Forum posts
nodrawwwwwwww >:3
Will there be a third pass?
VLC for audio? How quaint.
For textureshadows to have any effect, you were correct in your previous post when you said you thought it required another file. It's very simple, however, and is just a lights.rad entry. When I'm less busy I'll go into detail about it, because it can be used to unbelievable effect. If you want to see it used incredibly well, check out Rabid's 2nd place compo entry cs_laundry.
Remember to compile with the RAD settings I posted on the previous page to get the shadows looking good. You can really see the difference StaticPropPolys makes (it casts model shadows based on their actual model, whereas by default the engine casts shadows based on their collision model) with stuff like the billboard frame on top of Urby's building.
$bsp_exe
-game $gamedir $path$file
$vis_exe
-game $gamedir $path$file
$light_exe
-both -final -TextureShadows -StaticPropPolys -game $gamedir $path$file
Copy File
$path$file.bsp $bspdir$file.bsp
Less than 1-unit. The building is off-grid. Zoom right in and you'll see.
Where are you seeing that message, ninja?
Increasing the sun spread angle has really killed the sharp shadows, so I'm going to lower it again. Uploading now, though. Exactly 23 hours, 45 minutes after starting :P.
Got a good solid ~6 hours done last night then another hour this morning when I woke up.
Please read the readme I've included. I've gone into a bit of detail about where I think the layout should go. Obviously this is optional and just my opinion, but please read it so you have it as an option.
Download
Oh, I forgot to mention - I recommend the following compile to achieve the best lighting:
Got a good solid ~6 hours done last night then another hour this morning when I woke up.
Please read the readme I've included. I've gone into a bit of detail about where I think the layout should go. Obviously this is optional and just my opinion, but please read it so you have it as an option.
Oh, I forgot to mention - I recommend the following compile to achieve the best lighting:
$bsp_exe
-game $gamedir $path$file
$vis_exe
-game $gamedir $path$file
$light_exe
-both -final -TextureShadows -StaticPropPolys -game $gamedir $path$file
Copy File
$path$file.bsp $bspdir$file.bsp
Progress:
Firstly spent a few hours just debugging & optimising. You guys really need to read your compile logs. There were about 20 physics models set to prop_static that just weren't rendering. Also someone had used displacements with a power of 4 (to make the leaf piles). Source gets reeeeallly unhappy about power of 4 displacements, and the compile log was screaming about them. Nodraw'd as many unseen faces as I could be bothered doing, but I really would like you guys to start doing that yourselves as you're brushworking. Makes such a huge difference.
Then I tightened up the lightmap grid and compiled with -textureshadows which makes everything sexier. Decreased the ambient light a tad on your light_env to make the shadows a bit darker, Blitz, but I like that you used the official Valve values for that skybox. Very nice!
Then I tweaked the garage area lighting and made a new stair texture.
Since then I've been working on layout. CT spawn, a sneaky route through an office and the hostage room, specifically.
Then I tightened up the lightmap grid and compiled with -textureshadows which makes everything sexier. Decreased the ambient light a tad on your light_env to make the shadows a bit darker, Blitz, but I like that you used the official Valve values for that skybox. Very nice!
Then I tweaked the garage area lighting and made a new stair texture. Since then I've been working on layout. CT spawn, a sneaky route through an office and the hostage room, specifically.
I just wasn't sure if you'd done it like that for a specific reason.
I recommend you make a test map with a high-res lightmap grid and a whole bunch of light_spots facing different directions but casting onto walls equal distances away. Play about with the constant values and falloff values to see the effects you can create with just one light_spot. It's a remarkably powerful entity.
I recommend you make a test map with a high-res lightmap grid and a whole bunch of light_spots facing different directions but casting onto walls equal distances away. Play about with the constant values and falloff values to see the effects you can create with just one light_spot. It's a remarkably powerful entity.
I prefer doing my details then finding a texture to suit them. If I can't find one, I make it there and then.
