us if the map will be less or more hardware demanding if you mostly use static models?

It doesn't matter so *what* you use, it's *how* that counts. The map above will run like ass regardless - sorry about that Lumpy!

THe main issue here is the open-ness of the map, there's nothing there to do any blocking and using distance portals won't work becasue nothing to blocking the line of site to hide their effects.

The ceiling is better off as a normal map as rich_ said above, an object like that is drawing and wasting to many tris as a result of it being 'real', add light volumes to that and you've got problems with that many shadow volumes being cast at any one time.

The idea is great, but it's more of a visualistion than a game map so it'll need some serious 'compromising' to be done to idea (don't let that put you off btw Lumpengnom.

couldn't lumpy use that ceiling model as it is but with a noshadows casting material?

that would cut down on the number of shadow volumes being generated. also remember we can add in nodraw shadow casting parts of the models for when we want to fake the appearance of the visible model casting shadows.

Yeah that can be done but you always have to ask, is the feature worth the collosal polycount it's uses to be drawn?

Each light volume that hits it will draw polys (shadows or not) and the chances are it's going to need more than one light to fully illuminate it which will dublicate the tris for a feature that 'not important', i.e. it doesn't do anything other than look good.

So yes technically it can be done but it'll need a lot of work to make it FPS/runtime friendly.

I'm not even going to guess how many wholes there are that would need a shadowcasting texture btw!!

understood and good points kat

If you ask me (which I realize none of you actually did, but I can't shut up so here goes) nothing in that scene would really warrant the modeling of the scene.
The ceiling is too heavy to manage really, this is a really quite big and open scene and that's gonna be taxing for the engine to handle. And since it's fairly high up, a texture tiled on a patch mesh would do the trick just fine. As for all the ground stuff, nothing of it is very intricate or complex. It easily handled by brushes and some patch meshes.

ot: can the admins here ban Encetenpo45, all his posts seem to be totally ot and contain links to porn sites. its getting very annoying thinking there is an update to a thread I've been following, only to find its Encetenpo45 posting spam again

ratty i think its a spambot

understood and thanks, but could it still be blocked or banned somehow?

This is an early WiP of my first serious map i´m doing for Doom 3. It´s a subway station. I guess i can´t export everything as one model but will have to export it as several different models.
And i hope i can normal map these inset thingies on the outer hull of the structure.

You can normal map the dome but you won't get near the same illusion of depth.

Actually you can export everything as one model, and for an open space like this there is no reason not to (although this is easier in Max than it is in Maya). I would redesign the vault so you have a combinaton of some realtime shadows, and normal maps. Make a few prominent beams etc. It should be fun to light. You may have issues with AI in such a space, but I'm sure it will look good!

The dome will kill the framerate if you let it like this (for casting dynamic shadows), so maybe you should try to double dome's squares...
For AI, you could "inline" your model, and create the basic structures with "clip "brushes and "AIclip" to let them walk where you allow them.

I really love to see really new architectures maps, especially "open" areas!

I would like to see an open area mapped out and run with a decent framerate.

Visually impressive, nice work!

Then comes texturing!, sounds, triggers, relays, script......
and ah yes lighting....an art in itself. have moving trains? see how it goes best luck with it looks great!

It is very impressive work, what modeling app did you use?

You really should make a lot of that in the map editor. Don't make everything static models when you can more easily make level geometry. It helps to know the strengths and weaknesses of both the level editor and a modelling program, so please do not become too dependent on making entire levels out of imported models.

Have "fun" texturing your model, because they will have pretty low-resolution textures if you export them as big models. (No megatexture technology for Doom 3 ) That's why it would be better to do the whole thing in the Doom 3 editor.

The design looks quite nice though, perhaps you should include a couple of round shapes (i.e. the handrails).

You can also use the Doom textures tiled into your 3D app, and for certain persons (like me ), it's easiest to UVWmap exactly like you want into Max (for example) rather than into Doomeditor...

