PDA

View Full Version : Realtime Physics Forest for Lightwave



3dhotshot
08-28-2018, 04:22 PM
http://apagom.com/physicsforests/

http://apagom.com/physicsforests/videos/

They say it will be released as a native plugin for use inside of Lightwave ...

JohnMarchant
08-28-2018, 05:36 PM
But make sure you vote for it.

paulk
08-28-2018, 05:36 PM
And they recommend an Nvidia card > GTX 1080. Beer and cigarette money . . .

Bitboy
08-28-2018, 05:39 PM
Voted for Lightwave in the survey on the download page :-)

3dhotshot
08-28-2018, 06:51 PM
We need to be at the top of that list more votes lets go .... ^^ hate seeing C4d plugin monster as 2nd place kick them off ... the entire forum should hit that button

Oldcode
08-28-2018, 07:11 PM
Very cool! :D

Ernest
08-29-2018, 12:03 AM
Come on people!
This afternoon, we were #3 and now we're in the next to the last place!
But we're just 4 votes from #3 and, if everyone who voted on the Hub poll voted here, we'd be at #1 by a long margin!

stevecullum
08-29-2018, 01:12 AM
We’re back to 3....for now.

pinkmouse
08-29-2018, 04:43 AM
Voted, even though I haven't opened LW in about 6 months...

COBRASoft
08-29-2018, 07:43 AM
Second place again...

Axis3d
08-29-2018, 07:51 AM
My vote is in (LW, of course).

The good thing is that many of the other 3d programs already have options for doing this. So, maybe their users won't see as great a need as the LW community does.

ianr
08-29-2018, 08:24 AM
http://apagom.com/physicsforests/

http://apagom.com/physicsforests/videos/

They say it will be released as a native plugin for use inside of Lightwave ...


Can you post on lightwiki also?

Antonio De Lorenzo has a post up on lightwiki too

Photogram
08-29-2018, 08:45 AM
Voted!!!

Felipe Soto
08-29-2018, 11:07 AM
Voted! I want a mac version!

sadkkf
08-29-2018, 11:29 AM
And they recommend an Nvidia card > GTX 1080. Beer and cigarette money . . .

I don't think it's a recommendation. They say it won't run on < GTX 980

RPSchmidt
08-29-2018, 12:02 PM
I voted... would definitely be a huge plus for Lightwave!

OFF
08-29-2018, 12:05 PM
I voted twice! ) Why not? ))

shrox
08-29-2018, 12:18 PM
I don't see much description about it. What are the demo limitations?

kopperdrake
08-29-2018, 12:23 PM
We're at the top of the list! Any ideas on pricing?

squarewulf
08-29-2018, 01:00 PM
What are the demo limitations?

Demo appears to be literally just a demonstration. Resembles a bench mark test.

Airwaves
08-29-2018, 02:18 PM
I voted for Lightwave.

shrox
08-29-2018, 03:31 PM
Demo appears to be literally just a demonstration. Resembles a bench mark test.

Ok, thank you sir.

Dillon
08-29-2018, 06:42 PM
And some incredible demos they are! One of them being a beach with a sandcastle that erodes when water runs over it.

The sand simulations are as equally awesome as the water sims.

Have a 1080ti card, and seeing this happen in real time made my mouth agape for a solid 2 minutes while watching some of the samples.

Mesmerizing!




Demo appears to be literally just a demonstration. Resembles a bench mark test.

ianr
08-30-2018, 05:02 AM
And some incredible demos they are! One of them being a beach with a sandcastle that erodes when water runs over it.

The sand simulations are as equally awesome as the water sims.

Have a 1080ti card, and seeing this happen in real time made my mouth agape for a solid 2 minutes while watching some of the samples.

Mesmerizing!


I am happy if WE ALL VOTE for THIS (see its download page).

( Even if you are thinking about a new card)

But it's although very welcome ,it is a bit granular in the videos.

