Triton Oceans and 3D Water for Unity Pro / Windows [DEPRECATED]

Triton Oceans and 3D Water for Unity / Windows & Mac

TRITON FOR UNITY HAS BEEN DISCONTINUED.
This thread will remain for support purposes for existing customers.


Thank you,
Frank Kane
Founder, Sundog Software LLC

 

Very nice plugin!

I sent you PM.

 

Been using the demo version while waiting for it to reach unity store
I haven found a few bgs and the average time, not for an answer but for al actual fix, was less than 10minuts!
Has someone who spend a lot on assets (i've probably bought over 50assets including the eluSive substance database) I'm pretty happy to see top end plugins coming for unity.
There alreay are nice water solutions fo unity, mobile compatible, so instead of more of the same it's really good to have something to use to mane top windows game with! Also this runs silky smooth ol my nvidia 540M !
This plugin impressed me much i gave a small project to his author and i can confirm from that that he's extremly nice, open, hard working and quick and produces great solutions!

 

Hi mr_Necturus, thanks for having a look! I can't tell if my reply to your PM went through or not - if not, the answer is "no, we aren't redistributing that"

 

Looks really amazing, it's a shame that it isn't build in as a pro feature.

 

Looks interesting. I'm still using the "community" ocean water (which happily works quite well in 3.4 but not so well in 3.5) but development of that water seems pretty stagnant. Might be worth the investment in this plugin if it's clearly superior and will be continued further.

Couple of questions:

1. Any videos of these features in action?
* Foam and particle-based spray effects
* Ship wakes with 3D Kelvin wakes and propeller backwash effects

2. Will this support Unity 4.x in the (near) future?

 

+1 By the way, what ever did happen with the development of the Unity 3.x water? Seems it just stopped without ever being completed / fixed.

 

Thanks for checking it out, bigkahuna.

Foam and spray is visible in the second demo video above, once I crank up the wind - at around 0:16.

Ship wake effects are shown in the second half of the tutorial video.

I am planning to support Unity 4 when it comes out, barring unforeseen changes to Unity's plugin model - Triton supports DirectX11 and the code is already in place in our plugin for it.

 

truly impressive!!!

 

i am sorry but i can’t say: wow!
i just do not have the time to test the trail version right now, so i can just judge from what i see in the videos:
– no soft faded shore lines
– no refrection
– pretty weired reflection on this picture:
– how does it react to lighting?
– i can’t see any specular
– no caustics

writing this might not be fair – i just don’t have the time to dig deeper into it.

lars

 

Looks really nice especially that picture with semi transparent ocean. No refractions and proper reflections is bit drawndown tough...

 

Hi,

The screenshots were all taken using dynamically-generated environmental cube map reflections, which don't handle local reflections very well. In the demo, the environmental reflections are generated directly from the skybox which works much better. However, Triton does allow planar reflection textures to be passed into it which are better suited for local reflections.

Triton does compute refractions using the full Fresnel equation, but as it's targeted mainly toward ocean simulation, the refraction color is a constant deep water color rather than refracting geometry underneath the water (which may not be present.) It wouldn't be hard to modify our shaders to do a texture lookup from a cubemap on the refraction vectors instead of using a constant color.

 

Thanks. The tutorial video is pretty good, seems pretty easy and straightforward to use. This is pretty close to what I've been waiting for. Would love to test this in a couple projects that I'm using the community ocean shader in to compare one versus the other. No time for that at the moment but hopefully soon.

A couple more questions:

- I noticed that the trailing end of the boat wake very noticeably "pops" instead of smoothly fading away. Any way to fade that more realistically?
- Is the plugin using an external .DLL?
- Does the ocean surface support projectors for effects like blood / oil on the water?
- What happens if the camera moves below the surface of the water?

 

- I noticed that the trailing end of the boat wake very noticeably "pops" instead of smoothly fading away. Any way to fade that more realistically?

