WytRaven's Robot WIP

Get feedback from others on your works in progress
Deus
Posts: 336
Joined: Sun Feb 04, 2007 3:47 am

Post by Deus » Wed Dec 24, 2008 2:15 pm

Wyt really impressive work and I admire you for doing things the hard way...

But since this is a Deus post there will always be a "but". Wouldn't boolean operations and mesh simplification give you the same result? (perhaps even better)

User avatar
pixie
Indigo 100
Posts: 2332
Joined: Sat Dec 29, 2007 4:54 am
Location: Away from paradise
3D Software: Cinema 4D
Contact:

Post by pixie » Wed Dec 24, 2008 2:34 pm

Deus wrote:Wyt really impressive work and I admire you for doing things the hard way...

But since this is a Deus post there will always be a "but". Wouldn't boolean operations and mesh simplification give you the same result? (perhaps even better)
OMG!! He already talks of himself in the third person! :shock: :shock:



:D

User avatar
WytRaven
Indigo 100
Posts: 905
Joined: Mon Aug 27, 2007 8:24 pm
Location: Dubbo, Australia
Contact:

Post by WytRaven » Fri Dec 26, 2008 11:56 am

Deus wrote:Wouldn't boolean operations and mesh simplification give you the same result? (perhaps even better)
Booleans create unnecessarily complicated meshes. They may look pretty when in solid mode and may render just fine but when it comes to fine tuning, the mess booleans make just causes headaches.

It's not easy to explain in words but l can guarantee that building this same mesh using booleans would result in a significantly "heavier" mesh from a poly count perspective.

Complete control over the polys also makes UV mapping less of a nightmare too ;)
:idea: "A foolish consistency is the hobgoblin of little minds..." - Emerson 1841

User avatar
WytRaven
Indigo 100
Posts: 905
Joined: Mon Aug 27, 2007 8:24 pm
Location: Dubbo, Australia
Contact:

Odd Artifact

Post by WytRaven » Fri Jan 09, 2009 10:47 am

Anyone got any idea what is causing this odd artifact under Indigo 1.1.15?

At first I thought it was just due to changes in the way indigo renders since the last time I did any rendering of these objects (back on version 0.9 i think...) but after about 48 hours and no change to the black spots I'm starting to wonder if this is some sort of bug...

The material on the effected parts is aluminium nk data modified by a bump map which simulates very fine mold dimples...that's it...nothing "special".
Attachments
OddArtifact.png
OddArtifact.png (1.57 MiB) Viewed 3037 times
:idea: "A foolish consistency is the hobgoblin of little minds..." - Emerson 1841

User avatar
Kram1032
Posts: 6649
Joined: Tue Jan 23, 2007 3:55 am
Location: Austria near Vienna

Post by Kram1032 » Fri Jan 09, 2009 11:17 am

It's a known bug... something with the KD-tree... simply set the now missnomed BIH-threshold (or is it already the BVH-threshold? :? ) to 0 to fix it. Poorly, the render will be a bit slower, but not as much slower as with old BIH :)

User avatar
WytRaven
Indigo 100
Posts: 905
Joined: Mon Aug 27, 2007 8:24 pm
Location: Dubbo, Australia
Contact:

Post by WytRaven » Fri Jan 09, 2009 12:16 pm

Thanks Kram. I'll give it a shot.
:idea: "A foolish consistency is the hobgoblin of little minds..." - Emerson 1841

User avatar
WytRaven
Indigo 100
Posts: 905
Joined: Mon Aug 27, 2007 8:24 pm
Location: Dubbo, Australia
Contact:

Post by WytRaven » Fri Jan 09, 2009 7:34 pm

Ok Kram I set bih_tri_threshold to 0...it's not a bit slower; it's more like half the speed. msps went from 26 to 56 :shock:

...but it does fix the problem :?
:idea: "A foolish consistency is the hobgoblin of little minds..." - Emerson 1841

