Can c4d fake GI?

[verti]

actually if you tweak your radiosity settings, the redering does not take so long. besides that, it will be always slower than raytracing. sometimes it's good to use "hybrid"... that is i.e. set the strengh of radiosity to around 50%. it depends on what kind of scene you want to render...

v.

[neilyb]

You could download the plugin 50%Grey which assigns lights to points on a poly object, the more points the more lights. The results can be OK but you need to play with it a little.

Don't know where I found it though?

[say-g]

whoa it only took 5 mins for a reply!

thanks verti, ill do some tweakin of settings

ill also have a look for this 50%grey plugin.
do you still have it neilyb ? could u post an example of its style ?

thanks again


-sage

[neilyb]

I'll have a look and post something when time allows!

[AdamT]

I believe the best plugin for fake radiosity is "softRadiosity", which you can check out here: http://www.c4d-plugins.com/
I haven't tried it, but the gallery images look pretty good.

But before giving up on real radiosity I'd definitely suggest that you do plenty of experimentation--if you haven't done so already. Just as an example, there was a recent thread on Postforum from someone who was getting awful artifacts even though his renders were taking *days*. After troubleshooting the scene and fine-tuning the radiosity settings he was able to get a *very* nice render in about two hours. BTW, the scene had almost 1/2 million polys and over 10 light sources, including 5 area lights.

[neilyb]

Heres the 50% example using a skydome, took 20 secs to render on a P3.
You can use a map in the luminance channel and the plugin will change the colors of the lights to match the luminance.....although it don't work for me!

[verti]

actually... we've bought softRadiosity for our studio just few months before xl 7 release. yeah, it's great radiosity FAKE... but if you got a good eye you will notice it's just a FAKE. in my opinion you should just experiment with the xl 7 built in (real) radiosity.

check out those settings:

strengh: 90%
accuracy: 75%
prepass: 1/2
diffuse depth: 3
stochastic samples: 120
min res: 20
max res: 60

i got not bad results redering interiors with this settings...

remember... it all depends on the scene you want to render... maybe you can post some pics of your raytraced scene and we/i can help with the setup?

cheers,
verti.

[LucentDreams]

Its pretty easy to fake GI in anyap with softshadows, I remeber doing it all the time in XL 6 before I got 7 with radiosity, and theone advantage was speed, but you can almost always pinpoint a faked scene, usually because of the lack of cproper colour bleed.

[say-g]

i have no scene in particular to show, i was just doing experiments, and i did a render of a simple 2x4 type block cloned into a spiral staircase with a floor that took 45mins an lookedn o better than the render without radiosity...

thanks for all ur help, judging by the screenies of 50% (unless the box is elevated above the floor) i think ill stick with real radiosity, ill try out those settings verti.

ps: thanks for taking the time to make those examples neilyb

-sage

[kiwi]

Check out MJV brilliant tute on radiosity as well Sage.....{link to his site in the thread}



http://www.cgtalk.com/showthread.php?s=&threadid=18116



Stu.

[LucentDreams]

Originally posted by say-g
i have no scene in particular to show, i was just doing experiments, and i did a render of a simple 2x4 type block cloned into a spiral staircase with a floor that took 45mins an lookedn o better than the render without radiosity...

thanks for all ur help, judging by the screenies of 50% (unless the box is elevated above the floor) i think ill stick with real radiosity, ill try out those settings verti.

ps: thanks for taking the time to make those examples neilyb

-sage

do you still have thaat scene, 45 minutes sounds rediculously long for such a smple scene, let me have alook at it and see what your doing wrong, though if MV would look at it that would be even better, he has Radiosity down to a true art.

[neilyb]

I agree, if a simple scene takes so long there must be something up.......BUT you don't say what machine/setup you are using?

[LucentDreams]

I'm right on your tail for responding it seams :)

I never considered what the machine would be like, but even a PIII 450 shouldn't take that long on such a smple sounding scene.

[neilyb]

Yeh but a PIII with too little RAM would struggle....It is a little annoying that minimum specs take only the running of the prog into account, not what is really necessary to render a scene...!

The sooner we know the spec the sooner we can say....AHAAHAA....!

[say-g]

the spiral stairs scene also had env in it, so that could have slowed it down, i did use default settings for that scene as it was my 1st experiment...

my machine is a AMD 1.4ghz, 256ddr ram, only wish cin4d would use my gf3 to render D:

that tutorials helps abit (i just found out it had a 2nd page keke)
thanks guys


-sage

[LucentDreams]

Never ever use the default settings, they are the worse default settings ever, pretty useless for anyscene. Thats your problem.

[AdamT]

Yeah, the defaults aren't great. But the real difficulty with radiosity is that it's hard to predict which settings will produce the best results in the least amount of time. It just takes a bunch of experimenting. That said, I highly recommend Arndt Von Koenigsmark's XL7 book. The whole thing is great, but he has an excellent section on radiosity and a very good analytical approach to arriving at the best settings. I give it :) :) :) :)

[amapimaster]

hmm, this book sounds interesting.. Where can I get it? Whats it called?

Cheers,

[AdamT]

Read all about it!

[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.