Ultimate Water System - The most depreciated realtime water simulation [Depreciated Support thread]

This is all very confusing! I bought UWS yesterday in the Unity Asset Store! I hope there will be no problems with future updates, etc.! I did not know about all the problems here!

But what I can say after some tests with UWS is that this system seems to be the long-awaited well-functioning water system for Unity!

Best regards Mike

 

Just waiting for the upgrade. Will support whenever they're done fiddling around.

 

So I didn't see anything specific about this in the docs but it looks like in the vidoes you have the main feature I was missing with Playway. Which was to be able to put some type of edge blend on meshes like boats. If so I'd be upgrading for sure.

 

@pat_trick the package contains simple-demo scenes for most of the features mentioned.

@stevenwanhk I've optimised few things but the overall performance is comparable to PW.

@snacktime If i understand correctly, you're speaking about the "foam" that is placed where the object surface meets water. For now, this is pretty basic implementation without any advanced stuff.

 

Hello everyone!

We finally have some good news

Unity has finally finished their reviewing process - package is now available on Asset Store at link:
http://u3d.as/XrV

Owners of PlayWay Water System can now upgrade it to UWS for 50 USD. Promotional price will be available till the end of October, after which upgrade will still be available for 75 USD.

We'd like to apologize for all the delays and confusion with account changes.
All customers who already bought UWS please email us at support@moonlit.pl with invoice number to receive voucher enabling purchase on right account.

We're looking forward for your feedback and comments.

All the best,
Moonlit Team

 

Great, but I hope that the playway comments linked to it are only seen by Playway owners and not anyone else.

 

Yeah, you should see if you can get rid of the bad reviews. Not a good way to start.

 

This is my first upgrade experience with the asset store. While I'm not sure how much I'd jump back and use it, does purchasing this upgrade still allow me to use PWS? Will I have two entries in the asset store after the upgrade for PWS and UWS?

Best of luck to you guys on this asset, and the rest of us users for that matter. This will be the fourth water asset I've been through since Triton trying to get my project off the ground. I hope you guys are here to stay.

 

Hello, all

I bought the upgrade earlier this evening. I won't call this a comprehensive test by any means, but I thought others might be waiting for some "first impressions", so here they are:

1. In the last release candidate for Unity 2017.2.0 (I know, the release is out there, but I just got home from a business trip and haven't downloaded yet), there are warnings about obsolete script APIs. Unity was able to auto-upgrade those, and there were zero compiler errors.
2. I have not even attempted to add UWS to any of my own scenes yet (I have a sandbox project just for asset testing, so I'll try that this weekend). The demos included with the UWS package all work correctly for me in this version, on Windows 10 Pro 64-bit with an Nvidia 1080 GPU. My system is not a great place to test performance because it's a really high-end laptop.
3. In my opinion, the demos leave something to be desired in terms of interactivity -- they're just "the water doing its thing". That being said, there is an argument to be made for simple demos because they are easier to study and learn. My respectful suggestion to the devs is to keep these simple ones but add one or two more complex, interactive demos in future releases. This is a very small nit, but the devs can't react to customer suggestions if we never send them.
4. The visual quality of the water, the surface ripple effects, underwater meniscus transitions (very important to me), and raindrops are extremely good. The shoreline effects are, in my opinion, decent but not stellar. To be fair, I have not yet seen *any* water asset for Unity that I think really nails this, and UWS is as good as its competitors. The transition from above to below the waterline is spectacularly good. Thanks to the devs for this, because it's an important factor to me!
5. I'm not seeing any foam on the wave crests, but to be fair I haven't really tried to enable that yet. Call this one "inconclusive" for now.
6. In the "volumes" demo scene, I was able to generate some non-fatal runtime errors about screen resolution being outside camera frustum, but I generated these by doing some crazy parameter changes at runtime. That being the case, I call this one "pilot error" by me and not a bug. In fact, I'll give them props that whatever I was breaking didn't visually impact the scene, just spammed my console log.
7. I played around with moving the subtractive volume in some fairly crazy ways, and didn't break anything.
8. The camera-side components don't appear to be overly intrusive, and the scenes don not contain a complex clutter of "manager" game objects.
9. Wake effects on the galleon: Good but not spectacular (again, I haven't tried changing settings).
10. Physics demo: As an engineer who's picky about physics simulations, I call this a win. It is completely believable to me.