How long do I have left, btw? I only started ~6 hours ago, but CT posted the link ~16 hours ago.
Also, I have a query to whoever lit the garage area. You've used a very odd method which is producing some really weird, unrealistic results:
What made you use your method? Spotlighting near the ground like that is normally only used if the primary light source is a regular light entity, rather than a light_spot (which has so many variables that you can emulate the effect you've created very easily using constant and falloff).
I'm not a fan, anyhoo. Hope you don't mind if I change it a bit.
How long do I have left, btw? I only started ~6 hours ago, but CT posted the link ~16 hours ago.
Also, I have a query to whoever lit the garage area. You've used a very odd method which is producing some really weird, unrealistic results:
I'm not a fan, anyhoo. Hope you don't mind if I change it a bit.
Soak up the benefits of using overlays instead of decals, especially whoever put the dead leaves on the displacements. Always use overlays on displacements to avoid rendering issues in both Hammer and in-game... In fact I don't think there's any reason to use decals at all anymore tbh.
I'm gonna concentrate on actually making a playable layout, though.
Also, still seeing loads of brush based entities that haven't had their unseen faces nodraw'd. Those massive shelves, for example. If you're forgetting to do it - texture every brush you make completely nodraw at first, then use the face editor to individually texture faces once you've completed your brushwork. It also helps you take more care over your texturing (and whoever textured the shelves in unaligned combine metal could definitely use that extra care).
The stairs next to the shelves as well. That's 80+ unseen faces being rendered right there.
It's really shaping up nicely, though, guys. Really good work.
I'm gonna concentrate on actually making a playable layout, though.
Also, still seeing loads of brush based entities that haven't had their unseen faces nodraw'd. Those massive shelves, for example. If you're forgetting to do it - texture every brush you make completely nodraw at first, then use the face editor to individually texture faces once you've completed your brushwork. It also helps you take more care over your texturing (and whoever textured the shelves in unaligned combine metal could definitely use that extra care).
The stairs next to the shelves as well. That's 80+ unseen faces being rendered right there.
It's really shaping up nicely, though, guys. Really good work.
hlife_hotdog just described exactly to the letter my last experience with an AMD card.
Scale the texture to 1.01 or 1.022012 & Source. Bad TJB, bad!
Ha! Or do some original lines in that deedee style.
5 versions of 'she's turned the weans against us!'
This computer supposedly was just out a few months agoI'm afraid that's meaningless when it comes to prefab PCs. All the parts in it are at absolute least a year old.
I'm going to be at a media festival for the next 2 days (the one I'm pitching to the BBC in) but come Friday I will be in a mega mapping mood.
No! CT has actually done a really good job optimising that skybox, although it's a bit rushed around the tunnel area.
I notice a compo 31 carbon fibre material has been used on the roofs for some reason. That should probably be replaced.
I notice a compo 31 carbon fibre material has been used on the roofs for some reason. That should probably be replaced.
Go go go, Dimbark. You got yourself 24 real pretty hours.
Compiles'll be worse with everything func_detailed than otherwise.
Keep it to smallish structures & especially those with non square or triangular geometry.
Keep it to smallish structures & especially those with non square or triangular geometry.
Was gonna say it was pretty overpriced for what you're actually getting, but I suppose the dollar is fairly weak right now so it's actually less than £500 which ain't too shabby at all.
If you built it yourself, you'd save enough to be able to get a much better card, though. The GTX 560Ti is still out-performing everything ATI throw at it, and it'd give you a much more future-proof system.
I'd also be a bit nervous about getting an AMD processor over an intel, especially since non-sandybridge versions of the i7 series are plummeting in price (because sandybridge has so much more to give) but they're still more powerful than the similarly priced AMD alternatives.
Enjoy the new system, though, mate. Nothing like a shiny new PC
If you built it yourself, you'd save enough to be able to get a much better card, though. The GTX 560Ti is still out-performing everything ATI throw at it, and it'd give you a much more future-proof system.