User avatar
CTZn
Posts: 7240
Joined: Thu Nov 16, 2006 4:34 pm
Location: Paris, France

Post by CTZn » Sat Jan 10, 2009 12:48 am

Come on, you are a tough guy ;)

I'm wondering if that bih issue is not just because the affected faces are coplanar with bounding boxes sides, see what I mean ? Try to give a small rotation amount to the whole scene with bih, and see, coz it's an untested fix (if any) :oops:

dmn
3rd Place Winner
Posts: 243
Joined: Thu Dec 21, 2006 6:42 pm

Post by dmn » Sat Jan 10, 2009 5:38 am

This is sure to be wrong, but did you try without the displacement map, or lowering the adaptive subdivision on the mesh?

EDIT: Oh, you said bump map. Try turning that off?

User avatar
PureSpider
Posts: 1459
Joined: Tue Apr 08, 2008 9:37 am
Location: Karlsruhe, BW, Germany
Contact:

Post by PureSpider » Sat Jan 10, 2009 5:52 am

Its 100% not related to bump or whatever.
This is a KD-Tree error...

User avatar
Zom-B
1st Place 100
Posts: 4700
Joined: Tue Jul 04, 2006 4:18 pm
Location: ´'`\_(ò_Ó)_/´'`
Contact:

Post by Zom-B » Sat Jan 10, 2009 6:17 am

Herea bugreporting thread for this issue!

Anyway, your speed drop down from Kd-tree to BVH is realy creapy!!1!
The slowdown should be minimal :?

How much polys?
polygonmanufaktur.de

User avatar
WytRaven
Indigo 100
Posts: 905
Joined: Mon Aug 27, 2007 8:24 pm
Location: Dubbo, Australia
Contact:

Post by WytRaven » Sat Jan 10, 2009 11:29 am

CTZn: I'll try rotating and see what happens. Won't this just move the problem though? I mean stop it occurring in those places but cause it to appear in others?

dmn: I can't turn off the bump or the object will look completely unrealistic :?

PureSpider: That is encouraging due to the above response to dmn.

ZomB: Creepy and for me painful...I've gone down to 10 samples/pixel/hour! :shock: The only change was switching bih_tri_threshold from about 50000 (or whatever the blendigo default is) down to 0 :(
:idea: "A foolish consistency is the hobgoblin of little minds..." - Emerson 1841

User avatar
WytRaven
Indigo 100
Posts: 905
Joined: Mon Aug 27, 2007 8:24 pm
Location: Dubbo, Australia
Contact:

Post by WytRaven » Sat Jan 10, 2009 11:31 am

Zomb: Poly count (blender quads at least) is 276,688.
:idea: "A foolish consistency is the hobgoblin of little minds..." - Emerson 1841

User avatar
CTZn
Posts: 7240
Joined: Thu Nov 16, 2006 4:34 pm
Location: Paris, France

Post by CTZn » Sun Jan 11, 2009 12:21 am

WytRaven wrote:CTZn: I'll try rotating and see what happens. Won't this just move the problem though? I mean stop it occurring in those places but cause it to appear in others?
Forget what I said, the issue is related with bih and the geometry, even while the later is of clean/simple topology. Modifying the geometry would just move the issue around the mesh. This is a severe bih bug wich I could work around only by redoing the geometry, and I don't see this as an option in your case, WytRaven :/

User avatar
WytRaven
Indigo 100
Posts: 905
Joined: Mon Aug 27, 2007 8:24 pm
Location: Dubbo, Australia
Contact:

Post by WytRaven » Sun Jan 11, 2009 12:26 am

Thx CTZn :)

I've just bitten the bullet and am rendering with bih_tri_threshold 0. 6h 16m - 106spp :cry:

Estimated time to arrival (at my desired 10000spp) aproximately 24 days :shock: ...this is insane :x Ono help me! ;)
:idea: "A foolish consistency is the hobgoblin of little minds..." - Emerson 1841

Post Reply
108 posts

Who is online

Users browsing this forum: No registered users and 48 guests