As I said, I'll be doing some further testing in the coming days, but I wanted to share an early impression in case others are wondering if this is compatible with Unity 2017.2. The answer is yes, with some warnings during import.

Overall, my first impression is good, pending deeper testing. I consider the US$50 upgrade price from PlayWay to be a fair compromise between "accepting obligations from the asset's transfer" vs. "being saddled with a problem they did not personally create" and "having added new features."

 

Oh, one other really small nit: The import of the galleon model discarded some polygons that were self-intersecting. The devs may want to clean up that model in a future release.

 

@syscrusher
Strange, i am on Unity 2017.2.0f3, and getting empty water when running (no running water), i am on Nvidia Optimus with Intel HD 4600 and Geforce GTX 860M. I have already tried running it with Nvidia GPU but with the same results, deleting Library folder or reimporting the asset also getting the same result.

The objects seems affected by an invisible water or ocean :-(.

 

I'll download the final and retest. My test tonight was with the RC.

EDIT: 2017.2.0f3 is working for me. I solved the missing foam problem. It just wasn't in the demo scene by default, but when I added the foam texture it's working.

I'm playing with some physics things now.

 

@Moonlit-Games-Studio, I'm stuck on a physics issue. I created a mesh GameObject in a scene with boundless water. The mesh is the Galeon mesh from the samples. I added a RigidBody with gravity and non-kinematic, materialized its center of mass and moved that near the keel, per the manual. I set the mass and buoyancy such that the gravity balance percentage is around 175%, which seems to work well for the floating barrels.

My ship floats, and moves from wave action with remarkable realism, but it floats a few meters below the water's surface, as if the calculation point for the physics is the center of the mesh and not the center of mass of the mesh. The masts stick out of the water, but the hull is submerged. That makes a certain amount of sense, but I can't seem to figure out how to build an object hierarchy that allows me to correct this with a Y axis offset.

Also, I'm finding that flotation only works if the MeshCollider is marked convex. Is that by design?

Finally, it seems that things become more buoyant as their mass is *increased* for the same collider volume. That is actually backward from real-world physics. Can you explain if this is by design, or is a user interface bug?

Thanks!

 

hi @MCPGNZ !!!

I have already updated from PWW and playing a little with UWS I have seen some things that I do not know if it is a mistake or that I do not configure something well. I'm just using the sample scenes with Unity 2017.1.0f3 version

In the Physic scene there is a very rare transparency in the water in the part of the barrilles that are submerged. It looks all white. I have touched water parameters and it is not corrected. Nor have I found anything in the manual about this problem.
Go with all water profiles except Sea - 0 Calm, Sea - 2 Light Breeze, Sea - 6 Strong Breeze,


Another question. Is there no particle spray for the wave crests as there was in PWW?

What exactly is the "Shore Color" parameter for? I've been playing it in the Shore example scene but I do not notice any changes. Or does it still not work? I know that this features put zuzzu in PWW in the Beta versions, but I never got to know what exactly.

Otherwise I do not know if I do something wrong, but in the same scene of Shore, the foam shows a black line at the end where the water is attached to the ground. What am I doing wrong? I leave video

regards

 

Screen shots illustrating the physics problem I mentioned previously, alongside two showing the lovely reflection behavior sunlight (just for fun). I'm using the Unity standard post-processing stack here, again Unity 2017.2.0f3.

 

Out of curiosity, are you doing anything special with ambient occlusion? For example, are you using SSAO?

