is lagoa dead?

[gfxman]

i got this email from the lagoa newsletter i signed some time ago:

Thanks for hanging on for so long. I appreciate your interest in Lagoa.
It's been a while since the last update, but I'm now happy to announce a new project that I've been working hard to bring to you.

The state of physics simulation changed considerable since Lagoa came out. I'm quite happy to see so many new tools that make simulating Visual Effects more real.

However, I realized that the problems in the computer graphics industry are much deeper than the tools themselves. So based on meeting many of you and hearing feedback I decided to make something to truly improve the experience of creating 3D content.

Back in 2010, shortly after launching Lagoa I realized how difficult it is for an artist to use a physics software like Lagoa. First you need an initial investment of thousands of dollars for equipment and software. Secondly you need the investment of time to learn a tool like Softimage. Finally to pull it all together you need to buy and learn yet another rendering software... so you can make it look good.

Taken together, this makes the art of 3D content creation inaccessible to most users.

The new kid on the block - TeamUp
TeamUp was born out of the combination of efforts between me and Arno Zinke whose reputation on Optics simulation speaks for itself. Our mission is to make 3D content creation, and the results - like real life.

I've put together a team of very talented Software Engineers, Researchers and Digital Artists to build a new set of technologies that make content creation much more fun.

TeamUp is a new kind of platform that lets you bring in your assets, then use physically based cameras, materials and lights to interactively edit your 3D scenes.

It would be easy if we had stopped there, but we decided to make it truly collaborative. You can edit 3D scenes simultaneously with colleagues, or show your clients and get feedback on your work from any device, from anywhere.

As valued Lagoa users, I'm extending an invitation to you, to participate in the beta and launch of the TeamUp platform. Please register now for the beta.

There's a lot more to TeamUp and the Multi-Optics technology, we hope you like it.

Siggraph 2012
If you are coming to Siggraph, signup to get details on our exclusive demos and after-show parties!

best regards,



Thiago da Costa

CEO / co-founder
TeamUp Technologies
getteamup.com

I imagine that this will be a standalone application, and that lagoa will have no more devolpments.....

[bottleofram]

Opportunities around the cloud technologies are hard to resist. Anyone got on the beta yet? Interested in first impressions if they are allowed to be shared.

As for Lagoa, to me it doesn't seem very likely it'll be updated any time soon. I would assume most users would like to see a faster lagoa which is hard to achieve without moving computation to GPU. He probably figured its just too big of a hassle to deal with Autodesk and their snail-like development. Who can blame him?

[Hirazi Blue]

As far as I know the beta hasn't officially started yet. On trying to sign up for the beta I got a mail with the message "We will follow up when the Beta is ready for you". I'm guessing they do their big reveal at Siggraph and after that will get around to the beta test itself.

[Maximus]

"Back in 2010, shortly after launching Lagoa I realized how difficult it is for an artist to use a physics software like Lagoa."

Funny cause when i was saying this, people started bashing to death! ICE is for artist! Easy to use!
Doesnt really look like
Anyway it was inevitable, cause it was too complex and slow, so i'm not surprised "its dead". Even if some people probably use it for something.
But i never saw so many things done with Lagoa except the usual "r&d" or some videos on vimeo/youtube.

[iamVFX]

Finally to pull it all together you need to buy and learn yet another rendering software... so you can make it look good.

I like how Thiago is pointing out that built-in version of mental ray can't produce good looking images

[Pancho]

Ice is definitely for artists. Fluid sims are for ..., nerds?!?! The horrible thing is that none of these packages offer presets for real world settings. Even realflow didn't offer water presets.

You'll always get funny excuses that everythings depends on scale, density of your mesh or the particle simulation which I think is bollocks. Scale should be everything which should be necessary for the artist to choose.

Disappointing is Thiago himself admits that his own software is a pain to use. Of course it's very versatile, but a which price?

He should have been able to make it user friendly. That's the pitty.

