PDA

View Full Version : Compare Grass Sasquatch versus FiberFX



simpfendoerfer
10-15-2009, 06:19 PM
I was just exploring grass, close up. That's how good I could get it with each plugin. Sas first jpg. Fiber effects second.

Biggest draw back for fiberFX was that I couldn't make the leafs fatter. It was already set at 10000000.

shrox
10-15-2009, 06:33 PM
The Sas definitely looks better.

RTSchramm
10-15-2009, 07:00 PM
What were the render times?

Rich

simpfendoerfer
10-15-2009, 07:09 PM
Yeah. Sas about 6 min. FiberFX about 30 min.

SplineGod
10-15-2009, 08:05 PM
Well obviously you should be using fiberfx! :)

simpfendoerfer
10-16-2009, 07:36 AM
I suggest distribution of tasks. Don't know how long NT will take to fix all of FiberFX.

After improvement FiberFX could take over long hair guides modeling.

The future Sasquatch-Core could do the hair rendering.

I am looking forward to be able to use all kinds of lights on sas hair and that sas hair shows in reflections..

ingo
10-16-2009, 08:53 AM
.......and that sas hair shows in reflections..

and in Fprime..........one can dream and wait.

jaxtone
10-16-2009, 08:54 AM
You must be kidding... 6 minutes in SAS and 30 minutes in FFX tells me something must be wrong and definitely not worth waiting for. As it is now thereīs a lot of tweaking and bug crashes to get things work the way you wanīt.

Iīve said this earlier and each time thereīve been a handful of people that want everything to be as it always have been, "slow and complicated". But since I am not interested in torturing myself with slow and jumpy software limitations I guess this is still a wish:

I like to have all program and plugin features delivered with "fantastic presets" containing loads of combinations ready to use. Of course editable combinations shall be optional for personal desires.


Yeah. Sas about 6 min. FiberFX about 30 min.

Hieron
10-16-2009, 03:25 PM
Yes, and add a "make stunning" button please



nice Sas render btw

simpfendoerfer
10-16-2009, 03:33 PM
Jaxton, I am not as pessimistic. With Core coming up there is an opportunity to get this figured out.

simpfendoerfer
10-16-2009, 06:52 PM
FiberFX came with good videos.

Yes FiberFX was a bit of an over promise I think. And I also suppose it will be quite sometime until it gets better. Therefore I just mention, it would be efficient if Worley and Newtek would split the task. Strand guides modeling Newtek, Fiber rendering Worley.

Maybe they are listening in. You never know.

Hieron
10-17-2009, 05:07 PM
I don't think asking for presets is too much to ask for. Of course Sas did come with several scenes that you can pick apart. I assume that Fiber FX did not?


Yes, but there will probably be those who do not upgrade for quite some time and it would be nice to have a WORKING version of Fiber FX in 9.6 - since it is supposed to be working. Right? ;)


I was referring to:


I like to have all program and plugin features delivered with "fantastic presets" containing loads of combinations ready to use. Of course editable combinations shall be optional for personal desires.

As it was in contrast to other people apparantly wanting quote: "slow and complicated"


It is a nice wish, my wish for a "make stunning button" is too. With FFX the issue is not with presets though, it is a blatant lack of stability. I feel it would be better if it would be production stable before worrying about presets, no?

Having FFX working well is something I really want too, when I finally have some time between projects I'll gladly join the beta program again and try my best to find bugs..

Hieron
10-17-2009, 08:29 PM
Yeah, I must say that I first scrolled straight down to the fixes concerning FFX in that 9.6.1 list. Only to find 1 measly thing..

Too bad, I could find myself using FFX in the future.. but not before it works reliably.

Jim M
10-18-2009, 01:46 AM
I think you should have used the same lighting for both scenes :)

simpfendoerfer
10-18-2009, 08:18 AM
I think you should have used the same lighting for both scenes :)

Unfortunately one can only use spot lights with Sas. So that I lit the ball with soft shadow lights separately.

But you may have a point. If I would have lit the FiberFX only with spotlights it might have rendered faster and may have been a bit crisper.

sandman300
10-19-2009, 11:31 AM
I've found that some combinations of settings in FFX make it take a lot longer to render than it normally would. This may be just a bug but I haven't been able to do anything that is reproduceable.

