Occlusion shader just to visualise geometry (mantra and renderman)

[Occlusion shader just to visualise geometry (mantra and renderman)]

I see some nice greyscale renders that use an occlusion shader, is there one for both mantra and renderman?

[Ambient Occlusion map in a principled shader]

On 27/07/2020 at 1:57 PM, dleonhardt said:

You multiply Ambient Occlusion over your Base Color.

Thank you, I know I have to multiply, but how and where?

[Ambient Occlusion map in a principled shader]

A beginner question, I just can't find how to use the AO texture map with the Principled shader, there are plenty of "use texture" for everything, but not for AO.

How do I use AO?

[scale towards the center of the geometry]

I need to scale down a geometry, but it has to be towards the center of the geometry.

[Python/VEX: create and assign material to each existing shop_materialpath]

10 hours ago, sadhu said:

Check this file

group_material.hipnc

I'm a bit lost here, I meant for each existing shop_materialpath, I do not need to create any shop_materialpath, I need to create groups and materials.

I have a 3D model pack that has some 30 shop_materialpaths, I just need each one of them to become a group and have a material in the mat context with the same name.

the shop_materialpath is a starting point, it's an input that we already have.

[Python/VEX: create and assign material to each existing shop_materialpath]

For each existing shop_materialpath (there are a lot of them):

- create a group, with the same name as the shop_materialpath

- create and assign a material with the same name as the shop_materialpath

[beginner question: vellum ropes]

I'm trying to simulate this vellum rope for the first time, everything falls including the tube that's supposed to suspend the rope. I'm new to vellum, and I'm sure this is a very simple thing, I've searched but didn't find something similar. How do I tell vellum that the tube shouldn't fall as well?

[transform to centroid then invert transform in a for each loop]

I got a little problem here. I've used transform with -$CEX -$CEY -$CEZ. Then I created a reference copy, and checked invert transformation, but it's not working. It's throwing an error

Unable to evaluate expression (Bad data type for function or operation (/obj/geo1/transform1/tx)). 
Unable to evaluate expression (Bad data type for function or operation (/obj/geo1/transform1/ty)). 
Unable to evaluate expression (Bad data type for function or operation (/obj/geo1/transform1/tz)). 
 

Is there a way to invert a transform to centroid? telling in for each connected primitive: go to the centroid, come back again to where you were.

[get rid of small islands of geometry and keep the bigger shapes]

How do I get rid of these islands of geometries inside these bigger shapes? I can do it manually but I have a lot more than 2. I want to keep only the bigger shapes.

[Delete unused groups]

I have a lot of geometry but only want to use part of it. I have selected what I want, delete non selected, but I'm left with all the groups from all other geometry. Can I keep only the groups that are in the current geometry?

[VEX: find the max of the max of bounding boxes of connected pieces]

On 25/06/2020 at 8:26 PM, tamagochy said:

Put to point wrangle 

int pts[] = neighbours(0,@ptnum);
float dists[];

foreach(int pt; pts){
    vector pos = point(0,"P",pt);
    vector dir = pos-@P;
    float dist = length(dir);
    append(dists,dist);
    
}
dists = sort(dists);
float d = dists[-1]; //longest 
f@d=d;
setdetailattrib(0,'dist',d,'max');

in first input connect all bounding boxes

Thank you, I'm almost there, I get an attribute d which has 7,9 and 21, I only need the 21, the highest number, one value, what do I need to add to get that?

[create mask for vdb to limit noise area in volume vop]

I have an vdb (sdf) to which I apply some noise in a volume vop. The noise is applied everywhere. I need it to be limited to some areas, so I need to create a mask. The mask can be driven by a noise.

Is it possible to create a mask for a vdb?

P.S. the vdb was a polygon geometry, would it be an option (and how) to use a mask from polygon into vdb? whichever way exists and works 

[VEX: find the max of the max of bounding boxes of connected pieces]

For each connected piece, get the bounding box, find the axis with the maximum scale and remember just the number. Say e.g. a box is 18 by 3 by 2, a grid is 4 by 7, a sphere is 15 by 15 by 15 then we need 18, 15, and 7.

Then find the max of these maxs (18 vs 7 vs 15) and assign it to an attribute, such that we end up f@maxattrib = 18.

[Give 2 names to a primitive]

I need to primitive to remember it belongs to a first set of primitives, and to remember it belongs to a second set of primitives.

If I call name1, the set with name one shows up. If I call name2 the set with name2 shows up. Each primitive should have two names.

 

[what could be causing distorted polygons in road tool?]

The two lines intersecting are not from the same geometry, I've merged them and fused them, some intersections look ok but a lot of the others are messed up like these two examples.

Do you know what could be causing this?

[smoke has a hairy look]

I have a very simple shader and here are also the mantra settings:

[smoke has a hairy look]

I've keyframed it yesterday several times, the problem is that the newly emitted smoke needs different settings than the older one that's been around for a while, and this is even in later frames because there is still new smoke being emitted, any change to satisfy the old one upsets the new, any that satisfies the new upsets the old.

[smoke has a hairy look]

Done simulation 18 times, the problem is still there: the smoke is hairy. I've used the control field with speed, tweaked gasdisturb settings, at different block sizes, different strength, tweaked temperature diffusion, the smoke is still hairy. I've watched countless courses / tutorials, I've even replicated their exact settings, the smoke is still hairy. If I bring down the strength of the gasdisturb the smoke is too smooth in the first 100 frames.

There is something else, I hope someone knows what's going on.

After 18 times, this is what the smoke looks like:

[rbd packed object reacts too late to the collider]

A sphere and a box, and rbd packed object and a collider, but the collider sphere goes very very fast, it goes in one direction then back again. The problem is that the box (rbd packed object) realises there is a collider only when the sphere is on its way back. Result: the box goes in the opposite direction than where it should go.

Any idea what puts that box to sleep when the sphere is crossing fast?

Here is a flipbook

 

slow_reaction.mp4

[why do small fractured wood pieces keep dancing on the ground?]

the collision padding is the cure, thank you.

I had to bring it down from the default 0,02 to 0,00001

[why do small fractured wood pieces keep dancing on the ground?]

I've changed the contact breaking threshold in both directions drastically but they keep dancing.

I couldn't find the collision padding, I'm using the bulletrbdsolver

[why do small fractured wood pieces keep dancing on the ground?]

The small fractured wood pieces are annoyingly agitated, they keep active all the time making the whole thing look unrealistic. How do you calm these things down?

here is a flipbook of their dance:

 

dancing_pieces.mp4

[smoke has a hairy look]

I have changed the control field of all three gasdisturbs to use the speed of the velocity, but for some reason the smoke loses its shape at the end (new hipfile included)

hairy_smoke.hiplc

[smoke has a hairy look]

Here is the hipfile, I kept changing the parameter and tweaking but the result is still the same. I don't know how to use the speed of the velocity field as a control field as user Skybar mentioned above.

hairy_smoke.hiplc

[smoke has a hairy look]

I would truly appreciate if you could tell me how I can do that. I do see that velocity is already the disturb field in bindings, do I need to so something else to add disturbance only where the smoke is moving faster?