Model pimpage Thread

Modelling, Textures, Animating and other general engine asset topics.

Re: Model pimpage Thread

Postby Smurftyours on Sat May 26, 2012 3:01 am

For question one you usually make the high poly model first, then you use the baking process in your 3d application to create normals and AO for the low poly. For the second question you also use texture baking, but instead of baking your model to a low poly model, you bake it to square plane. For max you use the bake to texture window, to bake normals you need to use a projection modifier, it would be inside the bake to texture window.
jangalomph wrote:Wise words from a wise man. ^
User avatar
Smurftyours
Senior Member
Senior Member
 
Joined: Tue Aug 04, 2009 6:11 pm
Location: California, US.

Re: Model pimpage Thread

Postby no00dylan on Sat May 26, 2012 3:18 am

How, modelling wise, do you go about making your high poly into a low poly, and preserve the UVW map?
User avatar
no00dylan
Veteran
Veteran
 
Joined: Sat Apr 19, 2008 10:34 pm
Location: NC, USA

Re: Model pimpage Thread

Postby Smurftyours on Sat May 26, 2012 3:21 am

You don't have to uv the high poly model. Basically you just removed edge loops rings and other details from the high poly and then UVW it.
jangalomph wrote:Wise words from a wise man. ^
User avatar
Smurftyours
Senior Member
Senior Member
 
Joined: Tue Aug 04, 2009 6:11 pm
Location: California, US.

Re: Model pimpage Thread

Postby no00dylan on Sat May 26, 2012 3:43 am

I'm a little confused, sorry. How exactly do I get the details from the highpoly into the lowpoly? I guess what I'm saying is I don't know how to bake the details from one model into another, just how to bake ambient occlusion.
User avatar
no00dylan
Veteran
Veteran
 
Joined: Sat Apr 19, 2008 10:34 pm
Location: NC, USA

Re: Model pimpage Thread

Postby Smurftyours on Sat May 26, 2012 4:06 am

Image
1. Skylight (for baking Ao/Normals)
2. Low Poly and High poly models directly stacked upon eachother
3. Render setup with light tracer enabled
4. Render to texture window (shortcut 0) With a projection modifier enabled and Normal and AO maps being baked

As for the projection modifier just select the HP as the model to project from, then reset the cage. For most simple models you then just need to use the select the entire cage and push it until all floating details are consumed by the cage. Same process for baking to planes, except the low poly model is just a plane and the cage is pushed to cover whatever details you are baking to the surface. Hope this helps.
jangalomph wrote:Wise words from a wise man. ^
User avatar
Smurftyours
Senior Member
Senior Member
 
Joined: Tue Aug 04, 2009 6:11 pm
Location: California, US.

Re: Model pimpage Thread

Postby no00dylan on Sat May 26, 2012 4:11 am

Thanks a bunch!
User avatar
no00dylan
Veteran
Veteran
 
Joined: Sat Apr 19, 2008 10:34 pm
Location: NC, USA

Re: Model pimpage Thread

Postby Sathor on Sat May 26, 2012 8:04 am

You should increase the Rays/Samples to something more than 250, though. More does not hurt. 5000 or so for the final pass.
User avatar
Sathor
Senior Member
Senior Member
 
Joined: Sat Jan 20, 2007 10:31 pm
Location: Germany

Re: Model pimpage Thread

Postby stoopdapoop on Sat May 26, 2012 8:40 am

Gambini wrote:@stopdapoop: How affects then fadeout distances for models regarding, again, performance?

And.. How is it that I converted a choppy map (average 15fps) into a well optimized map (30-40 fps) by merely adding a big occluder brush where a large building was and by adding lods to my models? And I´m talking of a lot of repeated models like bushes and fences.

And why then Valve uses up to 8 lod models with small polycount jumps (about -100 per lod in a 8000polys model). Aren´t they aware of all that?

No rethorical questions. I just want to learn too.