Also how are you making the fibers, looks like your just using the pixle filter. For what your asking for, I think you'd have more luck using the modeler side of FFX to make the grass and then send them over to layout.

Another way would be to use guides from modeler to control the FFX settings. The pixle filter seems to work better this way.

SplineGod
10-19-2009, 04:00 PM
One of the most frustrating things about FFX is that its problems seem to be like a dog chasing its own tail. Something gets fixed only to be replaced by a new problem. Then that gets fixed and then theres yet another problem. That gets fixed but then the same problems start popping back up again. This has been going on for way too long.

Mr Rid
10-19-2009, 07:15 PM
For my Sas grass, I have used no spotlights - only distant lights - and everything works fine.

Except Sas shadows wont work with a distant light. Your grass appears unnaturally bright without at least self shadows.

You can always light your Sas objects with Sas disabled while using any kind of lights and GI you want, and bake out a color UV image or sequence using the Surface Baking Camera, then apply the render in the color channel of Sas so any soft area or GI lighting will be baked right into the Sas, while saving on costly Sas shadow render time.

Mr Rid
10-19-2009, 10:34 PM
Thats good if it works for your client's purpose. I'd be curious to see the photo of real grass you are comparing. Is it a bright exposure? If I Google grass images, they all have more contrast and shadow than the example. If I were doing photoreal grass at this resolution I wouldnt render anything without shadows.

Other grass techniques I would use before messing with FFX-
http://www.newtek.com/forums/showthread.php?t=101604

http://www.newtek.com/forums/showthread.php?t=100274&highlight=volumetric+texture

Mr Rid
10-20-2009, 03:39 PM
... - no reason to alter a winning formula. :)

Right, my suggestions are not just to you but to anyone reading here who may ponder approaches to grass. Sas with no shadows is not ideal for every grass render.

Iain
10-20-2009, 04:21 PM
I don't think Saslite (or Sasquatch for that matter) is any good for realistic grass.
I wasted hours experimenting with it and then when I got something useable found I couldn't get it to match the rest of the render.

simpfendoerfer
10-20-2009, 04:24 PM
You were more successful with FiberFX? What did you use in the end?

SplineGod
10-20-2009, 06:11 PM
Well it must have been photoreal because you got a paycheck for it. :)

simpfendoerfer
10-20-2009, 07:19 PM
Would it be possible to elevate the argument a bit? We know it can be done with front projection of an image, even a photo texture if you wish.

It would be cool to be able to do it in one environment in 3D. That's why digital 3D is fascinating.

Mr Rid
10-20-2009, 07:30 PM
Which is pretty much the bottom line! As long as your clients are happy and you get paid... :thumbsup:

Well, not entirely. Your work represents you beyond the current paycheck. One client's satisfaction does not mean that other potential clients are satisfied with what they see in a demo. Personally, I have my own expectation for my work that has nothing to do with the client's, because I want the work to represent me for future jobs. I hope my contribution will represent the company I work for to help win future bids, but also hope it will rep LW to the CG community at large.

I worked on enough crappy direct-to-video movies where the client was satisfied, but I would not present the same work on a reel, and in some cases would not want my name associated with it at all. On the other hand, I might deliver results beyond expectation where the client would have been satisfied with less, but I was putting in extra effort for my reel. I was paid the same steady check regardless, but there is a greater payoff to consider.

Mr Rid
10-20-2009, 07:55 PM
Would it be possible to elevate the argument a bit? We know it can be done with front projection of an image, even a photo texture if you wish.

It would be cool to be able to do it in one environment in 3D. That's why digital 3D is fascinating.


Yes, I would always suggest using stock images with mapping trickery and compositing where possible, if CG can be avoided.

Mr Rid
10-20-2009, 11:58 PM
I don't recall saying that EVERYTHING I do I would put on a show reel.

Never said you did.



When you are working for yourself in your own company and the client LOVES your work and you are satisfied you have done a good job... THAT is pretty much the bottom line.

Sounds more like it.



Mr. Rid.... try to stop nitpicking what everyone says. PLEASE?

Not everyone, just you Meggy.

Iain
10-21-2009, 01:36 AM
I think Mr Rid is making a relevant point.

If the discussion is about whether or not you can satisfy clients' expectations then the example posted by Megalodon (and the results I achieved) would be just fine.

