Irrlicht Engine

For those who don't know the IrrLicht Engine: it's a free open-source
high-performance 3d-engine, capable of rendering nice scenes.

More info on http://irrlicht.sourceforge.net/

They've got a vid on the site showing off a new feature for the next
version: parallax mapping. This is some sort of technique (Andrew or
Reno might give more accurate/technical description) that gives normal
textures 'depth' by performing some kind of special bumpmap effect on
them.

Here are some screens from the video you can download from the website:





That's a pretty cool effect for a free engine, no ? :smile:

---

:: Forceflow.be :: Nuclear Dawn developer

You'd actually be surprised how powerful free/open source engines are. The thing that really separates them from professional engines are the tools to easily implement those features and debug them.

For those interested in Parallax mapping, I found this link from a gamedev.net article.

http://graphics.cs.brown.edu/games/SteepParallax/index.html

---

---

"Every time you make something idiot-proof, someone creates a better idiot." -Anonymous

Aelon.net for Gaming News

the Source HL2 engine is capable of that too, i read somewhere, just not enabled like HDR.

Exactly as forceflow said, parallax mapping is essential a special form
of bump/normal mapping that gives surfaces added depth. I don't know
exact details so if people know better please do correct me, but I'll
attempt to give some sort of explanation.

Bump mapping gives each texel a depth value. A texel is essentially a
pixel on the texture - therefore if you put a 256x256 texture onto a
256x256 surface at a 1x1 scale, there are 256x256 texels on the
surface, whereas at 2x2 scale there are 128x128 texels. Normal mapping
(a further development of bump mapping) gives each texel not only a
depth value, but a normal vector. This normal vector stores the
direction that each texel is facing, which allows for the surface to
react to different light directions as if it featured far more
geometric detail than it actually does. Parallax mapping, as far as I
am aware, lets the surface not only react to the way each texel is
facing, but react to the other texels as well. For example, imagine a
texture has a single block that protrudes outwards from the rest of the
surface the texture represents. This texture could be applied to a
single face on a brush, and even with normal mapping if viewed from the
side it would appear to be flat (even if the texels recieved realistic
lighting due to how they are portrayed to "face" with normal mapping).
Parallax mapping lets each texel be rendered based on its depth and the
direction it faces. This bump in the texture would now block out texels
that would realistically be behind it from the current viewpoint of the
camera, as well as being lit realistically. This lets bumps be rendered
over texels that would actually be hidden behind them, giving surfaces
actual depth even if they are made of very few polygons.

I'm mildly drunk and not 100% on the theory behind the technique, so
this may not make much sense the way I have explained it. I guess the
best idea would be to read up on the link Edge provided - I'm sure it
covers things both better and more accurately than I have!

---

[img]http://card.mygamercard.net/sig/Default/reno84.png[/img]
Designer @ Haiku Interactive | ReNo-vation.net

Is it just me or do those shots look mildly cartoony? Like cell-shaded?

Well... it's similar to Quake3 engine I daresay... lots to make
progress of...like realism :biggrin: They are getting there, but Source is
already closer to this. U3 engine, though, it is a breakthrough in the
gaming industry. At least it should be... :rolleyes:

Anyway, it has a long time to develop...

---

HL2 tutorials 'n' stuff: http://madedog.pri.ee
217.159.236.34:27050 - CSS Server - Clean | koffer.ee

Is it just me or do those shots look mildly cartoony? Like cell-shaded?

Those are screencaps I took from the vid. Might be a bit different, you should watch the video, it's only 1 mb.

---

:: Forceflow.be :: Nuclear Dawn developer

I just watched it and I still think it looks cartoony. But
perhaps its just that setting where everythign is made of cobblestone.

Reno - basically the way I understand it is that it takes the normal
vector from the normal map and then offests the texels based on that
and the view vector. So as you move views the actual texels
appear move on the surface giving it depth.

---

To get something to work, sometimes you just have to beat your head against the wall longer; the skin grows back, but the brick doesn't.

Source hates soup!