Strange displacement behaviour

Post Reply
Zybrand
Posts: 11
Joined: Tue Aug 07, 2012 9:04 pm

Strange displacement behaviour

Post by Zybrand » Thu Aug 29, 2013 9:28 pm

Hi Guys,

I have a simple scene where a logo breaks through a ocean surface and I'm using Realflow 2013 and hybrido 2. The simulation looks great but the mesh is giving some weird results which I don't understand.

See the image attached. The first block shows the mesh with splash attenuation, the displacement looks fine till the logo starts interacting with the surface. The I get strange pattern appearing it the core parts of the surface but the splash part of the mesh looks the best this way.
Block 2 shows the same mesh build without splash attenuation, it fixes the problem for the most part but the splashes are worse of this way and I lose detain in the splashes.
Also with this method 2 frame glitch out completely as shown in block 4. 2 frames out of 124 did that and the were 30 frames apart. I also tested that on a different machine with the same result.

Block 3 and 5 show a different project of mine where I am running into the same problem. I've gotten the same noisy patterns as in block 3 on some of my Realfow 2012 meshes before. And when that action happens in block 5 things just get worse.

Any idea what is causing this or where I am going wrong?

thanks,
Zybrand
You do not have the required permissions to view the files attached to this post.

User avatar
atena
Site Admin
Posts: 112
Joined: Fri Jun 04, 2010 1:25 pm

Re: Straight displacement behaviour

Post by atena » Fri Aug 30, 2013 9:15 am

Hello, the best way to solve that issue is to have a look at your scene. It is possible at all to get your scene? We can simulate here and then try to reproduce the meshing problems you have reported.

In the other hand to no stop your work I want to recommend you as a workaround to have a look at the simulation graphs we provide in the "Help->Demo Scenes" menu, in this particular case, the one in "graphs" named "apply_stat_spectrum_scale_base_on_speed". This simulation graphs allows you to apply a displacement to a sequence of meshes and attenuate the effect of displacement based on velocity. So in your scene the splash areas probably are above some speed threshold and those areas won't be affected by displacement, which is actually what the "Speed attenuation" parameter does.

I hope you can send us your scene and any doubts about the simulation graph I pointed you at please let us know.
Angel Tena
Head of RealFlow Technology
Next Limit Technologies

Zybrand
Posts: 11
Joined: Tue Aug 07, 2012 9:04 pm

Re: Strange displacement behaviour

Post by Zybrand » Fri Aug 30, 2013 10:16 am

Hi Angel, Ok cool thanks I can sent you the scene to have a look at. How should I send it to you?

In the meanwhile I will have a look at that demo scene you are talking about as well.

I've done some more test and made some progress on the one issue. The bit I marked in red that I said I found in 2012 as well seems to be because my polygon size or radius is to small, when I removed the auto size and radius and made it the same as the cell size it fixed that problem. I am still getting the other mesh artifacts though.

thanks,
Zybrand

User avatar
atena
Site Admin
Posts: 112
Joined: Fri Jun 04, 2010 1:25 pm

Re: Strange displacement behaviour

Post by atena » Fri Aug 30, 2013 11:24 am

We need the FLW file and the geometry files. Do you have dropbox or any other file sharing service where you can put the files so we can download them?.

Other way is to send an email to techdeskrf@nextlimit.com, but this will take longer as the email will be filtered by my colleagues before it reaches me.

Again, thank you.
Angel Tena
Head of RealFlow Technology
Next Limit Technologies

Zybrand
Posts: 11
Joined: Tue Aug 07, 2012 9:04 pm

Re: Strange displacement behaviour

Post by Zybrand » Fri Aug 30, 2013 2:48 pm

Ok here is the dropbox link:
https://www.dropbox.com/sh/i1b841qyfap5qlb/M1oh2i8MJI

I shared both scenes I am talking about because I think they are both relevant, and their SD files.

Thank you, for having a look at my sims.

User avatar
atena
Site Admin
Posts: 112
Joined: Fri Jun 04, 2010 1:25 pm

Re: Strange displacement behaviour

Post by atena » Mon Sep 02, 2013 8:28 am

Thank you very much, I will start the simulation right now and be back to you as soon as I found the issue.
Thank you again for the scenes.
Angel Tena
Head of RealFlow Technology
Next Limit Technologies

Zybrand
Posts: 11
Joined: Tue Aug 07, 2012 9:04 pm

Re: Straight displacement behaviour

Post by Zybrand » Mon Sep 02, 2013 11:33 am

atena wrote:In the other hand to no stop your work I want to recommend you as a workaround to have a look at the simulation graphs we provide in the "Help->Demo Scenes" menu, in this particular case, the one in "graphs" named "apply_stat_spectrum_scale_base_on_speed".
That could be an interesting alternative to my current workflow. I'm very keen to start learning graphs, I use a lot of Softimage ICE and this looks very similar.
I'm not that clued up on python and I learn the node based stuff a lot easier.

Probably a silly question because I haven't used graphs yet, but does that graph the bake that speed based displacement into the mesh?

User avatar
atena
Site Admin
Posts: 112
Joined: Fri Jun 04, 2010 1:25 pm

Re: Strange displacement behaviour

Post by atena » Mon Sep 02, 2013 3:45 pm

If I understood right your question then the answer is yes. That graph the final mesh is modified with the displacement and also the effect of displacement attenuated with the speed.

To learn more about graphs you can watch these videos:

http://www.youtube.com/watch?v=qtrv_fdAJ7s
http://www.youtube.com/watch?v=copIGgyHGSM
http://www.youtube.com/watch?v=RV9xBTJ3fvY
http://www.youtube.com/watch?v=YeLX0OANPeQ
http://www.youtube.com/watch?v=eAPtiTgyGAA
http://www.youtube.com/watch?v=E_tzyK6vOXk
http://www.youtube.com/watch?v=aeRzxp9KUoY
http://www.youtube.com/watch?v=19DNo0iGlaY
Angel Tena
Head of RealFlow Technology
Next Limit Technologies

User avatar
atena
Site Admin
Posts: 112
Joined: Fri Jun 04, 2010 1:25 pm

Re: Strange displacement behaviour

Post by atena » Tue Sep 03, 2013 12:51 pm

Hello, I have some news on that issue. Definitely is something wrong with the "splashity" channel that is computed for the attenuation. Seems that this channel has some noise that is causing the dimples. I filed a bug regarding this issue.

In the meantime I encourage you and other people trying to attenuation the displacement to follow the workflow of the simulation graphs. There are a couple of example scenes in the default installation. This is more powerful than the attenuation feature in the parameters.
Angel Tena
Head of RealFlow Technology
Next Limit Technologies

Zybrand
Posts: 11
Joined: Tue Aug 07, 2012 9:04 pm

Re: Strange displacement behaviour

Post by Zybrand » Wed Sep 04, 2013 12:49 pm

So I found a bug, glad I could help.
Going to do some displacement through Graphs today. Thanks for the help and advice.

Post Reply