It seems more the case though, that we are trying to decide if truly realistic grass can be obtained using LW's fur options and as far as I'm concerned, nothing posted here or elsewhere has shown that.

I can get decent stuff from Sas on a 2'x2' patch using lighting that looks awful on anything else. FFX results just elude me on anything other than tinsel simulation.

jaxtone
10-21-2009, 02:25 AM
As Mr Bowie once said:

An artist is never bigger than his latest hit!

Time flies!

Jim M
10-21-2009, 10:24 AM
My main issue with FFX is the way it works out how thick the fibres are. It is ridiculous. Sasquatch on the other hand, is as it should be, and this fact alone mean Sas wins out of the two. It is such an essential parameter for controlling the look of the fibers, that without the ability to control it properly, the tool is nigh on useless for grass, especially in animation.

I believe that you can create photoreal grass with Sasquatch, you just have to have the camera the right distance away... :) Once you have discovered that, the setup is simple. Close = Geometry / Middle = Sasquatch etc / Distant = Texture

And Megalodon, though your points are perfectly valid, it is still 'your' bottom line. You can't make it everyone elses, no matter how hard you press SHIFT :)

Jim M
10-21-2009, 02:20 PM
I am not disagreeing with your point of view. Though I also think alot of people do crap work and get paid for it, and continue to do so. It's all very subjective. We are all different. + the flipside is true, my brother is a genius when it comes to gfx, serious genius, he is crap at getting paid. It's a wide skillset to be successful, and is nowhere near limited to or as simple as 'do good work get paid'. Just my experience.

+ On a more pedantic level :) .... I think if you are going to fervently state things as facts, then you should make sure you get them right. Personally I am neither independently wealthy nor do I need money to survive. So there. :p

Peace out brother.

Mr Rid
10-21-2009, 07:33 PM
Actually... who cares? I never said ANYTHING about a demo. Why would I put crappy work into a demo? You decided to add that irrevelant factoid to the conversation - which.... has NOTHING to do with "the bottom line.":)

It has everything to do with it, you are just not understanding. 'Who cares?' about the quality of the work they do?! Its as if you are saying that the worth of a render is summed up only by whether or not the client liked it and paid for it. But I find that how you feel about your work is the bottom line because the images represent your abilities at large which will lead to future opportunities, beyond the last client or paycheck. I dont imagine you posted your example of Sas grass just because a client liked it and paid you for it, but rather because you felt it looked good. I commented that there was room for improvement and this set you off apparently. But I was also pointing out the limitation of distant lights in Sas for anyone who may not be aware.

I doubt anyone here thinks your example looks photoreal. All I can guess is that the photo you compared to is an odd exposure. If I Google 'grass' I can not find an image that is that soft and lacking shadows or contrast. The walkway also has a repeating pattern in the texture that makes it obviously artificial. I assume you may be doing arch exteriors where grass and sidewalks are not the focus of the render. But that image does not typify the expectation for photorealism that the clients I typically deal with have. Mr Schultz even joked about how it must be photoreal if you got paid for it, yet he did not come under Meg fire.

And 'crappy' is more often the result of a lack of time and resources, or a client who has differing ideas about what looks good, and is not necessarily that the artist or facility is incapable. And believe it or not, on some projects the clients may not even be concerned about the quality of the CG where it turns out to be a throwaway contract, obligations are just running out the clock, or ulterior agendas are at work and the post house is caught in the middle. But many times we are asked to steer a shot in a direction that makes it look worse to the entire department.. but that's what the stupid client wanted, and they are paying the bill. But the value of the render does not end there. Sometimes, we/I have time to go back and improve our/my own version of the shot to put on our/my demo, so we can hopefully lure more challenging jobs in future. The reputation for quality is a greater concern because our peers and other potential clients will judge our work in the end product or demo. No one prefers spending time on work that is ultimately not worth showing to anyone, regardless of the paycheck.

We may also sometimes turn down jobs that may be good money because we feel there is not enough time in the schedule to make the work as good as we want it to look. Again, meeting our own expectation for quality is more important than a paycheck. It can work against us to put mediocre work into a show regardless of the excuse, then have our name associated with something possibly subtracting from future opportunities. Our rep for quality is always the greater consideration, even while we may be a little desperate for work.

Most of us also know what it is to work below rate or for free because of an opportunity to work on something particularly interesting. Again, the bottom line is not a paycheck, but how we feel about our work.