An interactive Foam from Granular morpher it is a dial needed, I

cite Flip- Fliuds here, so I hope they create this connection in the

next outing. Is there a Forum if any knows, to join for this?

mav3rick
08-30-2018, 05:18 AM
even though we are 1st i doubt theyll develop for so small userbase.... i bet they'll go for cinema

ianr
08-30-2018, 05:22 AM
2hrs ago 191 for LW 1st place and with only 50 to second Sooooooooo....

'Let's push this over the top' to quote Kelly on Lightwiki

Bernie2Strokes
08-30-2018, 06:34 AM
And they recommend an Nvidia card > GTX 1080. Beer and cigarette money . . .

Better than GTX 1080? But that's the card I'll be getting for my new PC next month. :( Can it still peform if I have a dual card SLI setup?

Update: scratch that. They recommend the 1080. But I can't try it until I get my new computer built.

motiondigital
08-30-2018, 08:07 AM
Voted

creacon
08-30-2018, 08:37 AM
I tried it with a 1080Ti, it works, realtime.

creacon


Better than GTX 1080? But that's the card I'll be getting for my new PC next month. :( Can it still peform if I have a dual card SLI setup?

Update: scratch that. They recommend the 1080. But I can't try it until I get my new computer built.

3dhotshot
08-30-2018, 10:00 AM
Since nvidia flex already works inside of lightwave, they can easily port this to lightwave without any issues ....

Oldcode
08-30-2018, 10:40 AM
And some incredible demos they are! One of them being a beach with a sandcastle that erodes when water runs over it.

The sand simulations are as equally awesome as the water sims.

Have a 1080ti card, and seeing this happen in real time made my mouth agape for a solid 2 minutes while watching some of the samples.

Mesmerizing!


Real Flow has a granular mode as well and it looks pretty much the same. The problem is, with both I think, is you need to up your scene scale a lot to make the fluid, particles, or whatever you're trying to do look good. That increases the number of particles you need, which in turn increases the simulation time.

When I first go into Real Flow back in 2012, I would import Poser characters. The problem was the polygons would be so small, the particles would pass right through the mesh. Only when I increased the scene scale, thereby making the mesh and polygons larger, would the particles interact. The minimum I'd have to increase the scene scale by was a factor of 8. It worked, but the fluid look terrible. Not enough detail.

With Poser characters, because they are so heavy with very small polygons, I'd have to increase the scene scale up to 25 to make it look decent, and it's still not quite good enough for my tastes. Obviously, that dramatically increases the simulation time. If this new software is as fast as it appears, that may no longer be an issue! :D

DonJMyers
08-30-2018, 02:00 PM
It is amazing but I can only make the first three scenes work. The other scenes cause a blank stage or volume. And then no matter what scene I click on I get a bare floor. Anyone else have these problems? Win 7 with an AMD proc and Nvidia geforce 1060 3 gig gpu.

AnimeJoex
08-30-2018, 02:35 PM
Just voted. And now they should close the poll and declare Lightwave the winner. :D

Qexit
08-30-2018, 05:25 PM
It is amazing but I can only make the first three scenes work. The other scenes cause a blank stage or volume. And then no matter what scene I click on I get a bare floor. Anyone else have these problems? Win 7 with an AMD proc and Nvidia geforce 1060 3 gig gpu.Maybe your graphics card is below the level needed for those other scenes ? A 1060 is the minimum required just to run the software at all and you only have 3GB of RAM. I can't even test it as I only have a 1050ti with 4GB :oye:

scallahan1
08-30-2018, 05:31 PM
Voted, and downloading shortly to check it out. We got us 222 votes so far. :)

Steve

ianr
08-31-2018, 09:38 AM
Voted, and downloading shortly to check it out. We got us 222 votes so far. :)

Steve


SUPA as in DUPA

lubor
08-31-2018, 11:58 AM
Thank you for your interest in our technology. We did not expect any kind of mobilization of communities to win our plugin poll:)
The plugin should eventually run also on old GPUs (only much slower). We just wanted the demo to run real-time. You can find answers to other frequently asked questions here: http://apagom.com/physicsforests/faq/

We do not want to deliver a product, which nobody wants to use, so we would appreciate, if you could answer some questions we have. We are not artists, so our opinions might be a bit skewed, and your answers would really help us in delivering what you want.

