Ship in Ocean

[flipbook]

I would put Liquid-Like on at least 1, it will help not produce the linear splashes .. as seen in the max manual (scroll down a bit )

https://docs.chaosgroup.com/display/...Mist-FoamOnHit

Higher it is the slower the simulation tho I've ran it at 100 before and slows it down a lot.

[waaazoo]

I would put Liquid-Like on at least 1, it will help not produce the linear splashes .. as seen in the max manual (scroll down a bit )

https://docs.chaosgroup.com/display/...Mist-FoamOnHit

Higher it is the slower the simulation tho I've ran it at 100 before and slows it down a lot.

hmm, let me try this. I read also the Affect Liquid should be at 1... thanks for the tip !

[flipbook]

Yea best put that on too, splash can merge back to liquid again etc.

[waaazoo]

New sim, new findings.
This time it starts to look right

It seems that activating "Affect Liquid" and/or "Liquid Like" removes the blobs that I had before. I will test what exactly makes this on a smaller grid.

I wanted to render out that sim during the week-end, but had suddenly a problem that I see for the first time. When trying to render (in Maya) or batch render, it seemed that simulation was starting again.
May I have ticked something that tells Phoenix not to use the cached sim? When opening the sim, it reads the cache, and as you can see I can make a playblast. But as soon as I try to render (even without animation render activated, it gets back to frame 1 and calculates the sim).

https://vimeo.com/202715705

https://vimeo.com/202722228

https://vimeo.com/202738575

[flipbook]

No idea about that one ... one for Svetlin to answer I don't see why it would start simulating again.

[waaazoo]

It's clearly the "Affect Liquid" which removes the blobs. If I have just Liquid-Like at 2, the blobs are still there. Good to know.

[Svetlin.Nikolov]

When batch simulating, it seems that it takes into account what is in the time slider and/or range slider, rather than what is typed into the simulation "Start Frame" and "Stop Frame".
Example: Start Frame:0, Stop Frame:1500, Rangeslider 0 to 100.... when batch simulating it will calculate 100 frames.
Is that the behavior you expect ?

New sim, new findings.
I wanted to render out that sim during the week-end, but had suddenly a problem that I see for the first time. When trying to render (in Maya) or batch render, it seemed that simulation was starting again.
May I have ticked something that tells Phoenix not to use the cached sim? When opening the sim, it reads the cache, and as you can see I can make a playblast. But as soon as I try to render (even without animation render activated, it gets back to frame 1 and calculates the sim).

Thanks! Gonna check these!

[waaazoo]

Hi,

I started again with some basic testing of the seat preset and rolled back to Maya 2014. At the beginning I thought that Phoenix for Maya2017 would be more stable/efficient, but it doesn't seem so.
As I prefer much more Maya2014, I'll be running my new tests from there.

So, I decided to make some simple tests, with a simple Ball at different sizes, just to see what happens, how the grid is defined and how it looks.
I know that these tests are not exactly comparable as I changed a little my mind, during the testing, on how to setup the scene. In the beginning (starting from the 1cm ball) I used a movement based on the size of the object (and with a grid following the object), and then changed to have a movement of 10knots regardless of object size (and with a static grid). I should probably redo the first three with a 10 knots movement.
ALL scenes are based on a centimeter setup in Maya (1 Maya Unit= 1 cm)

It's quite interesting to see how the grid changes. Remember: each time, Ball is created and then quick preset is pressed to create the ocean.
The best effect is given with the 5m Ball (in my point of view), but the Grid itself (units scale, cell size, Total grid size) makes no sense to me...

I would like to have the setup of the grid differently. The Width, Length and Height size should be specified in system units (so I can think how big I want my grid around my object), and then I can give a cell size (and it tells you how much cells will be created), or maybe even just a precision/definition parameter. Units scale would be just there to exaggerate or damp the effect. Makes that sense?

Ball 1cm:



https://vimeo.com/203090589

Ball 10cm:



https://vimeo.com/203090488

Ball 100cm:



https://vimeo.com/203090417

Ball 500cm:



https://vimeo.com/203090286

Ball 1000cm:



https://vimeo.com/203090146

[Svetlin.Nikolov]

Hey,

Using the toolbar presets, you should get identical results no matter what scale you use. We should investigate the differences between the first 3 scales and do some changes - thanks for noting this!

