Nosslaks stuff

[i30817]

A hair (on the upper strap) clips a bit too much. Isn't there a way to fix that without lots of work?

[Nosslak]

I was thinking of trying to increase my modeling pace, 1 model/week shouldn't be too hard to do. I was just wondering what kind of models would mappers want? The more specific you are the better it is. If possible I'd like to keep away from too generic (and some stuff that are way too easy such as meat hooks and melons) stuff.

 

I don't know how many people know about it but there's also this page on the wiki that could be updated as well with some of the requests so other people will have an easier time finding them (I know that the team is too busy to do it, but maybe the mappers can add some suggestions).

[Nosslak]

Still got a bunch left to do.

[7upMan]

Awesome! Also, can it move realistically?

[Nosslak]

Awesome! Also, can it move realistically?

Thanks! This is the highpoly so this won't move anywhere. It would look really sweet if I could get the lowpoly to move realistically along a rail but I'll probably need to make the CM with <16 tris so that won't happen. Hopefully it could be set up to go along a curve by a mapper when you push it down a slope or push a button or something like that instead.

[rich_is_bored]

That looks awfully modern. But take that with a grain of salt. I have no idea what flies in a universe with steam powered sentry robots.

[Melan]

It seems 18th - 19th century, which is pretty appropriate for a Thief-inspired game. Actually, the mine carts in TDP/TMA were more advanced, more late 19th or early 20th century tech (e.g. made of metal).

Edited September 20, 2011 by Melan

[Nosslak]

That looks awfully modern. But take that with a grain of salt. I have no idea what flies in a universe with steam powered sentry robots.

I've been wanting to make a minecart for quite some time now and I've basically just seen two varitions:

- Metal

- Wood

I just went with the one that looked the oldest.

[Glyph Seeker]

Loving your work Nosslak - thanks for posting here. Don't worry about the "era" for things like minecarts. Lord knows this universe is an aesthetic mish mash.

[Bikerdude]

Oooh pretty pretty - we need some mine carts, what about a steam engine to go with..?! Oh now that give me an idea for my next mission....

[Baddcog]

after seeing the pic of the wood one I think it's looking much better.

 

before that i thought it was a bit tall/cartoony, with the big blocky base and hieght to length ratio. But i was thinking those block below were metal..

 

I also figured it was still WIP.

 

On this should make 2 collision models (since it won't be a moveable it doesn't matter how complex it is).

 

So collision mesh one is wood parts (for arrows) and collision 2 would be for metal parts.

 

I see no reason why it can't move. It's 'movables' that need simple collision. things you pick up and throw, that bounce around.

 

But like your mask, you can attach an object to something and have it 'move'. In this case you'd need 1 cart model, and 1 wheel model. The wheel could be made to rotate with scripts (and could also be a misc moveable item), it can be attached to the cart, and the cart could be scripted to move.

 

Look at the wheel barrow also, would be great to have a 'materials' plane inside that can be skinned to empty, coal, rocks...

 

break it down into models. 1 complete stationary model, one with no wheels and one moveable collision wheel.

[Nosslak]

Loving your work Nosslak - thanks for posting here. Don't worry about the "era" for things like minecarts. Lord knows this universe is an aesthetic mish mash.

Thanks! I figured it wasn't that important but I felt that I wanted to rationalize why I chose it anyway.

 

Oooh pretty pretty - we need some mine carts, what about a steam engine to go with..?! Oh now that give me an idea for my next mission....

Thanks! Attaching a steam engine to it could be cool, I might look into it.

 

On this should make 2 collision models (since it won't be a moveable it doesn't matter how complex it is).

 

So collision mesh one is wood parts (for arrows) and collision 2 would be for metal parts.

I would've just made one but sure I can make 2 instead if you think that is worthwhile (only metalparts are some framing and the wheels).

 

I see no reason why it can't move. It's 'movables' that need simple collision. things you pick up and throw, that bounce around.

 

But like your mask, you can attach an object to something and have it 'move'. In this case you'd need 1 cart model, and 1 wheel model. The wheel could be made to rotate with scripts (and could also be a misc moveable item), it can be attached to the cart, and the cart could be scripted to move.

