New Renderer In the Works...Need Input

[AnimBot]

You are my new best friend!!!! Seriously I wish you guys the best of luck from the start to finish of this project. We as the 3d commuity really apreciate the effort of bringing quality at an affordable price (free):bounce: Good luck

[dvornik]

I know nothing about lights but our instructor wasn't able to preview the shape of photometric lights loaded into max 5 from IES files. So I guess this feature would be useful. Especially if you consider how the files are named.

[ambient-whisper]

if your going to make a renderer. then try to do something like a micropoly displacement freature. all max renderers have GI and all that . so this renderer would just fall into the pool. but a micropoly displacement unit would set you apart. :) what im saying is go into the areas that havent been developed much by others.

[Flypaper]

I totally agree, and this is one of the key issues that is being resolved. We are studying the effects of light from every possible angles. From atmosphere (the composition of atmosphere has a lot to do with the way light reacts, for example, elevations and such), To temperature and such. The big thing that we are noticing is that it seems (and we don't have their source or anything, but based on papers they've published) is that the majority of renderers are simply angular based. Reflection and bounces to put it simply. We are taking an open approach to this.
Since detail is the key to reproducing real world lighting we are going to attack any possible angle.

Some properties that have noticable effects on lighting are:
Atmosphere & Atmosphere Debris
Stratoshpere
Object Density
Types of Light
Temperature

We are looking to have an open caustic engine as well, which is dependent on an objects physical composition and make up, rather than max's material and light settings. Here you will be able to define your object's translucency, object mass, density and composition. We are hoping to make an editor which will allow for a large library of actual materials, in the literal sense of the word. Need frosted glass? It's easy to slap together a raytraced material with a frosted glass bump or displacement, but our material skining parameters will allow you to in 1's & 0's create all the pyhsics of frosted glass, from skin inward, while still keeping all of Max's main attributes. We are excited about this feature above any other, and the architecture of this will be something amazing in itself. The main goal is to create all this while keeping rendering times down. We've snagged one person in particular, who's dynamic approach to software architecture is truly mind boggling.

[dvornik]

Sounds like you're financed by Pentagon or something :)

[Flypaper]

LOL, Pentagon... nope, just some really cool investors who've picked up 3D as a Hobby and have an empty shell on the NASDEQ, with some money they need to get rid of. I'm not personally taking any money on this right now, instead I'm using it as research funding. And of course paying some of the programmers. Most of the people in on this are doing it purly for the challenge. It's a really cool thing to see.

[cdinic]

are you goiong to put up a site of any type? I would love to hear the backgrounds on the member of this project. some outlines of the goals, maybe even test renders as you guys produce them... anything would be facinating.

-Chris

[Fozzie]

Flypaper, this definately sounds pretty cool and I can't wait to see what you guys come up with for a result. Good to see some of the brainpower at the UofA going where it is needed and can be applied to something for a change :)

As for other features, here are some ideas off of the top of my head...
- Additional shaders
- Soft Shadows
- Option for shared rendering of a single frame across multiple machines on a network
- Exposure control
- Improved motion blur
- Adaptive subdivision for both lighting and rendering (based on camera distance)
- Sub Surface Scattering (translucency)

Well those are my ideas. Good luck on the project I look forward to seeing news of you guys in the local papers sometime soon.

Foz

[xynaria]

Going by your pics of your Max scripts in action....I think this is going to definately be worth keeping an eye glued on. Congratulations on getting the funding and the sincerest best wishes.
I'm glad someone is actually taking about the materials at last....like how the light is diffused by most surfaces before reflecting. Surafce roughness is definately the factor that seems to have been ignored by many as also is light absorption.

Personal obsessions...
That aa is adjustable on both an object and material basis, possibly with a means of fine tuning based on a colour picker.
That should you also go down the route of also offering alternative quick fakes solutions, that areas could be defined and key framed dependant on camera view, so as to be treat in a certain way, perhaps in the same way as you might in post.

:)

[gaggle]

Wow Flypaper, that sounds very cool, best of luck with that! A willing tester I be :)

Feature-wise the only thing I can think of that is a feature I tend to deem neccessary in a rendere for it to be useful:

Non-flickering global illumination.

That is all :) Few renderes currently has that as far as my experience goes, I know only of Brazil, which I've heard but not tested is supposed to feature that.

[trybul]

Flypaper.....Congratulations!

This is great....and it sounds very promising...

