PDA

View Full Version : can't get it to work



OOZZEE
05-08-2007, 08:16 AM
hi, can someone post a screen shot of a Dielectric Node on a simple mesh where you actually get results... What ever I am trying, the surfacing is working half the time....sometimes not showing up in the node window or sometimes it is showing... however never been able to render it or view it in Fprime... I'm thinking I should maybe change the graphics card ? but first want some confirmation that it's not a bug.

I have simply plugged the dielectric into the mesh via material to material / up'd the absorbtion to 5, refraction to 3, color to orange and it goes black on me.... or orange but no show in the F9 ?:help:

stevecullum
05-08-2007, 08:59 AM
This is for glass right?

Make sure you have double sided on and raytrace transparency etc..

stevecullum
05-08-2007, 09:02 AM
nodes are on the venus statue

OOZZEE
05-08-2007, 09:51 AM
thanks Steve, it's for glass and I did have both of the options you mentionned checked...

you see, your surface node result is also black....but sometimes it changes to the color I chose but sometimes stays black even if I chose a color....

Do you really need all the extra switch, spot, and standard nodes to make the dielectric work ? - I only plugged it directly into the surface node.

thanks for your help.

stevecullum
05-08-2007, 10:31 AM
You do need the extra switch for the opposite side of the polys, otherwise the refraction will be all wrong (Glass is 1.5). Its just instead of 'air polys' which was the old way.

There are numerous threads on glass shaders, search some out they will prove useful ;)

hrgiger
05-08-2007, 10:56 AM
Well that's interesting, I didn't know that either. Going to have to read up on it some more.

nolan
05-08-2007, 11:54 AM
Steve . . . Not to get off topic but, did you use image maps or layers or nodes for the marble texture in the venus statue render?

Thanks & great work.

Nolan

cagey5
05-08-2007, 11:59 AM
thanks Steve, it's for glass and I did have both of the options you mentionned checked...

you see, your surface node result is also black....but sometimes it changes to the color I chose but sometimes stays black even if I chose a color....

Do you really need all the extra switch, spot, and standard nodes to make the dielectric work ? - I only plugged it directly into the surface node.

thanks for your help.

I reported this many moons ago as I can't get the ball to register anything less than totally black with typical absorption values. If you use relatively low values of 1 - 10 absorption you'll get a nicely coloured ball, anything
greater and it'll be pretty much black, or at least it is for me, though no one else seems to have the problem.
With absorption, the greater the object size, the lower the absorption number you need so I would expect this to be reflected in the sample surface when the sphere size is altered, but it shows the same no matter what size is entered.

stevecullum
05-08-2007, 12:08 PM
Steve . . . Not to get off topic but, did you use image maps or layers or nodes for the marble texture in the venus statue render?

I used the Delta node. Its pretty simple setup really.

OOZZEE
05-08-2007, 12:24 PM
cagey5 - I have the problem even if values are NOT changed...it will show me color sometimes and other times black... and when I finally get color, it doesn't render at all - it's invisible... Are you at least getting a render when you see the colored ball ( surface node )

and what you are saying is the size of the mesh has someting to do with it ?

zatara
05-08-2007, 12:48 PM
Thesize of the objects it is very important. An object with two meters, and the same with 2 cm, will have diferent absorption.
It is like real word. Think of A giant glass, and a small glass.

cagey5
05-08-2007, 12:49 PM
I don't have any problem getting a render it's just that it would be better if I could get an accurate representation in the editor before I got there, rather than have a dense black ball most of the time.
Re. mesh size. Yes it does effect the final render. If you have a very small or thin piece of glass then you will need a relatively high absorption figure [50 - 200] to see some colour in the final render, othwise it may look completely transparent especially if you have the refraction set at 1.0. Likewise a huge sphere will need a much smaller absorption number [0.5 - 10] to show a similar effect.

OOZZEE
05-08-2007, 01:55 PM
ok - becoming more clear now... (subtle joke hehe ).. I also found this on SpinQuad which might answer my F9 prob.

http://www.spinquad.com/forums/showpost.php?p=187851&postcount=11

thanks folks. - I will attempt when I get home later.

OOZZEE
05-10-2007, 02:23 PM
ok attempted - getting better results....BUT not quite -gasp - what I would really like to see is how the heck this was done.

http://www.newtek.com/lightwave/gallery/main.php?g2_itemId=5212

oh if I could get my hands on that node setup. that reddish statue head is really nice and I want to try capture similar effect.

gotta keep trying !

Amurrell
05-10-2007, 05:35 PM
It looks like the same node setup except that the dielectric node is a red color and the absorption is turned up so that most of the light passing throught the object proper is absorbed and around the edges where the material is thinner, light is able to pass through more readily.

In reality, after you have the standard node set up and plugged into the false of the switch and the spot info node tied into the switch as well, then all you have to do is mess with the dielectric settings to get a wide variety of results.

cagey5
05-10-2007, 10:54 PM
I'd agree with Amurrell. The attached was loaded into the default layout scene, a typical dielectric node tree added to the surface with an absorption of 4. One large, bright poly added to reflect and that was it No additional lights other than the default distant light and no radiosiity.
Nowhere near as good as the link but not a bad jump off point for a 1 minute set-up.

v1u1ant
06-18-2007, 07:40 AM
Wow, so hold on.....are you saying if im using dielectrics i dont need air polys?

cagey5
06-18-2007, 08:33 AM
Not a separate air poly as such, but you do need double sided on with a switch node to determine which side is which so the end result is very similar to having the traditional air polys.