PDA

View Full Version : Feature Request < NVIDIA PhysX FleX >



OnlineRender
12-11-2014, 10:44 AM
I don't ask for much ... pretty please implement this.


http://youtu.be/1o0Nuq71gI4

ianr
12-11-2014, 11:12 AM
I ssem to remeber that LW3Dg have an Ex-Nvidia guy now onboard.

Surely he could help throw light on the possibilty of getting this into

LW 2015.5 etcetra?

Oedo 808
12-11-2014, 11:46 AM
I ssem to remeber that LW3Dg have an Ex-Nvidia guy now onboard.

Surely he could help throw light on the possibilty of getting this into

LW 2015.5 etcetra?

I'm not entirely sure but I think he disembarked not too long after coming aboard. Though I could be getting him mixed up with someone else.

Kaptive
12-11-2014, 12:45 PM
It would be good to know if it was possible to get this data into LW. If we could tap on the resources of the graphics card for fluids that'd be great. I guess it might have problems interating with other FX solutions like bullet though... just a guess. But really, if LW3DG could get this in, then a very fast liquid solution would be in reach without massive investment.

Thing is, even if it wasn't able to interact with other dynamic LW items, I'd still personally prefer to have it there than not.

ianr
01-02-2015, 09:53 AM
Yep Kaptive,
" without massive investment" send that part to Mr.Powers...
As a extra fillip Bullet V3 will be GPU'ed, so we hear. All good stuff to
slip Mr.P into a road-map CEO decision, me thinks?

vonpietro
01-03-2015, 01:39 PM
Wow, thats some nice realtime effects stuff. If thats the future of game stuff - were going to be wasting alot of time.

this nvidia stuff answers the problem of weak smoke features in lightwave - "everyone uses houdini, so dont bother with giving lw a fully featured smoke sim?" might be what they are saying, houdini is so far ahead, dont bother? Yet turbulencefd is nice but it's an add on, and doesn't solve all the smoke needs. As far as that thing goes it helps, but its also not the easiest thing to use.

Being able to do nice smoke effects is some artists main focus - i see some reels that are beautiful with pyro and smoke effects that lightwavers just drool over.

as smoke is used in an amazing amount of films and television (its in almost everything)- why we dont have a good implementation is like WHY? Its something that is left for other programs, and that is why lightwave loses a huge marketshare i think. Has houdini and 3dstudio max cornered the market with smoke effects work? Is newtek ever going to put in a decent smoke feature?
HV are dated and slow, and suck up alot of time trying to get mediocre results. and thats time i'd rather be finishing my shots not troubleshooting why HV isn't doing what i want it to do.
You can't do the movie Pompeii with HV's. Not in any reasonable time frame anyways.

its something everyone would use all the time me thinks.

my 2 cents.

prometheus
01-03-2015, 08:08 PM
Wow, thats some nice realtime effects stuff. If thats the future of game stuff - were going to be wasting alot of time.

this nvidia stuff answers the problem of weak smoke features in lightwave - "everyone uses houdini, so dont bother with giving lw a fully featured smoke sim?" might be what they are saying, houdini is so far ahead, dont bother? Yet turbulencefd is nice but it's an add on, and doesn't solve all the smoke needs. As far as that thing goes it helps, but its also not the easiest thing to use.

Being able to do nice smoke effects is some artists main focus - i see some reels that are beautiful with pyro and smoke effects that lightwavers just drool over.

as smoke is used in an amazing amount of films and television (its in almost everything)- why we dont have a good implementation is like WHY? Its something that is left for other programs, and that is why lightwave loses a huge marketshare i think. Has houdini and 3dstudio max cornered the market with smoke effects work? Is newtek ever going to put in a decent smoke feature?
HV are dated and slow, and suck up alot of time trying to get mediocre results. and thats time i'd rather be finishing my shots not troubleshooting why HV isn't doing what i want it to do.
You can't do the movie Pompeii with HV's. Not in any reasonable time frame anyways.

its something everyone would use all the time me thinks.

my 2 cents.


I canīt speak for the lw team, but I am not surprized if they simply thought it would take to much time to develop at the time ..some years ago when fluids popped up, and then jawset and turbulenceFD showed up, and I am sure they thought there was better ways to spend their resources when jawset was doing his thing, I recall someone saying it would require a full time assignement only for fluids, or something like that.
I simply donīt think they have the man resources to implement fluids nativly, and they would need to go scouting for a very skilled person doing his own fluid solution, then also make sure it can be ported to Lw rather than start from scratch, but
whereīs the incitament to do that when Jawset is already doing his thing?

I see some issues with a third party solution...

1. if jawset dropīs it, or get ill, injured etc..and thus canīt continue development, in such case lightwave will really fall behind in that area.

2. jawset is working with multi platforms, like cinema4d, after effects, lightwave..and now possible modo, I am afraid that will slow down the fluid development for the plugin, both for lightwave and for other platforms too, not sure if he in such case can expand on more employeeīs than himself.

3. as a third party plugin, it isnīt developed with focus to work with other lightwave items as a Main objective, if a fluid tool were developed within lightwave, I think the development process could be guided to work just like that, meaning it should be developed with many other lightwave tools in mind, particle advection...I am still looking for that in lightwave version, it has been there for a long time with cinema4d.

Though I would gladely see some Lightwave Inhouse fluids, I suspect the lightwave team donīt have the resources to do that coding, besides...I think they need to get other dynamics workin proberly with each other, such as the particle system, legacy or a new one to work with bullet properly, the same particle driven by bullet forces, or the same particles hitting and affecting bullet parts...should be made working with fluids, as fluid source..but also making the actual fluids particle advect and push the particles...so a fluid simulation is driving the particles that in the end are hitting bullet items.

Hypervoxels...yeah we all know about that one, I had my hopes up for 2015, but unfortunatly not..I am now comparing modo volume item against hypervoxels, and I will get back to that later, first impression, modo volume item has much better smoothness or density thickness that fades out much better than lightwave voxels, not even with dissolve local gradients can I get that smoothness, the light scattering seems much better too, rendering speed..I am not sure, I think modoīs previewer actually is slower than VPR in showcasing the volumetrics.

Setting up modo volumetrics, well..I prefer lightwave hypervoxels menu, than adding noise and density functions the modo Item shader way, but...I suspect modoīs volumetrics are more flexible in terms replacing/drive certain things thanks to the nature of the shader tree, but as I said..also requires more to understand and get going with.
I am just in the beginnings of testing modo volumetric items...so there is a lot left to try out.

The physX-FleX...is looking awesome, but I think it is just a silly dream thinking it will be adapted by the lightwave group, it is a tool that seem to work on most possible dynamic forces itemīs there is, in conjuction with each other, wether or not it is liquid fluids, smoke fluids, rigidg bodies,fratures, particles etc..all possibilities to drive Any dynamic item with wich ever dynamic item there is.

thatīs what the Cor.........Stop, not going there:)

I think focus now would be to keep working on bullet, and get particles to work with that in all kinds of scenarios..forces and with rigid bodies and fractures, I think the new bullet hinges,motors, bone dynamics are great..though I felt it was unstable in the first release...it is a good direction.
apart from getting particles in there, continuing making soft body, or a clothfx tool working with bullet better, as well as improving on the fracture tool.