The only advantage of using brushes is being able to generate visportals. Its always going to be easier and advantageous to make geometry in a dedicated modelling program.



The UV mapping abilities of any modelling program are far more sophisitcated than DoomEdit's. You cant unwrap in DoomEdit silly!

There are other advantages to using geometry. Game A.I. recognizes soild brushes as obstacles and don't need to be clipped.

You should also consider map performance. I don't have enough knowledge with the latter, so can anyone tell us if the map will be less or more hardware demanding if you mostly use static models?

It doesn't matter so *what* you use, it's *how* that counts. The map above will run like ass regardless - sorry about that Lumpy!

THe main issue here is the open-ness of the map, there's nothing there to do any blocking and using distance portals won't work becasue nothing to blocking the line of site to hide their effects.

The ceiling is better off as a normal map as rich_ said above, an object like that is drawing and wasting to many tris as a result of it being 'real', add light volumes to that and you've got problems with that many shadow volumes being cast at any one time.

The idea is great, but it's more of a visualistion than a game map so it'll need some serious 'compromising' to be done to idea (don't let that put you off btw Lumpengnom.

couldn't lumpy use that ceiling model as it is but with a noshadows casting material?

that would cut down on the number of shadow volumes being generated. also remember we can add in nodraw shadow casting parts of the models for when we want to fake the appearance of the visible model casting shadows.

Yeah that can be done but you always have to ask, is the feature worth the collosal polycount it's uses to be drawn?

Each light volume that hits it will draw polys (shadows or not) and the chances are it's going to need more than one light to fully illuminate it which will dublicate the tris for a feature that 'not important', i.e. it doesn't do anything other than look good.

So yes technically it can be done but it'll need a lot of work to make it FPS/runtime friendly.

I'm not even going to guess how many wholes there are that would need a shadowcasting texture btw!!

understood and good points kat

If you ask me (which I realize none of you actually did, but I can't shut up so here goes) nothing in that scene would really warrant the modeling of the scene.
The ceiling is too heavy to manage really, this is a really quite big and open scene and that's gonna be taxing for the engine to handle. And since it's fairly high up, a texture tiled on a patch mesh would do the trick just fine. As for all the ground stuff, nothing of it is very intricate or complex. It easily handled by brushes and some patch meshes.

ot: can the admins here ban Encetenpo45, all his posts seem to be totally ot and contain links to porn sites. its getting very annoying thinking there is an update to a thread I've been following, only to find its Encetenpo45 posting spam again

ratty i think its a spambot

understood and thanks, but could it still be blocked or banned somehow?

This is an early WiP of my first serious map i´m doing for Doom 3. It´s a subway station. I guess i can´t export everything as one model but will have to export it as several different models.
And i hope i can normal map these inset thingies on the outer hull of the structure.

You can normal map the dome but you won't get near the same illusion of depth.

Actually you can export everything as one model, and for an open space like this there is no reason not to (although this is easier in Max than it is in Maya). I would redesign the vault so you have a combinaton of some realtime shadows, and normal maps. Make a few prominent beams etc. It should be fun to light. You may have issues with AI in such a space, but I'm sure it will look good!

The dome will kill the framerate if you let it like this (for casting dynamic shadows), so maybe you should try to double dome's squares...
For AI, you could "inline" your model, and create the basic structures with "clip "brushes and "AIclip" to let them walk where you allow them.

I really love to see really new architectures maps, especially "open" areas!

I would like to see an open area mapped out and run with a decent framerate.

Visually impressive, nice work!

Then comes texturing!, sounds, triggers, relays, script......
and ah yes lighting....an art in itself. have moving trains? see how it goes best luck with it looks great!

It is very impressive work, what modeling app did you use?

You really should make a lot of that in the map editor. Don't make everything static models when you can more easily make level geometry. It helps to know the strengths and weaknesses of both the level editor and a modelling program, so please do not become too dependent on making entire levels out of imported models.