[ActionArt]

I've used it a fair bit and I don't find it exceptionally difficult. I think the main drawback is just speed. It takes so long to preview results. I think this goes for any fluid sim at the moment. It just needs some serious GPU acceleration and it would be great

[EricTRocks]

Ice is definitely for artists. Fluid sims are for ..., nerds?!?! The horrible thing is that none of these packages offer presets for real world settings. Even realflow didn't offer water presets.

You'll always get funny excuses that everythings depends on scale, density of your mesh or the particle simulation which I think is bollocks. Scale should be everything which should be necessary for the artist to choose.

Disappointing is Thiago himself admits that his own software is a pain to use. Of course it's very versatile, but a which price?

He should have been able to make it user friendly. That's the pitty.

Well when you're dealing with simulations that need to be life like and physically accurate you need to adhere to certain parameters to get it to look that way. Not to mention "good" simulations aren't just one click things. They are going to take time to develop and tweak and customize. If the whole industry just wanted one button clicks for everthing, we'd all be out of jobs. I'm not saying there shouldn't be easier interfaces and presets though.

Realistically, animation is technical. The more sophisticated effects you're going after is going to escalate the knowledge requirements on the users end. I think what isn't taken into account is that Lagoa is not just fluid simulator. It's an uber "everything" simulator. The fact that he packaged it up into a single simulator should be taken into consideration in regards to how easy it is to use. I'm no Lagoa expert myself have only dealt with it a hand fulllof times but regardless the above still holds true.

I do believe that more work should have been done to improve performance but who knows what has been going on in Thiago's world. The cool thing about the TeamUp stuff is that its going to be cloud based and seems like it will be stand alone as well which makes me hope that it will be more cross platform and be able to be developed more rapidly with less concern about which 3d app its running in.

[gustavoeb]

Thiago certainly is a very clever guy, and I bet so is everyone in his team. But I really wonder how can one program change the whole paradigm of 3d simulation software. I mean, making something easier is all about the design decisions that guide your software. Most (but not all) of the times hiding complexity means less customizability (think of instagram) and results in everything looking the same...

+ one gets that feeling of "oh no, one more raytracer..." it feels like eventually many rendering companies will go bankrupt as it is not possible for so many of them to coexist

[EricTRocks]

Depends on your target audience for sure. If you're aiming for every knuckle dragging goon who can use Photoshop to be able to make realistic simulations and customize it with a few clicks, yeah you're going to end up with an Instagram type setup. I'd rather see very smart people who understand the complexities of physics, mathematics, and calculus get paid crazy money to make the cool effects than fledgling photoshop artists who get paid a decent amount for doing nearly nothing.

Sure renderers are becoming a hot new thing that seem to be popping up everywhere. I don't think this is a bad thing because a long time it seems that there was really only 1 dominant one where you didn't really have a choice in what you could use. More options will mean more innovations as we can already see as the tide is shifting away from Mental Ray.

[Kzin]

Sure renderers are becoming a hot new thing that seem to be popping up everywhere. I don't think this is a bad thing because a long time it seems that there was really only 1 dominant one where you didn't really have a choice in what you could use. More options will mean more innovations as we can already see as the tide is shifting away from Mental Ray.

innovations are a big problem because you have only 2 big players, renderman and nvidia arc. if you want to code a new, fast production renderer you have to go to these two and ask for permission to use their tech. a good example is sonys arnold version which uses nvidia arc's rqmc sampling (aka unified sampling). nvidia take over mi because of all the patents mi owned.
i would be aware of saying mr is old because there is alot of stuff going on (check out some tech stuff nvidia owns, take a look an some annoucements or take a look an speedtrace and mental core).
otherwise, all the patent means that other companys have to develope new techs, which is often not easy by very basic tech stuff patents.

the most renderer that comes out are truly brute force path tracer which are not production ready for vfx. sure, you can render nice archviz in 8 hours on your gpu, but try the same with a vfx scene und you will fail. coding a production ready vfx renderer is a big challange, look at arnold. they are doing a great job, but it took some time to make it ready for the masses.

anyway, lets see what thiago's team is working on.