Hm, you're right, I'll have to tweak that for an upcoming revision. For now, I think it can be worked around by opening up the Triton.config file inside the TritonResources folder, and reducing the setting for wake-wash-length.

- Is the plugin using an external .DLL?

It consists of two DLL's. One is a C# class library that wraps Triton's native C++ code and is largely auto-generated using SWIG. The other is the actual native DLL for Triton itself. Both are included in the Plugins folder, but the native DLL does need to be installed alongside the final executable in order for Unity to see it outside of the editor.

- Does the ocean surface support projectors for effects like blood / oil on the water?

Not at this time, but that's a good idea.

- What happens if the camera moves below the surface of the water?

The surface of the water, which will still be visible, will be fogged more strongly to simulate reduced visibility underwater. The fog density and color are configurable for both under and above water. Triton does not attempt to fog the rest of your scene to match for you, however.

Thanks!

 

Any chance you can post a picture? Would we be able to adjust the scene's fog to match the ocean water's fog?

 

Yup, here's an underwater shot from our image gallery:



Yes, it shouldn't be hard to match our fog, since it uses user-supplied parameters. There's even an option to use the Render Settings fog values for above water fog. to make it automatic.

 

Nice! Looks a lot like the community shader but (hopefully) faster, more features and support for versions > 3.5. Can't wait to start digging into the demo. Any plans for an "introductory" special price?

 

It should be plenty fast - it's mostly native, and uses CUDA, OpenCL, OpenMP, and Intel's Integrated Performance Primitives to make sure no computing resource goes wasted. I haven't tried the community shader, so can't really say how it compares quantitatively. I can tell you that the C++ version of Triton can hit 300 FPS on a GTX690, and the Unity version is really just a thin wrapper over it.

No current plans for messing with the price; we did do some research with potential customers to arrive at it.

Thanks so much for having a look!

 

hi there,

it might sound even more harsh, but any water shader offering less possibilities like this one: http://www.ogre3d.org/tikiwiki/Hydrax is nothing more like a fx sahder to me.
the example above is from 2009 so in 2012 even in unity we should have much better water that the one shown in your videos or screen shoots.
sorry to have to say that –*it is just my 50 cent.

lars

 

Are you seriously asking for $400 for this water system? this isnt the US navy, we don't have billions to throw around..

I'm willing to pay upto $70 for this... i'm all over EasyWater 2.4

 

is it really 400$?

puhhh –*that is quite some money!


lars

 

The substance database does cost $900+... what do I think of that? I think that's a bargain.

What about this half finished water system for $475? i say LOL

 

I thought about how awesome this asset is, but the price... really? shut us down like that.. how many sale you reckon you'll make? 5?? 10??... I think if you lowered it to say $70

$70 x 100(easy) = $7000

$475 x ...10 max = $4750

 

@sundog - "I haven't tried the community shader, so can't really say how it compares quantitatively."

Here's a link to the project's page on the wiki: http://wiki.unity3d.com/index.php/Ocean_Shader_Project
It works quite well considering it was done 3 years ago and was result of donated efforts by TwinFox, Nikko, Trond and a few others. Unfortunately it hasn't been updated in years and when I last tested it no longer worked with any version of Unity above 3.4.x

 

Sexy tool, outrageous price...
We can re-create the same functionality of this plugin using Megafiers, for $150 bucks; also Megafiers can be used for many, many many things more.
The only reason I would consider this is because of CUDA and OCL, still would try anything else to not pay that price.

 

Really usefull tool and nice, but i have to agree price is way out of range. Not becouse product would not deserve it but just suggestion becouse you wont sell much that way.

 

That's only one side of the calculation, this may be the whole calculation for some asset authors who just drop an asset, support it for 1 week and leave never to be heard of again, but if you're professionally supporting a product then 7000 for 100 users to support isn't quite looking better than 4750 for 10 users.

 

Professionally supporting? does that include phonecalls and facetime?

 

You'd have to ask him that, but as i've said my emails were literally answered (with a fix included) in minutes multiple times, including going on skype if needed.