Have "fun" texturing your model, because they will have pretty low-resolution textures if you export them as big models. (No megatexture technology for Doom 3 ) That's why it would be better to do the whole thing in the Doom 3 editor.

The design looks quite nice though, perhaps you should include a couple of round shapes (i.e. the handrails).

You can also use the Doom textures tiled into your 3D app, and for certain persons (like me ), it's easiest to UVWmap exactly like you want into Max (for example) rather than into Doomeditor...

The only advantage of using brushes is being able to generate visportals. Its always going to be easier and advantageous to make geometry in a dedicated modelling program.



The UV mapping abilities of any modelling program are far more sophisitcated than DoomEdit's. You cant unwrap in DoomEdit silly!

There are other advantages to using geometry. Game A.I. recognizes soild brushes as obstacles and don't need to be clipped.

You should also consider map performance. I don't have enough knowledge with the latter, so can anyone tell us if the map will be less or more hardware demanding if you mostly use static models?

It doesn't matter so *what* you use, it's *how* that counts. The map above will run like ass regardless - sorry about that Lumpy!

THe main issue here is the open-ness of the map, there's nothing there to do any blocking and using distance portals won't work becasue nothing to blocking the line of site to hide their effects.

The ceiling is better off as a normal map as rich_ said above, an object like that is drawing and wasting to many tris as a result of it being 'real', add light volumes to that and you've got problems with that many shadow volumes being cast at any one time.

The idea is great, but it's more of a visualistion than a game map so it'll need some serious 'compromising' to be done to idea (don't let that put you off btw Lumpengnom.

couldn't lumpy use that ceiling model as it is but with a noshadows casting material?

that would cut down on the number of shadow volumes being generated. also remember we can add in nodraw shadow casting parts of the models for when we want to fake the appearance of the visible model casting shadows.

Yeah that can be done but you always have to ask, is the feature worth the collosal polycount it's uses to be drawn?

Each light volume that hits it will draw polys (shadows or not) and the chances are it's going to need more than one light to fully illuminate it which will dublicate the tris for a feature that 'not important', i.e. it doesn't do anything other than look good.

So yes technically it can be done but it'll need a lot of work to make it FPS/runtime friendly.

I'm not even going to guess how many wholes there are that would need a shadowcasting texture btw!!

understood and good points kat

If you ask me (which I realize none of you actually did, but I can't shut up so here goes) nothing in that scene would really warrant the modeling of the scene.
The ceiling is too heavy to manage really, this is a really quite big and open scene and that's gonna be taxing for the engine to handle. And since it's fairly high up, a texture tiled on a patch mesh would do the trick just fine. As for all the ground stuff, nothing of it is very intricate or complex. It easily handled by brushes and some patch meshes.

ot: can the admins here ban Encetenpo45, all his posts seem to be totally ot and contain links to porn sites. its getting very annoying thinking there is an update to a thread I've been following, only to find its Encetenpo45 posting spam again

ratty i think its a spambot

This is an early WiP of my first serious map i´m doing for Doom 3. It´s a subway station. I guess i can´t export everything as one model but will have to export it as several different models.
And i hope i can normal map these inset thingies on the outer hull of the structure.

You can normal map the dome but you won't get near the same illusion of depth.

Actually you can export everything as one model, and for an open space like this there is no reason not to (although this is easier in Max than it is in Maya). I would redesign the vault so you have a combinaton of some realtime shadows, and normal maps. Make a few prominent beams etc. It should be fun to light. You may have issues with AI in such a space, but I'm sure it will look good!

The dome will kill the framerate if you let it like this (for casting dynamic shadows), so maybe you should try to double dome's squares...
For AI, you could "inline" your model, and create the basic structures with "clip "brushes and "AIclip" to let them walk where you allow them.

I really love to see really new architectures maps, especially "open" areas!

I would like to see an open area mapped out and run with a decent framerate.

Visually impressive, nice work!

Then comes texturing!, sounds, triggers, relays, script......
and ah yes lighting....an art in itself. have moving trains? see how it goes best luck with it looks great!