1. The quality of our simulations does not reach the quality of offline solvers (yet). What kind of problems or artifacts do you find the most disturbing?

2. What kind of features NOT included in this demo do you find the most important for you? (for example simulation on the top of the procedurally generated waves)

3. How important do you find two-way interactions between our fluid and other objects in the scene? (one way interaction = baked content inserted in our solver, for example that mutant in the demo). Such interactions (if not implemented in our framework) would slow down the solved due to CPU-GPU transfers and also slow down the plugin development.

4. How important is real-time rendering? I assume people would eventually render offline anyway. (we should speed up foam rendering, which gets slow when zoomed)

5. Our framework allows for choosing the speed / quality trade-off, but we would still prefer to know, what range of simulation size (number of particles) do you plan to simulate, and what run time do you find acceptable? The implementation of the algorithm is not very optimized, so there is a large space for improvement. However, at some point quality starts to matter more.



I can't even test it as I only have a 1050ti with 4GB
The real requirement of the demo is Pascal architecture (Compatibility 6.1), so you be able to run it.


I can only make the first three scenes work.
We allocate a fixed amount of memory at the beginning to avoid reallocation when new particles are created and we might have over-estimate the requirements a bit.

prometheus
08-31-2018, 12:40 PM
Voted both Lightwave and blender, though I may probably shift gear towards using blender more and more nowadays, it has a free fire and smoke..and if these two babes could work together a little bit..that would be sweet.

I havenīt been satisfied with what I have seen with hurleyworks up, or deepfx solver for fluids in lightwave, and I havenīt been satisfied with turbulenceFD for fire and smoke, thus a blender plugin solution would probably my first choice and Lightwave second.

I have yet to try the demo out, having an asus Rog G20B with a GTX 1080 card

rdolishny
08-31-2018, 02:06 PM
Voted now #1 very nice LW community...

Qexit
08-31-2018, 02:38 PM
Thank you for your interest in our technology. We did not expect any kind of mobilization of communities to win our plugin poll:)

The real requirement of the demo is Pascal architecture (Compatibility 6.1), so you be able to run it.Thanks for that info. I will try to get the trial downloaded and try it out over the weekend :)

Farhad_azer
08-31-2018, 03:37 PM
Voted. Thanks for informing us.

raymondtrace
08-31-2018, 04:28 PM
Thanks for tracking back to the poll mobilization headquarters. :)


...

3. How important do you find two-way interactions between our fluid and other objects in the scene? (one way interaction = baked content inserted in our solver, for example that mutant in the demo). Such interactions (if not implemented in our framework) would slow down the solved due to CPU-GPU transfers and also slow down the plugin development.

...

This is the great challenge and I assume it is what most seek. There are many solvers available and they often require baking. Having control between fluid and objects within the scene, in a primary application, is ideal.

There are even free (or very inexpensive) tools in Blender that can be baked but I'd still pay more for a tool that works directly within LW.

At the very least, developing a sensible round-tripping mechanism of baked content (not a true plugin that provides full control within the host application) would be useful.

jbrookes
09-02-2018, 09:51 PM
I agree. Two-way interaction is very important.

So for those who fill out the survey and feel the same way, please give support to two-way interactions. Something that could be toggled on and off (depending on CPU power or how much time you have to wait for calculations).

jwiede
09-04-2018, 09:09 PM
I agree. Two-way interaction is very important.

So for those who fill out the survey and feel the same way, please give support to two-way interactions. Something that could be toggled on and off (depending on CPU power or how much time you have to wait for calculations).

Unfortunately, the limitation there stems from LW not offering any APIs for interacting with LW's Bullet engine, so other than implementing RBD & SBD engines within the other plugin, the hope for two-way advection is long-standing but likely stuck awaiting LWSDK changes.

- - - Updated - - -


I agree. Two-way interaction is very important.

