emFluid4 released

[Hirazi Blue]

Eric Mootz presents a short preview video of his upcoming emFluid4.
(The successor of, you guessed it, emFluid3)
You can watch it on Vimeo

if everything goes as planned the release date will be mid-January 2011

Quick browsing help : The part of the thread talking about the actual release starts here

[Rez007]

That looks really nice!

[brudney]

yeah, that looks really really nice! great work Eric!

now only if we had some decent, fume-like volumetric shaders... oh man...

[Maximus]

Animation and particle simulations is really great but the shading and render is just ugly in my opinion its too "blobby" in some parts, doesnt enhance the product, yeah its still a sneak peek alpha, also i suppose Eric didnt spend lots of time on shading, but more on simulation, i know hes really great artist so it probably shouldnt be a problem for him to create some great smoke look.

Its sad to not have a decent volumetric shader with all those ICE going on lately, sadly you always compare fume and smoke effects to FumeFx look and its not that easy to match that quality.
Cant do much but hope for a new volumetric shader :/

Great job as usual Eric, thanks for putting efforts into creating those awesome tools!

/cheers

Max

[ace63]

Looks really great in my opinion - but just as everyone knows (except for Autodesk?) we need proper volume shaders

[Rez007]

Not to derail the thred any further, but it is true...Autodesk touts ICE as the best feature, but who in the world is going to switch over to using it, if you can't get a nice looking end result. In my opinion, this hurts people like Eric, here who is making something really awesome! Keep up the great work Eric, and hopefully AD realizes this long overdue issue and picks up the slack.

[CiaranM]

Not to derail the thred any further, but it is true...Autodesk touts ICE as the best feature, but who in the world is going to switch over to using it, if you can't get a nice looking end result.

That might be true if ICE were only about fluid sims. I use fluid sims with ICE too, but probably less that 5% of my work. Besides volumetric stuff, what else have you had trouble with getting a nice end result?

In my opinion, this hurts people like Eric, here who is making something really awesome! Keep up the great work Eric, and hopefully AD realizes this long overdue issue and picks up the slack.

I don't necessarily see that as Autodesk's job. Unless they implement some kind of extremely generic voxel primitive that any voxel based tool could plug into (that would be very nice, but more involved that a simple shader as is suggested). The voxel simulation and the shader are pretty closely linked, so that's why they are usually developed together e.g. Maya fluids, FumeFx, etc. - I wouldn't expect that from Autodesk. However, there's not much stopping a shader TD from writing a volume shader for emFluid right now - it works.

[Rez007]

Like I said, I didn't want to derail the thread any further from what it was intended, but since you replied to my post, I will do the same.

1.) My post was supposed to be relative to what was mentioned earlier - the current volumetric shader is not that hot in my opinion, and that is what was being used in the demo (I assume). Also, other non-fluid sims can handle a users base shader such as an Arch shader just fine, but that is irrelavent to what I was getting at.

2.) It never hurts to have a native volumetric shader that works well, I see your point though. However, now with AD owning/licensing Lagoa and touts ICE as the next best thing, then yes, they should be responsible to see that simulations should come out nice as an end result as well, not just as wireframe sims or screen captures to show off what it can do, and then expect people to use Arnold or port to Maya for a decent result, like most of the videos/tutorials seem to do. Maybe I am just not that good making relistic fire with the naitve Mental Ray shader...I don't know, maybe it is just me...

[origin]

Rez is right
now anybody know why there are no alternatives to fumefx
is this stuff hard to code/no papers available?

maya shader looks like little brother to fume, houdini results on vimeo looks blurred/no detail, big houses probably have own solutions for fire&smoke, phoenixfd lacks detail (grid method?)

Why nobody came up with solution close to fumefx
I'm talking about final look and easy of usage

[Maximus]

Rez is right
now anybody know why there are no alternatives to fumefx
is this stuff hard to code/no papers available?

maya shader looks like little brother to fume, houdini results on vimeo looks blurred/no detail, big houses probably have own solutions for fire&smoke, phoenixfd lacks detail (grid method?)

Why nobody came up with solution close to fumefx
I'm talking about final look and easy of usage

The only one who can top fumefx quality is DPIT for cinema 4D in my opinion, that plug in is brilliant.
I'm no TD so i cant say how hard is it, but there are some impressive stuff out there, Lightwave does some outstanding smoke too, Houdini is insane but its slow at sim it and hard to use, but it does top noch smoke quality, probably blender has some great smoke too.

[origin]

this?
http://vimeo.com/6266622
http://www.dpit2.de/features/vids/effex/effextv10.mov

Its still nowhere close to fume IMO...

fluid (water) parts looks like phoenixfd [grid?] (= bad)
smoke/fire its okay I guess for low end work...still not fume quality

for houdini I haven't found v.good fire/smoke on vimeo

unless you got some cool videos from experienced users

Moderator edit: please use the URL tags - HB

[owei]

Hi there..!

Well, Eric´s tests are rendered with the standard volume shader as he described in his comment at vimeo. I think, considering that, it looks quite nice. This volume shader is particle based only, that means it always produces something "cottonballs", except you use only a small amount of big particles and let certain fractal nodes do the rest of the density calculations.
Even "our" volume shader is grid/voxel based, as every volume shader out there (as far as I know..). The problem with the SI shader is, that is not able to "fill" the gabs between the particles in a nice and convincing way to create something like realistic smoke or fire.
What fume does, is filling the volume voxel grid directly with information directly from the fluid simulation and calculating/reconstructing further information on top of that (adding wavelet turbulence or something like that..). That´s possible with emFluid right now, as Ciaran already said. One has to code it, though
Fume produces great quality and it´s a tool that is on the market for quite some years now. The programmers got lot of experience that will be very hard to top...we will see and hope, that emFluid gets a nice shader one day ;)

cheers,
oliver

[Mootzoid]

Hey guys,

Thanks for the feedback. I'm glad you like the video and I understand the criticism concerning the shading... we all know how tough it is to get something volumetric with Softimage|XSI and MR, and having only one workstation for the rendering I was forced to have render times under 5 minutes per frame, which of course results in less quality.

So yes, the little shader problem is indeed a problem. I got some plans of writing an own volume shader, but that will take some time. Have never written a shader before ;)

Take care,
Eric

By the way: the release of emFluid4 will be 11.01.11 (day.month.year). It's a cool date as I think !

[Hirazi Blue]

Eric Mootz has created a new video specifically showing the new features of his new emfluid4.
You can watch this video on Vimeo

And to those waiting for this release:
[edit]it has been released already,[/edit]
Visit the product page
;)

[Mootzoid]

Yep, it's release

[Rork]

plugged on the SI mailinglist!

great stuff as usual Eric!

Only thing we can wish for this year is some kind of proper krakatoa/fumefx like volume shader for SI ;)

rob