The client likes your work... you get paid. Argue all you want "Riddy".... it still comes down to that. :)

I had agreed with how you last modified your bottom line statement to include "and you are satisfied you have done a good job." But then you argued against the same post for the second time and have returned to saying it is just all about a paying client.

'Hey Rid, everyone thinks your shot looks crappy.'
'I thought so too, but I was wrong because the client liked it and I got paid. Thats all that matters. It's going right on my demo.'

SplineGod
10-21-2009, 09:31 PM
I worked on enough crappy direct-to-video movies where the client was satisfied, but I would not present the same work on a reel, and in some cases would not want my name associated with it at all. On the other hand, I might deliver results beyond expectation where the client would have been satisfied with less, but I was putting in extra effort for my reel. I was paid the same steady check regardless, but there is a greater payoff to consider.


The only downside sometimes with going way above client expectations is that it now resets their level of expectation to getting way more then theyre are willing or able to pay for. :)
In that case it is sometimes better to wait until the job is done and delivered to go back and jazz it up for the demo reel.

SplineGod
10-21-2009, 09:40 PM
I can get decent stuff from Sas on a 2'x2' patch using lighting that looks awful on anything else. FFX results just elude me on anything other than tinsel simulation.

Believe me its not just you. FFX has been a frustrating moving target to use since NT adopted it.
Heres a good example of the current FFX workflow:
http://www.youtube.com/watch?v=vmhk38oq194

Jim M
10-21-2009, 11:04 PM
That is about right Larry. :)

Fiber FX could be great... if if if.

Enough of the pissing contest, you both understand each other, you're both clever and your both right. Show me the grass.

gordonrobb
10-22-2009, 02:40 AM
Can't you just light teh Sasquatch poly's seperately. So have lighting for it, the excludes evertyning else, and have lighting for everything else that excludes it?

SplineGod
10-22-2009, 03:05 AM
One thing Ive had pretty good luck with is using PDPros particle paint system to create plants etc with alpha maps to map on polys or sprites.

Mr Rid
10-22-2009, 03:14 AM
Can't you just light teh Sasquatch poly's seperately. So have lighting for it, the excludes evertyning else, and have lighting for everything else that excludes it?

Umm, eh? What problem are you addressing?

dballesg
10-22-2009, 03:33 AM
The only downside sometimes with going way above client expectations is that it now resets their level of expectation to getting way more then theyre are willing or able to pay for. :)


That is such a big truth.

jaxtone
10-23-2009, 03:04 AM
Recently I produced a house on speculation for a friends company and it didnīt take long before I understood that the real problem wasnīt the house itself, it was to match the house with the organic nature surroundings where "grass" are a main factor. FFX caused me problems with crashes and stuff and Sasquatch didnīt look realistic enough even if I did spend loads of time figuring out different ways to reach a desired level.

When looking at images created architects and building companies the grass many times look really fancy, soft and detailed and I guess not they have more expertise than the total intelligence in this forum. They might work in totally different softwares than what we are using but still, itīs 3D generated and probably have loads of presets and stuff because I guess not many producers have time to discuss why plugs and program features donīt work. The main thing must be that your software shall make your workflow frictionless, not the opposite.

What I would like to know is how these guys in general can produce really good natural environments when we are still in an argue process of why things donīt work as expected when you chose Lightwave as your 3D tool?

Well I guess someone would say, you get what you pay for, and probably there will be other voices raised about ignorant users like myself that donīt wanna spend a lifetime to figure out the "riddle" behind scripts and bugs, but still:

I canīt hear the grass grow!

Iain
10-23-2009, 03:30 AM
When looking at images created architects and building companies the grass many times look really fancy, soft and detailed............They might work in totally different softwares than what we are using but still, itīs 3D generated and probably have loads of preset

You are probably talking about VRay fur which works using displacements and is the standard these days for 3d grass in arch viz (and general arch viz) or Maya which has always had great hair/fur features.

It's hard to compare VRay with anything as it is so well rounded and advanced but we should by now have a fur generator which isn't limited in such a way that it looks like its development stopped in 1998.

Red_Oddity
10-23-2009, 03:55 AM
http://www.youtube.com/watch?v=vmhk38oq194

That was some of the funniest stuff i have seen this week. Made everyone here actually LOL.