It is very impressive work, what modeling app did you use?

You really should make a lot of that in the map editor. Don't make everything static models when you can more easily make level geometry. It helps to know the strengths and weaknesses of both the level editor and a modelling program, so please do not become too dependent on making entire levels out of imported models.

Have "fun" texturing your model, because they will have pretty low-resolution textures if you export them as big models. (No megatexture technology for Doom 3 ) That's why it would be better to do the whole thing in the Doom 3 editor.

The design looks quite nice though, perhaps you should include a couple of round shapes (i.e. the handrails).

You can also use the Doom textures tiled into your 3D app, and for certain persons (like me ), it's easiest to UVWmap exactly like you want into Max (for example) rather than into Doomeditor...

The only advantage of using brushes is being able to generate visportals. Its always going to be easier and advantageous to make geometry in a dedicated modelling program.



The UV mapping abilities of any modelling program are far more sophisitcated than DoomEdit's. You cant unwrap in DoomEdit silly!

There are other advantages to using geometry. Game A.I. recognizes soild brushes as obstacles and don't need to be clipped.

You should also consider map performance. I don't have enough knowledge with the latter, so can anyone tell us if the map will be less or more hardware demanding if you mostly use static models?

It doesn't matter so *what* you use, it's *how* that counts. The map above will run like ass regardless - sorry about that Lumpy!

THe main issue here is the open-ness of the map, there's nothing there to do any blocking and using distance portals won't work becasue nothing to blocking the line of site to hide their effects.

The ceiling is better off as a normal map as rich_ said above, an object like that is drawing and wasting to many tris as a result of it being 'real', add light volumes to that and you've got problems with that many shadow volumes being cast at any one time.

The idea is great, but it's more of a visualistion than a game map so it'll need some serious 'compromising' to be done to idea (don't let that put you off btw Lumpengnom.

couldn't lumpy use that ceiling model as it is but with a noshadows casting material?

that would cut down on the number of shadow volumes being generated. also remember we can add in nodraw shadow casting parts of the models for when we want to fake the appearance of the visible model casting shadows.

Yeah that can be done but you always have to ask, is the feature worth the collosal polycount it's uses to be drawn?

Each light volume that hits it will draw polys (shadows or not) and the chances are it's going to need more than one light to fully illuminate it which will dublicate the tris for a feature that 'not important', i.e. it doesn't do anything other than look good.

So yes technically it can be done but it'll need a lot of work to make it FPS/runtime friendly.

I'm not even going to guess how many wholes there are that would need a shadowcasting texture btw!!

understood and good points kat

If you ask me (which I realize none of you actually did, but I can't shut up so here goes) nothing in that scene would really warrant the modeling of the scene.
The ceiling is too heavy to manage really, this is a really quite big and open scene and that's gonna be taxing for the engine to handle. And since it's fairly high up, a texture tiled on a patch mesh would do the trick just fine. As for all the ground stuff, nothing of it is very intricate or complex. It easily handled by brushes and some patch meshes.

ot: can the admins here ban Encetenpo45, all his posts seem to be totally ot and contain links to porn sites. its getting very annoying thinking there is an update to a thread I've been following, only to find its Encetenpo45 posting spam again

ratty i think its a spambot

This is an early WiP of my first serious map i´m doing for Doom 3. It´s a subway station. I guess i can´t export everything as one model but will have to export it as several different models.
And i hope i can normal map these inset thingies on the outer hull of the structure.

You can normal map the dome but you won't get near the same illusion of depth.

Actually you can export everything as one model, and for an open space like this there is no reason not to (although this is easier in Max than it is in Maya). I would redesign the vault so you have a combinaton of some realtime shadows, and normal maps. Make a few prominent beams etc. It should be fun to light. You may have issues with AI in such a space, but I'm sure it will look good!

The dome will kill the framerate if you let it like this (for casting dynamic shadows), so maybe you should try to double dome's squares...
For AI, you could "inline" your model, and create the basic structures with "clip "brushes and "AIclip" to let them walk where you allow them.

