PDA

View Full Version : New fluid soft



chris_lwgr
10-30-2014, 03:25 AM
Seems interesting--> http://www.52-technology.com/index.php

Netvudu
10-30-2014, 11:32 AM
Seems interesting but it looks too much like RF and it looks like it doesnīt have nodes. There is a reason for Realflow to jump into node-based workflow, just like Houdini and Naiad before. Menu-driven fluids has shown not to be the best approach.

Snosrap
11-03-2014, 09:18 AM
I thought I was looking at LW Core there at the end. :)

VirtualFM
11-21-2014, 09:34 AM
There are two kinds of people:
- those who get over-exited for little and start shouting "OMG, it's AMAZING, a toaster that toasts bread on BOTH SIDES at the same time!!!"
- those who quickly disregard something really outstanding as something that is not that important "Meh... landing a 3 foot object on a piece of rock 300 millions Km away? What's the use of that? Hey, look at that guys shirt, it's so corny... and offending, in a way"

Here it looks more like the later. Look how quickly you undervalue this new program just because the interface "looks like" Realflow. (Well, it really looks like Realflow, but maybe that's on purpose, to ease-in the learning curve, as most people interested in this stuff already know Realflow). As for the rest, I don't know the simulation times (on that village setup seems to be something like 2.5 minutes per frame), but the rest is quite outstanding:
- the turning washing machine would have particles flying all over the place, and probably a crash, in other programs
- 100 million particles?! I can't even figure out how that's even possible in one simulation without growing very old waiting for it to finish. In terms of comparison, loading a 8 million particles data into lightwave (already calculated by realflow) takes about 20 to 30 minutes. So dealing with 100 million, whatever it's state, is something that makes me turn my head into.
- simulating liquid through that complex messy tangled mesh sounds like a nightmare with the tools I know right now. I wouldn't dare to approach a project that required something like this!
- the village flood looks fast enough to calculate
- the mesh building looks consistent
- etc

So, even not knowing the simulation times, it has some great looking examples, solving complex problems that are not easy, and certainly not available for the common people with "out-of-the-box" software. So, why so much diminishing the deed?!

Netvudu
11-24-2014, 05:10 PM
Simply because I work my sims on a daily basis with Houdini for stuff like this:

https://vimeo.com/111436638

and that one ainīt alpha or beta but simply production-proven and I can use it for any commercial work for less than 200 dollars. Itīs not a promise of a software, but something that already works right now and really good, with nodes, render engine and all the bells and whistles.


Thatīs the reason when I see some other engine thatīs atmost similar but looks like a future version of Realflow Iīm not impressed. I like Realflow, and I work with it regularly. It got much better since it added nodes.

VirtualFM
11-25-2014, 03:13 AM
Great simulation you have there, congratulations.

Well, we know how capable Houdini is. I think the point here is "how long does that take to calculate"? An without glitches and particles jumping all over the place. One of the interesting things with this new fluid program is how quick it seems to solve the simulation. So I guess until we can get the hands on it and make a few tests (making a similar simulation in it, Houdini and Realflow), we can't say how better it really is.

Netvudu
11-25-2014, 10:56 AM
Just for clarification. Thatīs not MY simulation, thatīs the kind of work I do regularly (maybe not THAT nice).
I agree with you that until we try it we cannot know how good it is. Iīm just explaining that my lack of interest comes from my fluid needs being well covered by Houdini FLIPs.