SplineGod
10-23-2009, 05:12 AM
Yea it was funny and reminded me of the hell we went thru trying to use FFX on a show and then backing out and going to sasquatch. :)

Red_Oddity
10-23-2009, 07:25 AM
Come to think of it, a lot of productions we do feel that way with certain customers.

jaxtone
10-24-2009, 01:48 PM
If any one ever wonder why some Englishmen refuse to eat healty food this could be an explanation:

http://www.youtube.com/watch?v=NgsgPB0IXRY&feature=channel


That was some of the funniest stuff i have seen this week. Made everyone here actually LOL.

rednova
11-02-2009, 03:34 PM
Dear Friends:

I recently upgraded to lw 9.6
Without reading the manual, I was able to easily set up a fiberFX scene and it just took me a few minutes to do it. The render was really fast and it looks super. I showed my Mom a render with the rabbit from the SasLite scene..only I traded the sas settings for fiberFX. My mom really loved it. The render in fiberFX did not take long, and the results were amazing.
Just my 2 cents...

rednova :lwicon:

Hieron
11-02-2009, 05:12 PM
Hmm I suppose FFX is all clear then :)

SplineGod
11-02-2009, 05:23 PM
I wish the supervisor on the last project I worked on was rednovas mom. :)

Mr Rid
11-02-2009, 05:48 PM
Hmm I suppose FFX is all clear then :)

Absolutely. Rednova's Mom approves all of my renders as well.

Dont suppose we can see this super render? And how does it look animated in a practical instance?

jwiede
11-02-2009, 05:59 PM
The FFX problem must be entirely in SplineGod's head then, right? ;)

Hieron
11-02-2009, 06:29 PM
aww :)

Actually, I wished my mom would have been more positive on the images I made. Only thing I heard when I was younger was "hmm myeah, fine, so how did your exams go?"

:)

rednova
11-05-2009, 04:02 PM
Hi fellow Lightwavers:

This is my render of the rabbit using FiberFX for fur.
It rendered in less than one minute, and it only took 2 minutes to setup
the scene/FFX.
My mom really loved it. Not bad for a couple minutes work !!!

rednova

SplineGod
11-05-2009, 04:23 PM
The FFX problem must be entirely in SplineGod's head then, right? ;)

Ive been able to get a decent render or two from FFX but its like setting a bear trap with one hand. :)

cresshead
11-05-2009, 04:34 PM
ive been able to get a decent render or two from ffx but its like setting a bear trap with one hand. :)

lol!:d

Red_Oddity
11-05-2009, 06:02 PM
Ive been able to get a decent render or two from FFX but its like setting a bear trap with one hand. :)

I often think it feels more like placing my man bits inside the mouth of a sleeping lion while trying to flick a fly of its nuts with a wet towel (just when i think i have it i'm in for a world of pain.)

SplineGod
11-05-2009, 07:31 PM
I totally agree. Im sure NT will want to quote this in their advertising :)

Hieron
11-06-2009, 10:05 AM
I often think it feels more like placing my man bits inside the mouth of a sleeping lion while trying to flick a fly of its nuts with a wet towel (just when i think i have it i'm in for a world of pain.)


That's hilarious :)
So true



One time, it wasn't the problem to get the FFX to render, but to have it not crash when it didn't need to render FFX :) So for instance, that Rednova bunny was out of the view for a sec, *crash*. That bug got fixed after a while I think, but I had a hell of a time to keep "some" part in the view to circumvent it.

it's that odd stuff, when you slightly leave the beaten track of "import model, slap FFX on, render" that drives you nuts. Because you ofcourse did the testing with "import model, slap FFX on, render" and decided it was awesome and showed it to a client.

Then you move the camera, move the object, do something....



It would be awesome if it worked, and perhaps some bugs got fixed and it is quite good now. But untill a brave soul thoroughly uses it and can show some great moving (camera and object) shots, I won't use it soon again. Sadly that doesn't help with bugfixing but I'm short on time.

SplineGod
11-06-2009, 10:31 AM
Ive used FFX on some shots on show I worked on on moving
characters. Ill probably never do it again.

rednova
11-07-2009, 08:38 AM
Well, I tried to animate the rabbit with FFX, and the render got completely
screwed up. So I have to admit FFX is probably not as good as I tought.

rednova