Well, fadeout distances are pretty straightforward, by doing really really cheap checks to determine if a prop far enough away to not be rendered you can save both CPU time and GPU time, Even if the model is "free" you'll still get the benefit of not having to switch driver states, which can be a relatively big plus.

and about your map, I can't say for sure because I don't know the map or the props that you optimized, but when you use an occluder it has similar benefits to fade distancing. You can perform a pretty quick check and potentially save time on the CPU and GPU. Unlike LOD's which usually only save time on the GPU, and can take more cpu time(although if you collapse bones or reduce the number of textures use in the lod, then you can save a lot of CPU time).

So of course the occluder would help, especially in the case of repeated models (although they're probably why your framerate was so bad in the first place) because they cut down on the number of discrete models in the scene.

And as far as lod models with small polygon jumps, I've honestly never seen that. Which models are you referring to? I know that the character models jump by thousands of polygons for the first few lod levels, then after they stop jumping by alot they jump by smaller numbers but start collapsing bones. This makes sense because at some point the bone tranformations on each vertex start netting you a lot more than just decreasing polygon count.

sorry for the late reply, I've been projectile vomiting and sweating enough water to fill a pond for the past few days.
I'm Brown
Image
User avatar
stoopdapoop
Veteran
Veteran
 
Joined: Sun Aug 21, 2005 2:14 am
Location: Ann Arbor, MI

Re: Model pimpage Thread

Postby Gambini on Sat May 26, 2012 3:27 pm

sorry for the late reply, I've been projectile vomiting and sweating enough water to fill a pond for the past few days.


Sorry to hear that. Maybe if you change your nick to stoopdapuke j/k

Take a look at the stove in ep2 props/forest/stove01. It doesn´t have a short jump as i mentionted but it has 7 submodels (ref+6 lods) although the polycount in the last one is ten times lower than in the ref. But according to your theory, that model is unoptimal.

ps: no bones, it´s static.
User avatar
Gambini
Sir Post-a-lot
Sir Post-a-lot
 
Joined: Mon Oct 20, 2008 1:52 am
Location: Buenos Aires, Argentina.

Re: Model pimpage Thread

Postby Stormy on Sat May 26, 2012 9:16 pm

This isn't pimpage so much as it is a request for help. This model is an animatronic statue/doll of brer rabbit from disneyland. I am having a really hard time making the fur believable. Any tips will be much appreciated.

Image

I've never painted fur before and I'm really finding it hard. The most effective way to paint fur is to paint the volumes that light creates when reflecting through the hairs, but with a texture there's no real way to tell where the light will be coming from. Besides that, I suck at picking the colours.

I've also tried to give it a few engine shaders to sell the fur effect, as you can see how I've busted up the rimlight and given it an exponent, I don't want to give it any translucent hair planes because the fur is all over and once I put one in I'll have to put them all over the place.

Tips? please?
User avatar
Stormy
May Contain Skills
May Contain Skills
 
Joined: Sun Nov 28, 2010 6:03 am
Location: Cairns, QLD, AUS

Re: Model pimpage Thread

Postby Armageddon on Sat May 26, 2012 9:24 pm

I think it looks pretty awesome (and creepy) like that. Maybe add some thin strokes on the arms though.
User avatar
Armageddon
Forum Goer Elite™
Forum Goer Elite™
 
Joined: Sun Dec 14, 2008 5:53 am

Re: Model pimpage Thread

Postby joe_rogers_11155 on Sat May 26, 2012 11:06 pm

can we see the reference photo?
"Day breaks, but decay soon follows." - Ava Winona
Currently developing in radio silence... 99 Bolts
Here's another project of mine... Assault on Overwatch
Are you new to Source SDK? VDC
joe_rogers_11155
Veteran
Veteran
 
Joined: Wed Oct 08, 2008 11:11 pm
Location: United States

Re: Model pimpage Thread

Postby stoopdapoop on Sun May 27, 2012 10:10 am

gambini wrote:Take a look at the stove in ep2 props/forest/stove01. It doesn´t have a short jump as i mentionted but it has 7 submodels (ref+6 lods) although the polycount in the last one is ten times lower than in the ref. But according to your theory, that model is unoptimal.


