smoke has a hairy look

[anicg]

Any idea what could be behind the hairy look of the smoke?

When the smoke has just been emitted it looks detailed but overtime it turns into .. a hairy look. I've placed a light behind it so I can see the edges clearly. It has several gasdisturb nodes, no dissipation and no evaporation (both set to 0).

Actually even in the beginning (image 2) there is that hairy look around the edges, so I do not even know if it has to do with age.

 

Edited June 1, 2020 by anicg

[Skybar]

Too much disturbance probably.

[anicg]

I've used 3, here are the attachments, I can keep tweaking but maybe you have an idea from your experience.

 

[dleonhardt]

Those disturbance values seem too high for smoke of your scale. I'd personally try to dial them down to the 1-2 strength range. And maybe set the Gas Disturb nodes to disturb your velocity field instead in the Bindings tab, because I feel like this tends to produce more interesting results as the simulation goes on.

[anicg]

I've reduced all three gasdisturbs to strength of 1, and the bindings have vel by default as disturb field, but the results still look hairy, like cotton.

Here are 4 renders from 4 different angles (front, back, right, left).

As I mentioned I have dissipation and evaporation at 0.

Do you have any idea why this still looks like cotton / has hairy look

 

[Whatsinaname]

Use some dissipation, I think that might help a lot. 

[Skybar]

I think the main problem is that you are adding disturbance everywhere all the time. A popular method is to use the speed of the velocity field as a control field for example, so you scale the disturbance amount with how fast the smoke is moving.

[anicg]

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?

[fencer]

3 hours ago, anicg said:

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?

Share a hip file.

[anicg]

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

[fencer]

Check help about Control Field. Right now your Disturb and Turbulence doesn't have influence limit, they apply forces everywhere where density is. Use Control field to limited this, for example by temperature or density.
https://www.sidefx.com/docs/houdini/nodes/dop/gasdisturb.html

Create an explosion or billowy smoke from the shelf and examine what inside.

Edited June 4, 2020 by fencer

[anicg]

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

[fencer]

Good, right step, you are close.

Use Visualization in gasdisturb to see where forces are still present. Play only with gasdisturb1 first, disable the rest, adjust one than add one more, adjust and so on.

Play with control range, you are adding high values of gasdisturb but trying to limitate only from range 0-1, try 0-5, 0.1-2 or so...

Plus now limit only by velocity, try to add density as well.

 

[anicg]

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:

[mestela]

That's overcranked disturbance. You could try keyframing it down over the first 30 frames.

 

 

[anicg]

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.

Edited June 7, 2020 by anicg

[fencer]

Again, check shelf tool "billowy smoke".
You are using very low gas dissipate, smoke almost stay forever, at the same time there are no forces to move this smoke except gasdisturb and turbulence which tearing it apart.
Main motion of smoke is Buoyancy (which based on temperature) or wind, in your case temperature disappears very fast, Buoyancy stop working, smoke not rising anymore but gasdisturb keep working.
Find a good reference and examine real behaviour.

[ziconic]

How are you rendering the smoke volume? Is your shader doing any displacement or uprezzing of any kind? I could be mistaken but your renders seem to show a level of detail higher than what you would expect to get from the voxel res in your sim....

[HM_2020]

What are your noise setting in the shader?  Your new 18 sims, renders look as hairy as the old ones....looks like a mad overblown noise in the shader.

Its kinda cool for some soft stuff. 

[anicg]

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

Edited June 8, 2020 by anicg