 

Yeah but that sounds more like a service... I own a huge collection of assets from the store and i hardly ever have to contact the developer, and when i do its never a problem.. and if i do its usually because there is something wrong with the asset, I wouldn't want that if i spent $400

If this asset comes back down to earth I'll be the first one to buy it. but I would never go over $70

with that said i know developer simply wont comply.. we can talk about this all day

 

Aye for the service, but the advantage of professional support with an history beside it (even if you don't need support) is that you know it won't be droped in the next week and end up uncompatible with the next unity version (like the before quoted ocean shader).
It's also rather unfair to compare it to a single shader, it's not just the shader, it's the huge library and HW support driving it you're paying for, can't compare a 200 loc shader with a shader + 70Kloc of C++ in price.
I know i'll be the first one to buy it , at 400, but i guess the price will be too high for many yes.

 

Thanks for the feedback, Lars. Let me try to address some of your concerns:

– no soft faded shore lines

The transparency of the water is adjustable via the "depth" parameter, which also affects the shapes of the waves. For most of the screenshots, we kept this set somewhat high, since oceans typically aren't very transparent even near the shore.

– no refrection

The full Fresnel refraction equation is computed in our vertex shaders, but since Triton is optimized for open, infinite oceans, it just refracts to a constant deep water color, since we can't assume there is any sort of geometry visible underneath the water. However, it would be trivial to do a texture lookup from a cubemap instead of blending to that constant color. I like that as an idea for an upcoming rev.

– pretty weired reflection on this picture:

That's because that picture used nothing but an environmental cube map for its reflections, and those aren't well suited for local reflections. Really the ship should have been omitted when rendering that cube map. Triton does allow you to pass in arbitrary reflection textures and matrices to map them in our pixel shaders prior to being projection mapped, which is a better approach for local reflections.

Here's an image of a planar reflection map being used with Triton:



– how does it react to lighting?

You can tie Triton to a Directional Light, and it will use that information as well as the ambient light in the render settings to light the water surface.

– i can’t see any specular

I think this has more to do with my inability to produce good screenshots than anything else - I think the position of the sun in those images didn't line up with the directional light I set up quite right. There is a specular term in our shaders. Here's a picture of what it normally looks like:



– no caustics

No, there aren't. Underwater effects are limited; we just fog the ocean surface to simulate reduced visibility. By design, Triton doesn't mess with any of your scene's geometry or shading other than its own water surface and spray particles.

The niche Triton is trying to fill is for open oceans where large waves need to be rendered at very high framerates, as in for simulation and training applications. In those sorts of settings, you need Tessendorf-style FFT-powered waves, and that's where Triton shines. It does that faster than any other solutions I've seen; most of its 70K lines of C++ code are dedicated to conducting a realistic physical simulation of the waves themselves in as little time as possible, by using technologies such as CUDA and OpenCL.

Triton can also be used for shallow, calm water - but there are lots of other solutions for those sorts of scenes available.

 

hi sundog,

there some very important points you have made in your last post.
i would have liked to read them earlier…
also the new images look much better than those you have posted before.

may be you can also show how a correctly setup shore line might look like?

lars

 

Sure, here's an example of a shoreline (disclaimer - these latest images are from the standalone Triton and not from within Unity, but it's the same engine.)



Achieving this look in Unity would require a small bit of scripting, as you'd need to pass Triton the surface normal of the local seafloor in order to get proper blending along the shoreline's slope.

 

The ship's "stern wake" in this screen shot isn't how it would be in real life. In your screenshot it looks like an arc (off the ship's starboard quarter), while in real life it would be straight lines emanating at an angle from the stern (and bow). The foam prop trail would most likely fade more rapidly than in your screenshot and how quickly it fades should be effected by the sea-state.

 

The ship in this scene is moving along a circular path, which is why the wake is curved. It's a feature, not a bug

The rate at which the backwash fades is configurable within the TritonResources/Triton.config file. Propeller backwash is important in training applications as they are used for spotting and identifying ships from a distance, and they are more persistent than you might think.