So for those who fill out the survey and feel the same way, please give support to two-way interactions. Something that could be toggled on and off (depending on CPU power or how much time you have to wait for calculations).

Unfortunately, the limitation there stems from LW not offering any APIs for interacting with LW's Bullet engine, so other than implementing RBD & SBD engines within the other plugin, the hope for two-way advection is long-standing but likely stuck awaiting LWSDK changes.

ianr
09-05-2018, 06:33 AM
Unfortunately, the limitation there stems from LW not offering any APIs for interacting with LW's Bullet engine, so other than implementing RBD & SBD engines within the other plugin, the hope for two-way advection is long-standing but likely stuck awaiting LWSDK changes.

- - - Updated - - -



Unfortunately, the limitation there stems from LW not offering any APIs for interacting with LW's Bullet engine, so other than implementing RBD & SBD engines within the other plugin, the hope for two-way advection is long-standing but likely stuck awaiting LWSDK changes.

Thank you John, a very good point. Would you mail it to Chuck as an update, one more time

JamesCurtis
09-05-2018, 07:16 AM
Lightwave was #1 as of this morning!!!!

ianr
09-05-2018, 09:39 AM
Any more you can Cajole James?

jwiede
09-05-2018, 09:54 AM
Ugh, sorry for the double-post, posted during the forum's "punishment period" and then got pulled away. No longer editable or I'd fix.

Philbert
09-05-2018, 06:04 PM
Dropped in my vote, we're still on top! This looks really cool. I don't see anything about a price though.

lubor
09-06-2018, 01:58 PM
I don't see anything about a price though.
We haven't decided on the price yet (the plugin is not even done), but there will be a different price for indies and companies.

paulk
09-07-2018, 10:43 AM
We haven't decided on the price yet (the plugin is not even done), but there will be a different price for indies and companies.

Does that mean different capabilities for indie and company versions?

lubor
09-07-2018, 05:22 PM
Does that mean different capabilities for indie and company versions?
capabilities that might end up different are not implemented yet (cluster support for huge simulations, multiple GPUs support on a single computer, etc)

Oldcode
09-07-2018, 07:17 PM
Ludor,

Will there be an adjustable viscosity and sticky settings, or will there only be pre-sets?

Thanks,

JamesCurtis
09-07-2018, 09:31 PM
Ludor,

I have a GTX 970 card in my machine. I know the demo will not work with that card. But, will the release version? I don't have a means right now to get a more beefy card.

ianr
09-08-2018, 07:52 AM
Lubor, Bear in mind that FlipFluids for Blender is of attractive price,
maybe a hire path would be attractive to indie wavers for top code,
meaning cluster support for huge simulations?
GPU should be in Indie in the LW World, it is in 'Turbulence' plug in
I have included some Pixs here omitted from my last p.m.
142776
Flip Fluids in action.
142777
How NOT to do it(Scale Mismatch)
142778
'Flowline' in action. How TO Do it! (clustered)
[copyright Universal Pictures 2012]

Hope they give some perspective, a standout popular plugin is needed.

lubor
09-08-2018, 10:44 AM
Will there be an adjustable viscosity and sticky settings, or will there only be pre-sets?
viscosity and sticky settings are adjustable


I know the demo will not work with that card. But, will the release version? I
yes, it will


Hope they give some perspective, a standout popular plugin is needed.
we will try to do our best

Oldcode
09-08-2018, 08:06 PM
Thanks Lubor! :D

3dhotshot
09-26-2018, 12:51 PM
From the Blender Guy FLIP solver creator .....

Hello Zahir,

I do not have plans to port the Blender plugin to Lightwave. At the moment, developing and maintaining the Blender plugin is full time work for me.

A real-time fluid plugin might come to Lightwave in the future called Physics Forests that is based on very advanced simulation research. The developer is gauging plugin interest for different platforms and Lightwave is currently in second place. You can place your vote here: http://apagom.com/physicsforests/downloads/

Best Regards,
Ryan


Yannik is back continuing his development process >> https://vimeo.com/291115445?


meanwhile lightwave has no real solid native fluid solver with white water simulation