I'd also be a bit nervous about getting an AMD processor over an intel, especially since non-sandybridge versions of the i7 series are plummeting in price (because sandybridge has so much more to give) but they're still more powerful than the similarly priced AMD alternatives.
Enjoy the new system, though, mate. Nothing like a shiny new PC
Yeah, it's just coming up with a good ratio of effectiveness when it comes to func_detailing. Some schools of thought say that every brush not sealing the world should be an entity, but that's just not true. VIS calculations go through the roof with that method, and you have to meticulously nodraw every unseen face by hand.
You're always best simplifying brushwork rather than func_detailing all of it. That was your mistake with the first version, CT. The road and pavement was a total mess, and even though it wasn't causing face splitting, it was still wildly inefficient to both compile and render due to unseen faces being rendered and the fact that the game will render all parts of the map tied to that entity, even if only part of it is actually visible.
Stairs are a bitch for geometry splitting, and I have an example map with the best method for making them, but essentially it's a world-brush ramp with func_detailed steps leading up it. The ramp itself won't cause any face splitting because the brushes around it already get split into tris during the compile.
The biggest problem with the current version is the skybox, though. I don't know who made it, but it's horrendously inefficient, causing VIS to calculate masses of unused space.
You're always best simplifying brushwork rather than func_detailing all of it. That was your mistake with the first version, CT. The road and pavement was a total mess, and even though it wasn't causing face splitting, it was still wildly inefficient to both compile and render due to unseen faces being rendered and the fact that the game will render all parts of the map tied to that entity, even if only part of it is actually visible.
Stairs are a bitch for geometry splitting, and I have an example map with the best method for making them, but essentially it's a world-brush ramp with func_detailed steps leading up it. The ramp itself won't cause any face splitting because the brushes around it already get split into tris during the compile.
The biggest problem with the current version is the skybox, though. I don't know who made it, but it's horrendously inefficient, causing VIS to calculate masses of unused space.
no worries tet, all the best with your stuff <3
There's no point just texturing them without building upon them. I just stuck them there so people could get an idea of how it might work layout-wise with the accessible buildings. If you're not interested in working on that part of the map then don't waste your time, dude, I'll grab it on my second turn.
VIS isn't finishing. Check for leaks/microbrushes.
I have my broadband back so I can actually check out the progress! 
You have AI grandkids :>
It's getting a standard SDK rather than L4D and Portal 2's "authoring tools" but aside from that nothing is confirmed.
And yes, you can open .vmf files in any Source version of Hammer.
And yes, you can open .vmf files in any Source version of Hammer.
The colour was fine, but for future reference I prefer the DB9 ;]
Stick a really directional light_env on it and pull the lightmap grid down to 2. Amplify those sexy shapes with some sublime shadows.
I'm glad you never over-react and approach everything with a level head, Bruce. Good job.
No, but you have lost because you were the only member who didn't try to buy us off with tribute.
That Aston Martin you sent me is purring like a kitten, Captain Terror.
(this is jokes)
That Aston Martin you sent me is purring like a kitten, Captain Terror.
(this is jokes)
looks at skybox
looks at urby
There's your compile problem.
looks at urby
There's your compile problem.
I sent my judging to Daubster so any further delay is all his fault.
Get your pitchforks and flaming torches ready.
Spoiler:
We already used all the beta keys on ourselves and our friends.
Get your pitchforks and flaming torches ready.
Spoiler:
We already used all the beta keys on ourselves and our friends.
ninja is now ~4 hours overdue for posting the map.
1. You need to seal off and COMPILE the map before you pass it of to the next mapper. In really don't want to try to debug/compile the errors from 8 different mappers x2, for a map that had never been compiled once. just no.My bad. I can confirm it compiles without any problems when sealed.
2. I don't see why you left stuff untextured.Sorry, just my way of working. I don't texture structures until they're roughly complete, and the buildings I've added are basically just placeholder blocks. I expect them to be vastly different by my next turn. This is particularly relevant for the interior sections, which at this stage are literally just hollowed out areas so that people can start working a playable layout into this.