 

They are the UWS example scenes as they come. The only thing I change is to deferred the project and UWS component of the camera.

 

Is OpenGL fully supported? There seems to be a problem with shadows. Here's a shot of the Samples/Physics scene. Unity 2017.1f1. Fresh project. The shadows looked fine in directx, but are misplaced when switching to OpenGLCore. Am I missing something? Thanks.

 

Here's a closeup of my result, after opening the Shore demo with the following changes only:

1. Added an ExplCamera component (from the asset by Voxel Frog, which I highly recommend for prototyping), so I could zoom around the scene for a good look. This doesn't change rendering, just allows camera movement by keyboard.
2. My color mode is changed project-wide from Gamma to Linear (if you haven't done this already, you should, in just about every project).
3. I changed the camera to Deferred Render mode, to match your settings.

My Shore Color setting is unchanged from the default.

Observations: You are seeing much more pronounced foam effect than I am, but I am not seeing the black line that you have illustrated in your screen capture.

EDIT: Ummmm.... and now I will actually include the screenshot I mentioned. {doh!}

 

Suggestion to @Moonlit-Games-Studio: The documentation for UWS is good as far as it goes, but some additional detail would be nice in future versions. I say this with full understanding that it's a new asset and will evolve over time, of course. Just know that some of us do actually RTFM out here, so it's not a waste of your time to make it detailed.

 

Document reader here too!

 

Very glad the upgrade path got sorted out. Just bought, will test first on MacBook Pro 15" 2017

 

I have another question: With the need (at least, as I observe it) for a convex collider on any boat mesh that is to float with physics, that introduces a problem for any other RigidBody objects that would need to be on or in the boat. Marking the MeshCollider on the galleon as convex precludes a character walking around on the deck.

I'm planning to do some experimentation with floating a collision-only invisible mesh, then making the visible boat with a non-convex MeshCollider be a child of the physics mesh, but I would have to think there is a better way. What is the way that UWS intends this situation to be handled?

 

One other suggestion, if you aren't sick of hearing from me already: Please in future versions add a way to view a low-quality water proxy, for unbounded water, in the scene editor when not in play mode. It's really tricky to set other objects to their correct height relative to "sea level" if the sea is invisible. It needn't be animated or reflective, just something that looks watery and behaves as an infinite (or very large) plane like the water will render at runtime.

 

Hmm, looks like has not received basic testing on Mac.

Caustics: no caustics shown
Volumes: nothing shows at all, just the pre-rendered background
Dynamic Water: seems to work, nice waves
Profile Blend: doesn't seem to do anything, just shows waveless ocean and overstated vignette
Raindrops: fall up rather than down, otherwise nice
Shore: all black where contacting beach
Underwater: looks nice
Wake water: looks nice, presuming it's supposed to be a glowing trail in the water

Particles: non-stop errors to the console
UnityException: FindKernel failed
UltimateWater.GaussianShader.Assign (KernelType type) (at Assets/Ultimate Water System/Scripts/Shaders/GaussianShader.cs:128)
UltimateWater.GaussianShader.get_VerticalKernel () (at Assets/Ultimate Water System/Scripts/Shaders/GaussianShader.cs:21)
UltimateWater.GaussianShader.set_VerticalInput (UnityEngine.RenderTexture value) (at Assets/Ultimate Water System/Scripts/Shaders/GaussianShader.cs:60)
UltimateWater.WaterSimulationArea.Smooth () (at Assets/Ultimate Water System/Scripts/Dynamic/WaterSimulationArea.cs:261)
UltimateWater.WaterRipples.Simulate () (at Assets/Ultimate Water System/Scripts/Dynamic/WaterRipples.cs:120)
UltimateWater.WaterRipples.FixedUpdate () (at Assets/Ultimate Water System/Scripts/Dynamic/WaterRipples.cs:69)

Giving up on Mac for now. Not encouraged at the level of testing here.