The presets try to compensate for the scale of the scene, so the simulation would produce the same result no matter what. This is done by automatically adjusting the Units scale coefficient. The ocean preset is designed to look right when an object passes through the simulator in about 15-20 seconds, so the scale is adjusted so that the object would be roughly the size of a ship.

Definitely gotta rename the Width/Length/Height - this way they don't say much about what is actually going on.

Cheers!

[waaazoo]

Hey,

Using the toolbar presets, you should get identical results no matter what scale you use. We should investigate the differences between the first 3 scales and do some changes - thanks for noting this!

The presets try to compensate for the scale of the scene, so the simulation would produce the same result no matter what. This is done by automatically adjusting the Units scale coefficient. The ocean preset is designed to look right when an object passes through the simulator in about 15-20 seconds, so the scale is adjusted so that the object would be roughly the size of a ship.

This is where I totally disagree, Svetlin.
We are trying to simulate natural effects, fire, smoke, explosions, water, sea... we have to be aware that we have to model in a correct size to simulate what we want.
Having a sugar cube falling into water is not the same as having a car falling into water. It should NOT do the same ! it should take into account the system size settings, and the size of the object.
If I'm trying to simulate a sugar cube falling into water, but model my cube 1m wide, it's my fault if I don't get the desired effect!
If Paul simulates the space shuttle take off, but makes his space shuttle as big as a mobile phone, then he shouldn't be surprised to have a bad simulation

Definitely gotta rename the Width/Length/Height - this way they don't say much about what is actually going on.

Cheers!

I'm glad to read that!

cheers!

[Svetlin.Nikolov]

Hey,

I should explain the motivation behind the toolbar presets in more details. The toolbar is just a starting point - I don't expect that a person would pick Phoenix for the first time and know he's supposed to work in real-world scale or any of the specifics, so this is why the toolbar has to make the job as easy and straightforward as possible - you shouldn't fail with the very first click of a new software because you didn't know things that you are not really supposed to know until this point

With time, once a person learns what the unit scale does and how the sims behave on a low and high scale, he can alter them manually, like basically all the other simulation parameters. The Ocean preset would work as a starting point when setting up the ocean texture, the material and the general splash and foam settings, but if you want to make a small-scale water splash, it's not the right preset to start with in the first place. Another reason for the way it is right now is that most of the presets on the toolbar would not look good if they are simulated on half or double the scale, not to mention if the scale is a a thousand times off, so this is the reason the scale is always adjusted for those.

In fact, I've also thought about a dialogue that would ask you for the correct size of the simulation when you select a preset, but actually changing the unit scale manually is all that this comes down to, so I've left this out for now, until we gather more ideas and feedback.

Thank you very much for raising this interesting topic!

[waaazoo]

updated the first tests... and should have said before that the size represents the Radius of the ball....
I had to lower the speed of the small balls (obviously)

Ball 1cm:



https://vimeo.com/203119157

Ball 10cm:



https://vimeo.com/203118905

Ball 100cm:



https://vimeo.com/203119066

Ball 250cm:



https://vimeo.com/203118904

[Svetlin.Nikolov]

Yup, this behavior is expected with the Ocean quick setup - you can see how the unit scale changes for each setup to adjust for the scale of the objects, so each one would act like you have a meteorite running through the water at high speed

[waaazoo]

Yup, this behavior is expected with the Ocean quick setup - you can see how the unit scale changes for each setup to adjust for the scale of the objects, so each one would act like you have a meteorite running through the water at high speed

Maybe stupid to ask, but when I change the "Unit Scale", does it affect everything in the simulator or just the grid settings ?

PS: I still don't really agree with the quick setup, even if I understand your point. Maybe a simple warning "your scene needs to be properly scaled" would make the trick for those who start with Phoenix...

[Svetlin.Nikolov]

Ah, yeah - it does affect everything related to gravity or scale of the effect of sources.