3. packbsp your work so there is no question what assets are used in the map. Again, it is easy to extract the materials and such from the bsp using GCFScape.Yes, well, I didn't send a .bsp so that would have been tricky.
Just looked at the map file. I'm in way over my head already.Why?
He saved my marriage.
thinks everyone hates himThen he has still not read/understood anything of what was said to him.
TWHL members should not have to worry about having some sort of social responsibility for this kid. It creates a really uncomfortable environment.
just put the materials folder in your cstrike folder then restart Hammer
OK. I'm too tired to continue so I'm handing over the reigns 4 hours early.
Managed to get a good ~4 hours solid mapping done around college, though, so some progress has been made.
I pretty much rebuilt your brushwork from the ground up, CT, sorry about that but framerates would be at risk building the map the way you had it. It's still your map & layout, though.
Also made some road materials. That took a while, but they turned out not bad.
Here's the download and I believe the baton is now passed on to Ninja Defuse.
And here's a timer for your slot, ninja. I added on the 4 extra hours from my shift.
Managed to get a good ~4 hours solid mapping done around college, though, so some progress has been made.
I pretty much rebuilt your brushwork from the ground up, CT, sorry about that but framerates would be at risk building the map the way you had it. It's still your map & layout, though.
Also made some road materials. That took a while, but they turned out not bad.
Here's the download and I believe the baton is now passed on to Ninja Defuse.
And here's a timer for your slot, ninja. I added on the 4 extra hours from my shift.
When it's your turn - before you do anything - save the map as a new .vmf so as not to overwrite the previous person's.
Add a sequential number.
This way we can go back and see the specific changes made in each vmf.
edit:
Goddamn, CT. Who taught you how to optimise? That is not how you func_detail!
editedit:
And if you are going to func_detail massive world brushes like that to save on face splitting, make sure you nodraw every face that isn't seen because by making it an entity, you stop the unseen faces from being nulled during the compile.
editeditedit:
General tip for your style of mapping would be to always texture brushes entirely in Nodraw first, and then individually texture their visible faces. It's a good habit to get into and will really help your framerates if you're turning so many brushes into entities. It also tends to make you take a bit more care on each face which is always beneficial to the final product.
I like that you've left the 4 huge areas empty, by the way. Leaves a lot of room for creativity for everyone.
Add a sequential number.
This way we can go back and see the specific changes made in each vmf.
edit:
Goddamn, CT. Who taught you how to optimise?
That is not how you func_detail!editedit:
And if you are going to func_detail massive world brushes like that to save on face splitting, make sure you nodraw every face that isn't seen because by making it an entity, you stop the unseen faces from being nulled during the compile.
editeditedit:
General tip for your style of mapping would be to always texture brushes entirely in Nodraw first, and then individually texture their visible faces. It's a good habit to get into and will really help your framerates if you're turning so many brushes into entities. It also tends to make you take a bit more care on each face which is always beneficial to the final product.
I like that you've left the 4 huge areas empty, by the way. Leaves a lot of room for creativity for everyone.
Captain Terror is scart.
nah, i can dedicate a few hours on most days unless I'm shooting, which I don't think I am for the rest of March.
[simg]http://www.myfacewhen.com/images/52.jpg[/img]
The original way is not 'lazy', nor is your way 'more interesting'.
I'd much rather see someone's creative input to an existing map made by other people rather than limit them to, for example, texturing hallways. This isn't a job - there shouldn't be a hierarchy.
I'd much rather see someone's creative input to an existing map made by other people rather than limit them to, for example, texturing hallways. This isn't a job - there shouldn't be a hierarchy.
Go with the original idea, dude. Don't complicate this further.
Changing my vote to CS to push a winner and actually get this stone rolling.
V1
V2
What else does this hallway need? It's a level change transition so it doesn't need to be hyper-detailed, but it still feels really empty.