 

Hey have you guys optimized the method to get heights for waves by chance? Playway is ridiculously expensive here with no optimization whatsoever.

 

What's your Unity version? I also have a Mac, and if time permits I can try to replicate your results.

 

I did a test with the galleon, and for me it seems right. When i press the "Set Mass to obtain 100% Gravity Balance", the hull is under water like in your screenshots. But it comes out of the water, when I reduce the mass of the rigidbodey.

Btw.: It seems that the height calculations are a lot more precise then in the last PW beta i used. There are still a few glitches starting at windspeeds like on the "Gale" preset, but i think it looks verry good so far.

 

hi
Yes. I already had Linear in the project / scenes.
As I said before I do not have any additions to the camera (SSAO, etc). Any. They are the example scenes as they come. I just set the camera to deferred and Linear in the scene.

We give it the same, except the most pronounced foam that you noticed, because it touches the parameters of Foam in the profile. But I do not think that's why I do not see it influence the problem when I touch the parameters.

I have continued testing and it seems that the problem comes from the parameter "Edge Blend Factor"
I leave video:

regards

 

+1

 

Hello,

First off I’d like to agree with @AdamGoodrich

If I were an asset dev I wouldn’t make a tool like a water asset I would make something like 3D goat model, it requires very little support won’t need updating until PBR is superceeded and I can always copy it make some tweaks and call it a Sheep model.

So when Moonlit-Games-Studio steps up to make / continue the development of an asset like this water I like to say:

Thankyou Moonlit-Games-Studio !!!

without you I wouldn’t be making a game I would be trying to make some water

Also thank @MCPGNZ stepping into the crosshairs.

Cheers

Roy

 

Thanks for all the comments and suggestions!
I'll try to address the issues mentioned and answer all questions as soon as possible.

It's weekend, but I've wanted to update you so you don't feel like nobody's watching

 

Hear, hear! I have a couple of water assets, and from what I've seen so far I will continue to use both of them as the "right tool for the job" depending on my project needs.

I'm a programmer, but that being said I have no interest to hand-craft my own shaders because the graphics stack is not my field of interest. I'm very grateful for premade shaders (and related tools) from assets like UWS, Suimono, UBER, RTP, CTS, Alloy, and others, and for tools like Shader Sandwich and Shader Forge that let me interactively create and test custom shaders for when there's nothing off the shelf to meet my needs.

Yes, I'm reporting problems that I find with UWS here, because that's how bugs get fixed, but I never forget in doing so how glad I am that I did not have to create my own water tools.

 

Another screen shot of WIPz. I solved (mostly) my flotation issue for the galleon. There is a separate mesh provided for the lower hull, and I used that with a convex MeshCollider to handle the flotation. My visible galleon (the LOD0 mesh) is a child of the physics object, and I've offset the center of mass downward as recommended in the manual. The boat still seems a little snappy (too low inertia) even with proper level of buoyancy. (This is probably just a matter of me needing further adjustment to my physics settings, so dev please don't take that as a "bug" report...it's not.)



Still to be solved is this other issue: Because I have the RigidBody on the convex MeshCollider of the separate hull, it floats. But Unity also tries to include the non-convex MeshCollider from the visible ship render, which is the collider I'll need for characters to walk around on the deck. I'm looking for advice from the devs for this.

One approach I am considering: Float an invisible version of the hull-only mesh (no MeshRenderer component on that one). Then create my visible galleon as a separate GameObject tree. The visible mesh gets a non-convex MeshCollider, and a RigidBody set to kinematic mode (which Unity says can be done with a non-convex collider). I'll make a trivial C# script to move the kinematic visible object in sync with the invisible one controlled by UWS physics, with the sync happening on every FixedUpdate() method call since it's driven by physics and not frame update.

On the downside, my proposed method introduces some extra complexity, but it does have two interesting features:

1. I could use some Lerp or Slerp damping to make the visible boat seem more inertial and to make character movement aboard less problematic. In effect, the physics sync would be a "hook" where I could add temporal damping behavior.
2. That same hook would also be a useful place to sync the object across the network in multiplayer situations.