No, absolutely not.(well, I'd argue it was suboptimal for the PC, but for a different reason than you're asking).

Maybe I've not been very clear, but the essence of all these posts boils down to this: If the GPU is the bottleneck your "optimizations" need to save work for for the gpu by doing extra work on the cpu. This can be done on a per object basis(lods) and/or on a per scene basis(occluders, areaportals, hint brushes).

Somewhat related: Some modern engines and libraries can actually save work on the CPU by doing extra work on the GPU (the opposite of what I was talking about the last paragraph). But implementing systems like this is complicated even when working with consistent hardware or hardware with strict feature compliance (like all DirectX 10 and 11 cards) and it's show-stoppingly hard to implement on older hardware. Examples of this kind of work offloading can be found in BF3, where the Dice uses the "Compute shaders" no the GPU to generate textures for terrain. Another example is Ageia Physx, or the "Bullet" physics library.

If your CPU is the bottleneck, as is the case with all SMALL AND CHEAP props, then you lose performance by trying to save work for the GPU, by doing extra work on the CPU.

Back to stove01, that is not a small and cheap prop, so the GPU spends more time actually rendering the prop than the CPU takes telling the GPU to render the prop(because that takes time too). So if you can save work for the GPU by doing extra work on the CPU, then you'll save performance.

The CPU and GPU Run in parallel, In a perfect world the cpu will update the world state and issue commands to the gpu in the exact same amount of time that it takes the GPU to render the objects passed to it (albiet a frame behind). Unfortunately that is rarely the case, but these optimization techniques that we talk about exist elusively to make it closer to being true. In the usual case where one finishes before the other, the faster one will literally do nothing until the slower one finishes.

It's not straightforward. Mappers and modelers should have at least a basic understanding of the graphics pipeline and actual overhead of the techniques they're using to properly optimize a scene. If they use the wrong techniques then they can slow down your rendering, if they use the the right ones then they can speed it up. But most of the time I see people doing a mix of the two, they'll save time using some techniques but then gobble up some or all of their savings by using the wrong one somewhere else.
I'm Brown
Image
User avatar
stoopdapoop
Veteran
Veteran
 
Joined: Sun Aug 21, 2005 2:14 am
Location: Ann Arbor, MI

Re: Model pimpage Thread

Postby Gambini on Sun May 27, 2012 2:24 pm

I feel small reading such post. Understood everything i think.

In the usual case where one finishes before the other, the faster one will literally do nothing until the slower one finishes.


Except Vsync is enabled.
User avatar
Gambini
Sir Post-a-lot
Sir Post-a-lot
 
Joined: Mon Oct 20, 2008 1:52 am
Location: Buenos Aires, Argentina.

Re: Model pimpage Thread

Postby stoopdapoop on Mon May 28, 2012 1:00 am

bekey wrote:Someone copy paste this conversation to a new thread and maybe make a visual representation of the summary.


There are already a bunch of PDF's and PowerPoint presentations on the topic, just Google CPU GPU interaction and the various optimization techniques available in source.

gambini wrote:Except Vsync is enabled.


usually, yes. But if the gpu fills up it's buffers faster than it is able to display them, then the gpu will stall(2 buffers with double buffering, 3 with triple). Also if vsync is enabled but the cpu can't fill the render queue fast enough then the gpu will stall. So Vsync doesn't fully remedy the issue of the faster component waiting for the slower, it just stalls both of them when the they can both keep up with the refresh rate.

But yeah, I guess I'm done derailing the model pimpage thread.

MORE MODELS.

And stormy, I'm assuming it's some kind of Animatronic bunny like the ones you find at Chuck-E-Cheese? it looks really good, the only thing that sticks out to me are sharp edges on the leading edges of the foot.

somewhat related:
Last edited by stoopdapoop on Mon May 28, 2012 2:46 am, edited 1 time in total.
I'm Brown
Image
User avatar
stoopdapoop
Veteran
Veteran
 
Joined: Sun Aug 21, 2005 2:14 am
Location: Ann Arbor, MI
PreviousNext

Return to Art Creation

Who is online

Users browsing this forum: No registered users

cron