I really love to see really new architectures maps, especially "open" areas!

I would like to see an open area mapped out and run with a decent framerate.

Visually impressive, nice work!

Then comes texturing!, sounds, triggers, relays, script......
and ah yes lighting....an art in itself. have moving trains? see how it goes best luck with it looks great!

It is very impressive work, what modeling app did you use?

You really should make a lot of that in the map editor. Don't make everything static models when you can more easily make level geometry. It helps to know the strengths and weaknesses of both the level editor and a modelling program, so please do not become too dependent on making entire levels out of imported models.

Have "fun" texturing your model, because they will have pretty low-resolution textures if you export them as big models. (No megatexture technology for Doom 3 ) That's why it would be better to do the whole thing in the Doom 3 editor.

The design looks quite nice though, perhaps you should include a couple of round shapes (i.e. the handrails).

You can also use the Doom textures tiled into your 3D app, and for certain persons (like me ), it's easiest to UVWmap exactly like you want into Max (for example) rather than into Doomeditor...

The only advantage of using brushes is being able to generate visportals. Its always going to be easier and advantageous to make geometry in a dedicated modelling program.



The UV mapping abilities of any modelling program are far more sophisitcated than DoomEdit's. You cant unwrap in DoomEdit silly!

There are other advantages to using geometry. Game A.I. recognizes soild brushes as obstacles and don't need to be clipped.

You should also consider map performance. I don't have enough knowledge with the latter, so can anyone tell us if the map will be less or more hardware demanding if you mostly use static models?

It doesn't matter so *what* you use, it's *how* that counts. The map above will run like ass regardless - sorry about that Lumpy!

THe main issue here is the open-ness of the map, there's nothing there to do any blocking and using distance portals won't work becasue nothing to blocking the line of site to hide their effects.

The ceiling is better off as a normal map as rich_ said above, an object like that is drawing and wasting to many tris as a result of it being 'real', add light volumes to that and you've got problems with that many shadow volumes being cast at any one time.

The idea is great, but it's more of a visualistion than a game map so it'll need some serious 'compromising' to be done to idea (don't let that put you off btw Lumpengnom.

couldn't lumpy use that ceiling model as it is but with a noshadows casting material?

that would cut down on the number of shadow volumes being generated. also remember we can add in nodraw shadow casting parts of the models for when we want to fake the appearance of the visible model casting shadows.

Yeah that can be done but you always have to ask, is the feature worth the collosal polycount it's uses to be drawn?

Each light volume that hits it will draw polys (shadows or not) and the chances are it's going to need more than one light to fully illuminate it which will dublicate the tris for a feature that 'not important', i.e. it doesn't do anything other than look good.

So yes technically it can be done but it'll need a lot of work to make it FPS/runtime friendly.

I'm not even going to guess how many wholes there are that would need a shadowcasting texture btw!!

understood and good points kat

If you ask me (which I realize none of you actually did, but I can't shut up so here goes) nothing in that scene would really warrant the modeling of the scene.
The ceiling is too heavy to manage really, this is a really quite big and open scene and that's gonna be taxing for the engine to handle. And since it's fairly high up, a texture tiled on a patch mesh would do the trick just fine. As for all the ground stuff, nothing of it is very intricate or complex. It easily handled by brushes and some patch meshes.

ot: can the admins here ban Encetenpo45, all his posts seem to be totally ot and contain links to porn sites. its getting very annoying thinking there is an update to a thread I've been following, only to find its Encetenpo45 posting spam again

ratty i think its a spambot

This is an early WiP of my first serious map i´m doing for Doom 3. It´s a subway station. I guess i can´t export everything as one model but will have to export it as several different models.
And i hope i can normal map these inset thingies on the outer hull of the structure.

You can normal map the dome but you won't get near the same illusion of depth.

