The Speed Of Fusion
Posted 16 August 2011 - 12:49 PM
just a quick note on how fast you can do graphical 3D stuff if time is limited:
All model & pixels were done within fusion... in about 3 hours (for the graphical side) of production and redenring times.
Of course nothing special in there - but those are one of the things what fusion do best if time is your enemy
So wouldn't it be nice if eyeon puts a lot more energie into the 3D Render/Model System as many users requested...
The 3DText Tool could be greatly improved with a real Curved Bevel, Subdivisons on X,Y,Z for cool displace effects, etc, etc...
And if this has been done it should be easy for eyeon to take any user polymask(s) which could converted to a 3DShape with the same features like in the above tool.
This would give the Artist some nice more complex 3D Model building functionallity without the need for a 3rd party app.
p.s. An Open CL Raytracing Render would be Top Notch !!!
Posted 16 August 2011 - 01:32 PM
Nice use of the Fusion 3d text and shaders BTW.
It would be nice if you could upload a comp to the compshop section in vfxpedia for everyone to learn your tricks.
Cool stuff Marcus.
Posted 16 August 2011 - 02:35 PM
Yeah, I would also love a little example comp for the glass look shaders. The only thing I've tried so far with materials was the dragon model from the other thread and I found it a little bit frustrating to achieve something that looked cool without being able to rely on raytracing.
Posted 18 August 2011 - 01:41 AM
Awesome work! Congrats!
Definitely worth an "I am Fusion" tutorial, I'd say.....;-)
Keep on Fusioning.
Posted 18 August 2011 - 11:53 AM
Posted 18 August 2011 - 07:11 PM
sorry for the delay - no problem i will do a tut/flow within the next 2 weeks....
@BLADE - the entire project was done on a GTX 580 PHANTOM (3MB VRAM)
Posted 18 August 2011 - 09:17 PM
not as detailed as your but i think i got the cylinder lights going.
incidentally i did it software render with depth blur and fog node cause the old 3700 is just a bit too slow
gave u credit in the notes
Posted 19 August 2011 - 09:16 AM
Posted 19 August 2011 - 10:28 AM
well it is just a 'instance camera' flip(Y-axis) and reprojected with the 'orginal camera' onto the floor (CATCHER Material with some bool operations)
CODE (REAL SHOT)
CODE (FUSION MODEL OGL SHOT)
Posted 19 August 2011 - 10:34 AM
CODE (FUSION MODEL OGL SHOT)
- alexudell likes this
Posted 19 August 2011 - 11:26 AM
Posted 19 August 2011 - 02:20 PM
Either way, very inventive tricks!
Posted 21 August 2011 - 12:19 AM
Posted 30 August 2011 - 06:34 PM
Your grasp of how to put all this together is just light years ahead of mine, so I'm really looking forward to this!!!
Being technically-challenged, I'm still trying to decipher exactly what you mean by
"In my example i used a INST. CAM -> goes to a 3Dtransform (Y *-1) + Offset (depense on the floors Y-axis). The Floor gets invisible by a 3Doverride -> the Render output goes to the MAIN CAM as a projector onto the 3Dcatcher - and vo la !"
So, any comps you could include showing the different ways of getting reflections will be very much appreciated!!
0 user(s) are reading this topic
0 members, 0 guests, 0 anonymous users