Again, any recommendations or suggestions from the devs would be welcome. How did you intend for me to do this, and am I pursuing a reasonable course?

I have some other Unity project work that needs to get done today, so this is probably my last post on UWS for the day.

 

So upgraded from PlayWay and having some issues with linear/deferred and the camera set to deferred image effects. Foam is the obvious one. The other is those dark horizontal lines showing over the terrain.

If I set the camera to default queue these issues go away, but I lose foam.

 

@syscrusher

For starters you can't have characters walking on non convex moving mesh colliders. Colliders must be convex to collide, and your character has to collide to walk on the platform. For best results use box colliders. Moving platforms are tricky you will get bad interactions with anything but very simple collider shapes. Colliders will invariably intersect and then get pushed apart because characters will often be pushing against some edge. Usually resulting in the ship getting violently pushed away from the character (if using character controller).

I think if you have a top level empty gameobject, and then have your hull as a child and your platform as another child set to kinematic, it should work. It's been a while since I did this with moving platforms so could be missing something here.

 

Having fun Testing in Unity 5.6.3p4 |
*request How to create HUD that can track depth?

1. Objects are visible -Y 200 units -Z 10
   How to make objects less visible in deep water?
2. How to fade caustics with depth?
3. Request a demo scene with all settings on: (Caustics + Shore + Underwater particles, etc.)
   each demo scene works, but trying to add them together is not working for me.

 

Hi,
I am experimenting with that as well, try setting the Precision of the WaterPhysics component to 0.2 or something like that. It is still responding a bit too much to the small waves i feel, but it is a bit better. Also try to raise the Drag and Angular Drag of the RigidBodey component. Because of unity´s drag implementation i think you need to have unrealistically high values here!

PS: please correct me if my guess is wrong here!

 

Hi,
I cannot get the RadialGrid be rendered beyond 1000 units, regardless how far the far clipping of the camera is set and if boundless is active or not.
It works fine with ProjectedGrid, but there are some artefacts (as it can be found in the manual) when the camera is moving and the foam is nearly not visible.

Please, I would like to have some information on that?

 

I revisited the documentation, and did some testing, based on this comment, because I remembered things differently.

Non-convex mesh colliders can collide, as long as the mesh collider itself does not have a rigidbody, or the rigidbody is kinematic.

I tried my idea with the script, and it actually works except for one problem: my mesh and the flotation mesh are overlapping in space, so they collide with each other and make a mess of the physics engine. I wish Unity had collision layers, because this would be easy to fix then.

 

Works for me. Do this at your own risk!

You can chang the radialGrid in the "WaterRadialGrid.cs":

Go to line 213

// [todo] clip is clamped due to precision errors
// find a way to deal with them without clamping the value
float farClipPlane = Mathf.Min(camera.farClipPlane, 1000.0f);

You can see - the "farClipPlane" is clamed to 1000 due to precision errors!

Change the line 215 to

float farClipPlane = camera.farClipPlane;

Now the radialGrid goes with the clip plane of the main camera.

Works for me. Do this at your own risk - I do not know how this change affects physics or other things!

 

Thank you Mike!

I guess this is the fix for the precision errors in the height calculations i stumbled into while using the PW beta. I will try that, but i think i cannot handle the trade of for what i am trying to do.

@MCPGNZ
Is it somehow possible to use radial grid near the camera and a projected gird far away from the camera, because i need an open ocean, and a far clipping of 1000 units is much to near for my usage?

 

@Moonlit-Games-Studio
I'm experiencing some issues with shadows that look like a flashlight effect. I believe it has something to do with Planar Reflection because when I change those sliders the flashlight size changes. I do however want to remove the flashlight effect completely because it creates these dark shadows at the edges of the screen.
I did not experience this effect before upgrading to UWS.