Here's a reference photo:



I actually spent some time on a ferry to try and get it right.

 

Hmm... still doesn't look right but it might be the perspective view that's fooling me. What about a "bow wake", can that be added or can the stern wake be moved to the bow?

Sorry to be so fussy, but this is an area I have a lot of experience in.

 

Yes, when you create a wake generator object with Triton, you can specify an offset to position its origin at any position relative to the ship. So, you can have the wakes originate from the bow, stern, or anywhere in between.

I think it is the perspective... here's a top down view showing the curvature of the path a little more clearly:



I based the wake simulation off of these lecture notes from MIT:

http://dspace.mit.edu/bitstream/handle/1721.1/35268/13-022Spring-2002/NR/rdonlyres/Ocean-Engineering/13-022Surface-Waves-and-their-Interaction-With-Floating-BodiesSpring2002/C1D972B0-20A2-4386-9A13-FA537AC2D614/0/lecture8.pdf

And, the run-time implementation was inspired by this paper:

https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&ved=0CCIQFjAA&url=https%3A%2F%2Fcircle.ubc.ca%2Fbitstream%2Fhandle%2F2429%2F5372%2Fubc_1994-0475.pdf%3Fsequence%3D1&ei=Jx5eUIvvEors9ASU5YDgCg&usg=AFQjCNHZL7u7kNVkFTrGMSaEr9iQV0Ho3Q&sig2=mi2iBdz0eQN95p4GdYNJBw

 

Really impressive and impressive, but as said by previous devs, it's way too expensive for a lot of us !

The 2 other interessting solution for water are less than 100 $ (Suimono tarashan water)

 

You can't compare Suimono and Tarashan water which are both just water (plane) shaders with such a complex 3D solution.

 

rubbish

 

Seriously do you just crawl the forum to post 1 liners everywhere lol?

And no you can't compare them, different target audience, different goals, and diferent project scales. The only similarity is that all of it is water

 

I'd compare easy water to this

 

I just compare it with other water solution, and say that it is way to expensive.

With Direct X 11 and tesselation support in Unity 4, i believe that another system will come, less expensive.

 

@sundog - Thanks for the explanation, looks like it has a lot of potential.

 

+

 

I'm not sure, but in any case if another system comes with DX11, it'll be DX11 only

 

Yep thats the main problem of DX11 features, they are so different in creation process to DX9 features that you have to completly redo everything to fit it for both DX systems. Hair, detail tessellation, water, name it.

 

Fortunately the Triton engine already supports DX11 (it uses compute shaders when using DX11), so we're well positioned to support Unity 4 when it comes out.

Shifty, if you really think Triton is comparable to a $10 water shader, you clearly haven't taken the effort to actually try Triton out or understand what it does before commandeering this thread with your insults, and trying to impact my ability to make a living. Let me spell it out for you; here's what Triton does that Easy Water doesn't, at least based on what I could see in their demo.

• 3D Wave displacement
• FFT-powered waves
• Ship Kelvin wakes
• Propeller backwash
• Wave simulations for given wind conditions
• Particle-based spray effects
• CUDA and OpenCL acceleration
• Dynamic LOD
• Planar reflections
• Foam effects
• Height and normal queries
• Visibility effects
• Shoreline blending

On the flipside, Easy Water isn't limited to Pro/Windows as a platform, but I couldn't see anything it does that we don't.

If all you need is some eye candy for shallow water, then the inexpensive solutions are perfectly appropriate and you should stick with them. Some of them are really good, and I actually think Easy Water is a tremendously great value if it meets your needs. Triton is filling a niche for companies who need real 3D open ocean simulations at high framerates using Unity Pro on Windows. We know that's a relatively small market, which is why the price is relatively high.

 

Amen.
Welcome on this forum, where most people thinks that a plugin over 150$ is insane.