Oh, yeah, I suspected that might be what a moveable was. Thanks for the explanations.

 

Look at the wheel barrow also, would be great to have a 'materials' plane inside that can be skinned to empty, coal, rocks...

Clever, I'll make sure to add one of those.

 

break it down into models. 1 complete stationary model, one with no wheels and one moveable collision wheel.

Will do!

[Nosslak]

Here's an update for you too:

I'll just add some more wear and tear and then I think I'll move on to the lowpoly.

[Baddcog]

I'd do 2 collisions, it's big enough to warrant it. Would be weird to have an arrow stick in the wheel, or go clank and bounce off the wood.

 

But it could be fairly simple... Well, now that I look closer...

 

It's mainly only the wheels that aren't wood. Don't worry about collision for the metal bands.

 

I guess it depends.

 

If it's all one piece, yeah make metal for wheels and wood for rest.

 

If you just make the cart, with a seperate wheel, we could probably make an entity with attached wheels. Only issue is that's a bit harder for mappers to align to track as they can't actually see the wheels in editor. BUT, if they want it to move with rotating wheels, then maybe we just need a wiki entry 'Place mine cart ENTITY 48 units above track' or they can go in/out of game to check it.

 

after weighing options...

I'd say just export a wheel and a wheeless cart.

We'll make an entity with attached wheels, and it's easy enough to trigger the rotations (I have a test/machines map with switch/gears in a machine).

Might take a bit more magic to make the cart move around corners, but a straight track is pretty simple.

The cart itself might be able to be the trigger, player frobs it, it moves foawrd so many uints at such and such speed. and wheels rotate so far...

[Melan]

The wheels would also look good as random junk and... an old-school ventilator?

[Nosslak]

I'd do 2 collisions, it's big enough to warrant it. Would be weird to have an arrow stick in the wheel, or go clank and bounce off the wood.

 

But it could be fairly simple... Well, now that I look closer...

 

It's mainly only the wheels that aren't wood. Don't worry about collision for the metal bands.

 

I guess it depends.

 

If it's all one piece, yeah make metal for wheels and wood for rest.

 

If you just make the cart, with a seperate wheel, we could probably make an entity with attached wheels. Only issue is that's a bit harder for mappers to align to track as they can't actually see the wheels in editor. BUT, if they want it to move with rotating wheels, then maybe we just need a wiki entry 'Place mine cart ENTITY 48 units above track' or they can go in/out of game to check it.

 

after weighing options...

I'd say just export a wheel and a wheeless cart.

We'll make an entity with attached wheels, and it's easy enough to trigger the rotations (I have a test/machines map with switch/gears in a machine).

Might take a bit more magic to make the cart move around corners, but a straight track is pretty simple.

The cart itself might be able to be the trigger, player frobs it, it moves foawrd so many uints at such and such speed. and wheels rotate so far...

Sounds like a better plan. Thanks.

 

I've begun work on the lowpoly, but I think it is becoming way too highpoly. Right now it rests at about 1900 polys (the wheels are taking up a whopping 314 polys each) and I've still got some stuff left to add. Is ~2000 polys too much or will it be fine as they'd probably be used mostly in pretty closed caves/mines with not much else to render?

[Baddcog]

It seems like you could go pretty low on that. All the metal bands could just be baked. But hard to say without seeing wires.

 

But I don't think 2000 is really that much. After doing some mapping/lighting optimization again last week it just became really clear what's important. It's all about cast shadows.

I'd think a shadow mesh on that could be under 200 tris for the main cart (not wheels). And yeah, in a mine you probably won't have more than 203 lights on it at anytime.

And you should make an LOD too that could be half the tris of the up close version.

 

So with a very low poly shadow mesh and LOD's there's no reason we shouldn't start pushing a few more polys on nice objects imo.

[Nosslak]

It seems like you could go pretty low on that. All the metal bands could just be baked. But hard to say without seeing wires.