This picture is how it looked before i upgraded


This is how it looks with UWS.


Would love it if @Moonlit-Games-Studio had a solution for this or if anyone else has experienced the same issue.

Best regards
//Semih Parlayan

 

Progress! The visible ship is a separate GameObject that is neither child nor parent of the floating invisible convex collider. I have a very simple C# script that causes the visible mesh to follow the floating collider's position during LateUpdate(), but with an offset of +8.00 units Y, so the two colliders do not intersect.

The invisible floating collider is a convex MeshCollider of the bottom part (only) of the ship. The visible ship's MeshCollider is non-convex. The floating collider has a RigidBody with gravity and non-kinematic attached, while the visible ship has a RigidBody with gravity disabled and kinematic enabled.

I placed the swarm of barrels a few meters above the masts of the visible ship. As you can see from the screen shot, the RigidBody barrels correctly collide with the visible ship and roll around on the deck indefinitely.



@snacktime, I do agree with you on the collider physics issues with moving platforms in general, because if the platform or character is moving fast enough they may quantum leap through the collider's edge on a single update. I expect to have to take some precautions against that, as I would with non-water platforms, but my intent in this test was just to prove the concept that this could work at all.

Because of my 8 meter vertical offset, the ship is currently floating in midair above the water, but it's moving correctly. The solution to the offset is evident: I need to split the ship mesh into the hull part needed for flotation, plus the upper deck where other objects will interact. If I want realism of the interior, I could even fill the ship's keel with rocks, which is exactly what was done in period, thus conveniently filling the inside of the convex flotation collider.

EDIT: I forgot to mention, following the suggestion from @Funkeys to increase angular drag fixed the "snappiness" issue I reported earlier. The physics precision setting in UWS helped a little but didn't really solve it, but bumping the angular drag to 1.0 makes a very nice, smooth motion that seems realistic to my eye.

 

You have a new aspect for your game, Syscrusher! Dodge the barrels!

 

First things first.
I've created repository on github. This way i can easily track all the issues that arise, and you will be able to see the progress made. (Note that if we start discussing each problem in detail here, it can become a mess very fast).
I'm creating Issues based on your posts (it would be very helpful, if you could add your Issues directly there).

Now for the questions, there's a lot of them, so I'll split the answers in few posts.

@syscrusher
The obsolete api warnings are nothing to worry about, the package supports versions 5.4.5+, so some of the api's changed in newer versions, but the Unity manages to upgrade them without any problems and if-defing all of them does not really change anything but the code becomes uglier (just look at some c++ code with extensive if-defs - it's can become pretty much unreadable)

I personally think that very small demos are much easier to learn from, because you can focus on just the thing you learning without the need to look at all the other stuff and wonder if it's necessary or used for something completely different. I was considering creating more complicated stuff but it came to the point of: create more demos or spend more time on the codebase, and I've chosen the latter.

As for the issue with physics, I've added it to the Issues as the physics bug, even if it's not,
it will stay there (along with explanation) to help other people that stumble on the same problem.
(and it's a good sign that the documentation should be extended to explain it simply to others).

 

Great Idea!
I have to say I am really happy that this great water system has got support again!

PS: my sys specs for the issue

**spec:** Intel Core i5-4670K, Geforce GTX 970, Samsung SSD EVO 750, 16 GB Ram, Win 10 Home x64
forgot :/
**unity:** 2017.1.2f1

 

@Funkeys @brave_mike

The far-clipping plane fix mentioned by @brave_mike should be ok for now. This does not affect physics calculations but it can cause pixel-flickering at the horizon of the water. I've already added it to the Issue list.

 

I'm also having issues with the shore sample scene with opengl via macOS and an AMD firepro D300:

I tried switching my editor to METAL but it locks the editor up when I try and make the switch (which only seems to happen with UWS imported).

 

@mons00n
Hi, I've added the details you've provided to the existing Issue.