I think that you are focusing on some areas that others have left unattended, such as what affects the properties of light. and the most impressive undertaking of property, instead of material, based caustics.....

If you want to have a local tester that can give you feed back and can test it in some diverse aplications, just let me know......


trybul

[Lizard Head]

From the technology to the pointless observation as an end user consumer :

Lumicenter doesn't roll off the tounge as a catchy name,, one to many syllables

Lumipoint may work better


IMO

Adios

[WarlockXKZ]

Originally posted by Flypaper
Creating a new renderer...

Again my end goal is to keep this FREE and open source. We all pay through the roof for this software as is, and far as I'm concerned it's more of a personal goal to see what we can come up with.



Please keep it this way. Being open source (and free :) ) it can evolve much quickier.

Will this project be 3ds Max 5 only?

[Joebount]

Best of luck guys ! If you need beta testers, I'm your man :):thumbsup:

[Joebount]

Best of luck guys ! If you need beta testers, I'm your man :):thumbsup:

[ilasolomon]

with my best wishes for you & your friends... :beer:
count on me for beta testing!

[ambient-whisper]

Originally posted by WarlockXKZ


Please keep it this way. Being open source (and free :) ) it can evolve much quickier.

Will this project be 3ds Max 5 only?

i hope not. max has enough renderers. it would be nice if it was stand alone with plugs to apps. and not just for one app. ( plus if izware starts selling again. id love to see a mirai plug :D if this thing ends up being good:)

[WarlockXKZ]

Originally posted by ambient-whisper


i hope not. max has enough renderers. it would be nice if it was stand alone with plugs to apps. and not just for one app. ( plus if izware starts selling again. id love to see a mirai plug :D if this thing ends up being good:)

Me too, I hope this will be stand alone, so everybody could use it.

But if not, I would like to know if it will be max 4 compatible?

[Flypaper]

Well, the first version will be a Max Renderer, with a Maya Renderer following soon after. The problem with keeping it as a standalone application, is the framework needs to be much larger. Max/Maya handle their environment in a totally unrelated way, so each seperate renderer will have to be aptly customized for each software.

The actual renderer itself will be the easy part to port into other applications, It's the way the renderer handles the materials and light that worries me. This will definately be a challenge. Because of our Composition editor, we will definately have to rewrite our software in many different versions. I have funding, but I don't have that much. Since I am a Max man, and always have been, it will be the guinea pig. It's also a way for me to take max up the the serious production environment level as well, it's an amzing piece of software, that needs some serious render enhancements. The framework between Max 4 & Max 5 is similar enough that the plugin should work in both these versions, since we are all together bypassing the max 5 GI settings.

The thing that will set Lumicenter (or Lumipoint!) apart is the open source of the architecture. With module plugins the sky is the limit. This will open the horizon to different types of GI Algorithms and styles of GI, as well as stylized enhancements. After the release of the full version we will offer the source code for these areas. One team can only do so much, imagine what the world can do. (Look at the Unix open source environment!)

I've been emailed and asked how are we supposed to make money off this... I want to keep this free. Aside from the documentation provided online, we will offer live support, for which there will be a fee. The development kit will also be free, but any further documentation will require a small surcharge. I hope this sits well. Any other costs should be supplimented by advertising costs and such on the site.

If anybody knows of someone who is willing to host the official site from their server, for small charges, or possibly free, please let me know. I'm trying to cut costs as much as possible, so that i can funnel most of the resources in to development.

Thanks for all the support everyone!

[WarlockXKZ]

Sorry, I didn't read it carefully :hmm:

[ErickG]

Have you looked at the matropolis light transport work yet. Looks very deep. The big problem with all GI renderes these days is the time it takes to get feedback. If there was a way to speed-up the feedback prior to spending the time to test render, you would have a winner on your hands.

I have played around with basic GI code but I think there is a basic flaw in the way light is calculated. Thinking outside the box will help to crack this one. Photons are tricky little creatures and light has a spectrum that is mostly disreguarded by most production rendering systems, but Matropolis looks different. Let me know what you think.

-Erick

[dantea]

My suggestion is to make Lumicentre flexible in the way it chooses how to render different geometry. Using the same algorithm for everything will suck as either it will render everything at the same low quality or render everything at the same slow speed. Develop your own materials interface that is similar to how shaders work so that the artist can decide on where the computation cost will go when rendering a scene.

How fast are we expecting Lumicentre to be? Is it gonna be slower than photon-mapping?

[CGTalk Moderation]

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.