[luceric]

Depends on your target audience for sure. If you're aiming for every knuckle dragging goon who can use Photoshop to be able to make realistic simulations and customize it with a few clicks, yeah you're going to end up with an Instagram type setup. I'd rather see very smart people who understand the complexities of physics, mathematics, and calculus get paid crazy money to make the cool effects than fledgling photoshop artists who get paid a decent amount for doing nearly nothing.

the holy grail in simulation is artistic directability and interactive previewing that will scale to the higher res. there is going to be some learning curve, but you shouldn't have to understand calculus to get the shot. the math isn't suppose to be your problem.

BTW, at places like digital domain and ILM, there is one general fluid or volume modeling tool written by one or two person, and then dozens of artists using it with a simplified interface with a few dozen sliders and edit fields to enter python expressions to do something special.

[Nizar]

Finally to pull it all together you need to buy and learn yet another rendering software... so you can make it look good.

I like how Thiago is pointing out that built-in version of mental ray can't produce good looking images

As you did, I noticed this point

[Mathaeus]

Finally to pull it all together you need to buy and learn yet another rendering software... so you can make it look good.

I like how Thiago is pointing out that built-in version of mental ray can't produce good looking images

Well, good PR work Today, it seems to be some rule to show 'love and understanding' to all users... Somehow different than in good old times of XSI4 and old XSI Base.
Exactly for Lagoa, I think meshers are more important thing...
For smoke sims, I have impression that old-fashion renderers, like Max default scanline, or Blender default renderer, somehow are better 'drawers' for all that voxel data created by simulators.
Another interesting thing, I think now there are two public available renderers, exactly 3delight and VRay, able to calculate motion blur directly for ICE particles and strands ( no need for point velocity attribute) - which is great advantage for motion graphic work or like. Out-of-the-box render engine... still nothing...

[EricTRocks]

Depends on your target audience for sure. If you're aiming for every knuckle dragging goon who can use Photoshop to be able to make realistic simulations and customize it with a few clicks, yeah you're going to end up with an Instagram type setup. I'd rather see very smart people who understand the complexities of physics, mathematics, and calculus get paid crazy money to make the cool effects than fledgling photoshop artists who get paid a decent amount for doing nearly nothing.

the holy grail in simulation is artistic directability and interactive previewing that will scale to the higher res. there is going to be some learning curve, but you shouldn't have to understand calculus to get the shot. the math isn't suppose to be your problem.

BTW, at places like digital domain and ILM, there is one general fluid or volume modeling tool written by one or two person, and then dozens of artists using it with a simplified interface with a few dozen sliders and edit fields to enter python expressions to do something special.

Well wasn't necessarily meaning that you HAD to know calculus to do sims. Hell I do them from time to time and don't calc. I was meaning that for the more intricate and realistic effects that people who have a more expansive knowledge on the subject and know how things work behind the front end UI will be able to get out of it what is desired and hopefully be paid for knowing how to as well. Those dozens of people won't necessarily know how to add a new feature or tool that is needed. Those 1 or 2 people you mention will. Those are the ones I'm talking about.

[Chinny]

Update

To answer the thread question - Autodesk have the rights to continue Lagoa MultiPhysics, but as was mentioned in this thread Thiago has been working on something completely new. I left Autodesk last year to join Thiago on his new venture, knowing the amazing things he can do (because of course we worked together to get MultiPhysics inside of Softimage 2011.5).

The new Lagoa is a different to the MultiPhysics we know inside Soft. Lagoa MultiOptics deals with interactive photoreal cloud rendering, which is mainly targeted at Designers, Engineering and Architecture.

We have launched today, and yes this is a shameless plug ;)

More information on Lagoa MultiOptics
http://www.lagoa.com

Video
http://vimeo.com/63170024

I hope this gives some closure about Thiago and Softimage.

All the best

Chinny