Yeah that's what I thought too but the wheels take up most of those polys (about 1300 tris) and I don't feel that I can optimize those much more without sacrificing roundness. I also added some bevels on the wood after doing some way overdue normal-baking tests on cubes which concluded in that the bevelled versions looked a lot better that the hard edged ones. I've also read several posts on Polycount that said what is important is how many vertices your model has which will turn out to be about the same amount if you make hard edges as if you bevel them. Here's the wires:

 

But I don't think 2000 is really that much. After doing some mapping/lighting optimization again last week it just became really clear what's important. It's all about cast shadows.

I'd think a shadow mesh on that could be under 200 tris for the main cart (not wheels). And yeah, in a mine you probably won't have more than 203 lights on it at anytime.

And you should make an LOD too that could be half the tris of the up close version.

I keep forgetting that it's the shadow meshes that matter every time. I'm not sure if I'll be able to do a good and much lower poly shadowmesh for the wheels as that are so thin that if lower the polys on the circles they will most probably intersect. I'll try to do a lod as well.

[Baddcog]

Yeah, when you make hard edges you are basically doubling the verts there. Though a few more verts isn't going to have that much more impact either.

 

Still, on wood planks beveled edges will look better.

 

I wouldn't worry too much about roundness of shadows, unless hit at just the right angle anyway that is lost anyway.

 

You can just make flat planes, then copy/paste/flip normals. Shadow meshes don't need volume as much as objects either. So the bow up top would be the same count minus all the edge polys and bevels.

[Nosslak]

There was a bigger delay than I first anticipated but the Mask™ is finally ready:

http://dl.dropbox.co.../plaguemask.rar

It is being committed to SVN as we speak so it should be in the next version (1.07) but if you want to use it for the contest or anything else before that this is your chance.

 

Some schoolstuff has come up so I haven't been able to finish the minecart yet. I will start texturing it soon.

[Bikerdude]

 

 

Hi Noss,

 

Can you make things like the wheels out of patches, and for that matter make them solid, no need for fancy spokes to be honest..

[Nosslak]

Hi Noss,

 

Can you make things like the wheels out of patches, and for that matter make them solid, no need for fancy spokes to be honest..

Um, I have no idea how to make patches or mostly anything else in DR. Every time I've tried DR it's felt like a really slow and awkward program so I haven't bothered learning it properly (I basically only know how to make box, playerstart and place a prop). Is there any reason to make it out of patches instead of a mesh? AFAIK the only advantage of patches is that the parts that cannot be seen can be culled and with models you can only ever cull everything or nothing. I imagine this is a great advantage for big walls and such, but for wheels I can see no advantage.

[Baddcog]

Yeah, I see no reason to make them out of patches. Though a simple low poly no spoke model would be fine. You could even just bake spoke 'ribs' onto a low poly wheel and it would look fancy without the need for modeled spokes and see thru wheels.

 

And something as small as that, even if it was patches, you'd be culling a few tris quicker than a model... Not a reason to do that imo as models are typically easier to use anyway. (collision mesh, shadows, entities all inclusive).

And tris are tris so patch or model, there's no advantage for performance (other than models can have a shadow mesh and patches can't)

 

And if they are seperate from the main model they would cull individually as the cart passes out of view of a vis-portal. ie: the front wheels and cart would be rendering but not the back wheels...

[Nosslak]

Yeah, I see no reason to make them out of patches. Though a simple low poly no spoke model would be fine. You could even just bake spoke 'ribs' onto a low poly wheel and it would look fancy without the need for modeled spokes and see thru wheels.

 

And something as small as that, even if it was patches, you'd be culling a few tris quicker than a model... Not a reason to do that imo as models are typically easier to use anyway. (collision mesh, shadows, entities all inclusive).

And tris are tris so patch or model, there's no advantage for performance (other than models can have a shadow mesh and patches can't)

 

And if they are seperate from the main model they would cull individually as the cart passes out of view of a vis-portal. ie: the front wheels and cart would be rendering but not the back wheels...

That's basically what I thought too. I don't know how to make patches at all and there would be close to no real performance gains or anything else (AFAIK, please correct me if I'm wrong) so I'll just stick to making the wheels and everything else of models.

 

Also just to be sure: I should make the coal/stone textures separate from the other textures, right? So that they could be re-used and swapped between easier?

[Bikerdude]

well I will have a stabe at making a replica of your cart in DR so you have something for comparision