Results 1 to 11 of 11

Thread: VPR + caustics

  1. #1
    Explorer souzou's Avatar
    Join Date
    Sep 2003
    Location
    London, UK
    Posts
    423

    VPR + caustics

    Hi guys,

    Just wondering if anyone might be able to help with this. I've created this belljar but there's a big disparity between what VPR is showing for the glass surface and my F9 render. Surface is just a dielectric node. The thing is I really prefer how it looks in VPR so any pointers to how I can get that look in my F9 render would be really appreciated!!

    Render settings are MC radiosity 3-bounces with everything on except Directional Rays and Use Bumps.

    Thanks,
    Matt
    Attached Thumbnails Attached Thumbnails Click image for larger version. 

Name:	belljar_vpr.jpg 
Views:	194 
Size:	449.0 KB 
ID:	100666  
    Matt Knott / VERSUS
    http://www.versus.nu

  2. #2
    For sure VPR is not in "Draft Mode"?


  3. #3
    Explorer souzou's Avatar
    Join Date
    Sep 2003
    Location
    London, UK
    Posts
    423
    Hi Marcus, no VPR is not in draft mode.
    Matt Knott / VERSUS
    http://www.versus.nu

  4. #4
    2D is so Flat! oobievision's Avatar
    Join Date
    Dec 2006
    Location
    Boca Raton, Florida, USA
    Posts
    677
    looks as though your final render doesnt have double sided polygons and internal reflections.
    Oobievision

  5. #5
    Explorer souzou's Avatar
    Join Date
    Sep 2003
    Location
    London, UK
    Posts
    423
    Quote Originally Posted by oobievision View Post
    looks as though your final render doesnt have double sided polygons and internal reflections.
    The model has an outer 'shell' (polys facing out) and an inner (polys facing in), no double-sided surfaces. Followed William Vaughan's tut for glass node surfaces: http://www.youtube.com/watch?NR=1&fe...&v=XgH04RMVIQw

    If I turn on double-sided it all goes weird. I can live with the F9 render but just wondered if this was a VPR issue. I've probably posted this in the wrong section anyway - should've put it in support
    Matt Knott / VERSUS
    http://www.versus.nu

  6. #6
    Inner / outer surface (the old air glass layer trick) is old school Williams tut will no doubt be fairly old by now, Dialectic node likes one poly layer with double sided turned on, can't remember weather node plugs directly now or if you still need to use a 'Switch' node with 'Spot Info node's' 'Polygon Side' fed into the the 'Switch' end and 'Dialectic' in the 'True' end and the 'Standard node in the 'False ' end, then have the 'Switch' node plugged in to the Material node

  7. #7
    Explorer souzou's Avatar
    Join Date
    Sep 2003
    Location
    London, UK
    Posts
    423
    Quote Originally Posted by gerry_g View Post
    Inner / outer surface (the old air glass layer trick) is old school Williams tut will no doubt be fairly old by now, Dialectic node likes one poly layer with double sided turned on, can't remember weather node plugs directly now or if you still need to use a 'Switch' node with 'Spot Info node's' 'Polygon Side' fed into the the 'Switch' end and 'Dialectic' in the 'True' end and the 'Standard node in the 'False ' end, then have the 'Switch' node plugged in to the Material node
    Hmm but then how do you see/set the thickness of the glass without the model having thickness? Bearing in mind this is a belljar so not solid glass but with air inside it (if that makes sense), much like a glass tumbler. I tried a second model with one poly layer and double-sided turned on and it goes weird (see render - double-sided one-poly layer on the right).

    Thanks for all your replies btw
    Attached Thumbnails Attached Thumbnails Click image for larger version. 

Name:	belljar_singlevdoublesided.jpg 
Views:	84 
Size:	113.3 KB 
ID:	100716  
    Matt Knott / VERSUS
    http://www.versus.nu

  8. #8
    Registered User zardoz's Avatar
    Join Date
    Dec 2003
    Location
    Portugal
    Posts
    1,819
    it looks like the glass doesn't have thickness. how tick is your glass? dielectric uses real world units so maybe for better results you should model that glass with something like 5 or 6 mm.
    Intel Core i7-4770 @3.40GHz,GeForce GTX 660 - Ram 8Gb - Win 8.1 64b

  9. #9
    I was on about the air glass trick where you had a flipped set of polygons inside of the volume you intended to be a glass surface to make refraction work, post nodes and dialectic you no longer need to do this but you still need volume, an inside and outside surface with distance just like real glass. See pic for proof.




    http://dl.dropbox.com/u/30264782/Scr...2023.12.44.png

  10. #10
    Explorer souzou's Avatar
    Join Date
    Sep 2003
    Location
    London, UK
    Posts
    423
    heh I modelled it (more-or-less) real size so it is 6mm thick. Model screengrab attached (grid size there is 10mm).
    Attached Thumbnails Attached Thumbnails Click image for larger version. 

Name:	belljar_model.jpg 
Views:	62 
Size:	446.3 KB 
ID:	100717  
    Matt Knott / VERSUS
    http://www.versus.nu

  11. #11
    Explorer souzou's Avatar
    Join Date
    Sep 2003
    Location
    London, UK
    Posts
    423
    Quote Originally Posted by gerry_g View Post
    I was on about the air glass trick where you had a flipped set of polygons inside of the volume you intended to be a glass surface to make refraction work, post nodes and dialectic you no longer need to do this but you still need volume, an inside and outside surface with distance just like real glass. See pic for proof.
    Ah ok, no I hadn't done the air trick, this is just a simple volume.

    Thanks for the node setup. I duplicated it and here are three versions. One is with my original node, another with your setup and a third with your setup but with double-sided and exclude from vstack unticked. This ends up being an identical F9 render to my original setup, although it displays differently in VPR. I'm not sure what I've learned here, other than VPR lies!

    But either way I think I'm going to live with it as-is because the difference is marginal, and the clock is ticking...

    Thanks again!
    Attached Thumbnails Attached Thumbnails Click image for larger version. 

Name:	belljar_originalnodes.jpg 
Views:	72 
Size:	482.3 KB 
ID:	100719   Click image for larger version. 

Name:	belljar_gerrynodes.jpg 
Views:	76 
Size:	483.7 KB 
ID:	100720   Click image for larger version. 

Name:	belljar_gerrynodes_singlesided.jpg 
Views:	79 
Size:	484.6 KB 
ID:	100721  
    Matt Knott / VERSUS
    http://www.versus.nu

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •