This project is read-only.

Anything new in the pipeline?

Jul 24, 2009 at 8:19 AM

Hey i was just wondering if there was anything new being planned for Xen? I love xen and while im still getting used to it i was just wondering if any new features would be getting added in the next few weeks - months?

Anyway good work :)

Jul 24, 2009 at 8:39 AM

Also a feature i was wanting if you are looking for stuff to add was parallax mapping mainly because i dont really know where to begin on getting it to work.

Also some other things i was thinking of was HDR, Depth of feild, MotionBlur, Parallax mapping etc...

(Also i know that to do any one of these would be a huge task and you do this in your free time and i thank you for that)

i was just wondering if you had any plans...

Jul 24, 2009 at 2:23 PM

Well I'll get it out the way: I won't be implementing any of those. :-)
Those techniques you mention are all complex implementations of graphics algorithms that require quite a lot of 'content support' and good understanding of what is going on in the graphics pipeline - they can't really just be enabled by a switch, or use a special shader (well, except basic parallax mapping).

What I've tried to provide, in the form of MaterialShader, is a good starting point. Lighting is a tricky thing to get right, so I've provided a very simple OpenGL like lighting model. Even this required a lot of work - it might surprise you.
Beyond that, I take the opinion that if you are advanced enough to be able to understand and use complex algorithms like parallax occlusion mapping and the like, then you don't need my help in implementing them. I want to make implementing them as easy as possible, but not actually do it for you - because then I'd be making assumptions about how you are building your game.

This is especially true for the other effects. Tone mapping, DoF, motion blur, etc. I've provided a lot of tools to make these *much* easier to implement (accurate blur filters, texture down sampling, etc). But when they are implemented, they will potentially have a radical effect on the architecture of a game and it's content/shaders/drawing. I can't predict that.
The other risk you run, is the 'dump truck' mentality, where you just use everything because you can. Parallax occlusion mapping is *spectacularly* expensive. And has to be used with extreme care to not destroy your framerate.

With all that said...

I do have a update waiting. It has some changes left to make, the big one is I want to get avatars working cleanly with custom animations (preferably in as similar a way to ModelInstance as possible). There are also some other things that are already complete, mostly bug fixes and enhancements to the shader plugin.

But.. I'm very busy right now so I can't give you a timeframe. For the last few months I've been job hunting, and now that's over I'll be moving to the UK soon (I live in New Zealand). This is obviously a big disruption and means I can't give any solid dates (let alone reply to questions in a timely manner it seems...). So if I don't get something out in the next 2 weeks, then it'll be at least another 3 weeks for something more.

Anyway. Thanks for your feedback :-) I know that's probably not quite the answer you were hoping for, but I have to be quite strict. :-)

Jul 26, 2009 at 10:33 AM

To be honest this is kinda what i was looking for.

A very difinitive answer :)
Also i didnt really expect anything on my list to be getting done i just thought i would check to see if its still getting updated etc..

Avatar animations are cool and a nice thing to to add.

On another note you menchined "basic parallax mapping" do you know of any good places to find out how to do this? any good tutorials for it in general or ideas of how to get started.
i know it takes alot of processing time and can slow down a game alot, but how much do you think it would slow down the 360?
Say i have a 5,000 tri model being rendered 150 times and i were to turn on parallax mapping would the framerate half, goto 1/3rd, or 10 times slower? just wondering if you had any kind of ballpark figure.

also this is not really a Xen question but do you think that adding say 40% of the previose frames final image onto the current frames final image would look like an ok motion blur? it would be cheap to do wouldnt it?

Also i really like the fact that this is not so much an engine but more a renderer. I looked at 7 different engines (4 free, 3 paid) and yours was the one that was chosen mainly because of its small size and it not trying to take over everything in xna.
(in case you are wondering the one that came second was sunburn but it costs upto $1000 so you win we got trials of all the other ones and didnt like any of them you have done very well.)