Actually you can export everything as one model, and for an open space like this there is no reason not to (although this is easier in Max than it is in Maya). I would redesign the vault so you have a combinaton of some realtime shadows, and normal maps. Make a few prominent beams etc. It should be fun to light. You may have issues with AI in such a space, but I'm sure it will look good!

The dome will kill the framerate if you let it like this (for casting dynamic shadows), so maybe you should try to double dome's squares...
For AI, you could "inline" your model, and create the basic structures with "clip "brushes and "AIclip" to let them walk where you allow them.

I really love to see really new architectures maps, especially "open" areas!

I would like to see an open area mapped out and run with a decent framerate.

Visually impressive, nice work!

Then comes texturing!, sounds, triggers, relays, script......
and ah yes lighting....an art in itself. have moving trains? see how it goes best luck with it looks great!

It is very impressive work, what modeling app did you use?

You really should make a lot of that in the map editor. Don't make everything static models when you can more easily make level geometry. It helps to know the strengths and weaknesses of both the level editor and a modelling program, so please do not become too dependent on making entire levels out of imported models.

Have "fun" texturing your model, because they will have pretty low-resolution textures if you export them as big models. (No megatexture technology for Doom 3 ) That's why it would be better to do the whole thing in the Doom 3 editor.

The design looks quite nice though, perhaps you should include a couple of round shapes (i.e. the handrails).

You can also use the Doom textures tiled into your 3D app, and for certain persons (like me ), it's easiest to UVWmap exactly like you want into Max (for example) rather than into Doomeditor...

The only advantage of using brushes is being able to generate visportals. Its always going to be easier and advantageous to make geometry in a dedicated modelling program.



The UV mapping abilities of any modelling program are far more sophisitcated than DoomEdit's. You cant unwrap in DoomEdit silly!

There are other advantages to using geometry. Game A.I. recognizes soild brushes as obstacles and don't need to be clipped.

You should also consider map performance. I don't have enough knowledge with the latter, so can anyone tell us if the map will be less or more hardware demanding if you mostly use static models?

It doesn't matter so *what* you use, it's *how* that counts. The map above will run like ass regardless - sorry about that Lumpy!

THe main issue here is the open-ness of the map, there's nothing there to do any blocking and using distance portals won't work becasue nothing to blocking the line of site to hide their effects.

The ceiling is better off as a normal map as rich_ said above, an object like that is drawing and wasting to many tris as a result of it being 'real', add light volumes to that and you've got problems with that many shadow volumes being cast at any one time.

The idea is great, but it's more of a visualistion than a game map so it'll need some serious 'compromising' to be done to idea (don't let that put you off btw Lumpengnom.

couldn't lumpy use that ceiling model as it is but with a noshadows casting material?

that would cut down on the number of shadow volumes being generated. also remember we can add in nodraw shadow casting parts of the models for when we want to fake the appearance of the visible model casting shadows.

Yeah that can be done but you always have to ask, is the feature worth the collosal polycount it's uses to be drawn?

Each light volume that hits it will draw polys (shadows or not) and the chances are it's going to need more than one light to fully illuminate it which will dublicate the tris for a feature that 'not important', i.e. it doesn't do anything other than look good.

So yes technically it can be done but it'll need a lot of work to make it FPS/runtime friendly.

I'm not even going to guess how many wholes there are that would need a shadowcasting texture btw!!

understood and good points kat

If you ask me (which I realize none of you actually did, but I can't shut up so here goes) nothing in that scene would really warrant the modeling of the scene.
The ceiling is too heavy to manage really, this is a really quite big and open scene and that's gonna be taxing for the engine to handle. And since it's fairly high up, a texture tiled on a patch mesh would do the trick just fine. As for all the ground stuff, nothing of it is very intricate or complex. It easily handled by brushes and some patch meshes.

ot: can the admins here ban Encetenpo45, all his posts seem to be totally ot and contain links to porn sites. its getting very annoying thinking there is an update to a thread I've been following, only to find its Encetenpo45 posting spam again

ratty i think its a spambot