PDA

View Full Version : displaceApproximation + subdivApproximation on same mesh


dangerweenie
07-05-2009, 03:44 PM
does not seem to work any more in mental ray for maya 2009:

"// Warning: (Mayatomr.Geometry) : pSphereShape1: subdivision surface does not support separate displace approximation, ignored. Tune primary tessellation / approximation for displacement. //"

what the crap? used to be able to use both of these approximation nodes on the same shape in 8.5 ... feature displacement is checked off on the shape node.

anyone know what's up with this?

nfrancisj
07-05-2009, 06:39 PM
might just be an issue with your maya...seems to be fine on mine:p

dangerweenie
07-05-2009, 07:00 PM
what version of maya/mr are you using? what kind of mesh?

might just be an issue with your maya...seems to be fine on mine:p

Fumetsu
07-06-2009, 10:51 AM
It doesn't work on mine either, I am using maya 2009 service pack 1.

The error/warning won't show up unless you look for it in the log.

djx
07-06-2009, 12:57 PM
The warning just means you don't need the displacement approximation if you are using a subdivision approximation. Just tune the value in the subdiv-approx and your displacement should be ok.

-- David

Ash-Man
07-06-2009, 01:39 PM
you should not use BOTh at the same time

this was a very old workflow

nfrancisj
07-06-2009, 07:28 PM
Maya 2009

I think the 2 dudes might be correct. Maybe with the new versions...Autodesk decided that it wasn't nessessary and removed it. :shrug:

I don't have SP1 so I don't know... but I can add both at the same time with no errors. I recommend searching the forums to see if anyone has a good/updated displacement pipeline. I would love to know as well (if you do find something).

I'll see if I can find something tooooo.

thanks...

PS. ALL MAYA GURUS...COME UP WITH A UPDATED DISPLACEMENT PIPELINEEEE PLZZZZ!!! :D

dangerweenie
07-06-2009, 08:19 PM
you should not use BOTh at the same time

this was a very old workflow

This is the workflow laid out in the almost infamous z-brush hammerhead tutorial, and it worked perfectly in maya 8.5. However, I do think you are correct ... I'm posting a link to something one of the moderators @ the mental ray forums stated regarding these two approximation nodes.

dangerweenie
07-06-2009, 08:21 PM
posted here (http://forum.mentalimages.com/showthread.php?t=4670&highlight=displacement+approximation) by Bart about one year ago @ the mymentalray forums (might be regarding mental ray standalone, but I'm pretty sure it carries over to maya for mental ray):

For a subdiv, or a ccmesh, only one approximation is used for doing the displacement and the subdividing of the surface at the same time. If both subdiv and displacement

Over the versions, Maya has shifted which overrides which. I check from version to version by exporting to an mi file.

Note that this is different behavior from a pure surface in mental ray, which does both, first the surface approximation, and then the displacement approximation.

The point is that the newest ccmesh approximation implementations can make adjustments to tesselate overall faster. There are a few tips depending on which versions you are using, such as flags for auto uv matching at texture edges etc. I believe there is another post further down this Maya category which addresses that. Check toward the bottom of this topic: viewtopic.php?f=6&t=1441 (http://forum.lamrug.org/viewtopic.php?f=6&t=1441) approximations are specified, only one is passed to mental ray by Maya.
__________________
Bart
LAmrUG Forum Originator
Director of Training, mental images

dangerweenie
07-06-2009, 08:24 PM
The warning just means you don't need the displacement approximation if you are using a subdivision approximation. Just tune the value in the subdiv-approx and your displacement should be ok.

-- David

how would one go about setting up an approximation that would only add extra tesselation on areas that are displaced? jsut use a displacement approx. node instead of a subd one?

djx
07-06-2009, 11:26 PM
Not sure.
Seems inefficient, so I'd like to know too.
You'd think they changed it for a reason though. :curious:

Your quote from bart seems to suggest that this is something affecting the new ccmesh, so maybe by using the older non-ccmesh you can still use both approx's. I'll have to try to find out I suppose.

-- David

dangerweenie
07-07-2009, 02:12 AM
Not sure.
Seems inefficient, so I'd like to know too.
You'd think they changed it for a reason though. :curious:

yeah, well ... you think they'd at least document that reason ... and good luck calling autodesk, ha!


Your quote from bart seems to suggest that this is something affecting the new ccmesh, so maybe by using the older non-ccmesh you can still use both approx's. I'll have to try to find out I suppose.

-- David

yeah ... i think i am going to hit him up and see what he says, everyone seems to have a different outlook on the correct workflow for the approximation nodes and displacement. you mentioned using an older "ccmesh" ... is there a way to do this natively in maya, without using an older build of it?

also, almost forgot - love your blog ... found your swatch refresh script a lifesaver about a year ago (was sick of being afraid of opening the hypershade haha), and your gamma how-to + script taught me a ton about color. thanks a million, please keep it up!

djx
07-07-2009, 02:58 AM
you mentioned using an older "ccmesh" ... is there a way to do this natively in maya, without using an older build of it?

Yes, you can use this mel to tell your subdiv approx to use the older non-cc method

addAttr -ln "miExportCCMesh" -at bool mentalraySubdivApprox1;

(You replace mentalraySubdivApprox1 with the name of your subdiv approx node).

This used to be a fix for displacement seams and blurry artifacts, but I still have not tested to see if it allows you to use a seperate displacement approx at the same time.

Oh.. thanks for your nice words .. :blush:

-- David

orre
08-26-2009, 11:32 AM
Have anyone of you found a good workflow yet? Would be really interesting to know.

CGTalk Moderation
08-26-2009, 11:32 AM
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.