PDA

View Full Version : So what made SoftImage so special ? ( What can LW dev learn from it?)



robertoortiz
03-08-2014, 10:21 PM
As a fan of sotimage I want to use this thread to se what can the LW development community from Softimage.
SO tell us,
What made FaceRobot so special?
How about ICE?
And why people rave so much about the Sotimage Interface?


I am looking for your comments and tales.

-R

50one
03-09-2014, 01:14 AM
It's a great piece of software, once it was scrapped by Autodesk it just melted the hearts of some CEO's and exec's, so they now offering free courses, discounted software and other privileged to ex-softimage users':)

toeknee
03-09-2014, 06:36 AM
I can answer the ICE question from a uses perspective. I was using Houdini and Softoimage before they added ICE and the big deal was that they integrated tje idea of compound back when Houdini were the only guys on the block doing this. This made it very easy to create a tool and that let anyone use the same tool. Moreover if they wanted to tweak that tool all they had to do was drill in to it and change what they needed changed. The LW development team can take advantage of this by providing a huge list of functioal and flexible compound for Nodal.

mouse_art
03-09-2014, 07:27 AM
You can start with the Online Docs.
http://download.autodesk.com/global/docs/softimage2014/en_us/userguide/index.html
Or download the Free ModTool 7.5(which you can get from different Sites, which includes ICE.

No sorry, but if the Dev's had any interest to "copy"(or create inspired tools) some things from XSI they had already done it...

hrgiger
03-09-2014, 09:35 AM
one thing can be learned from Softimage is that if you want your application to still be around 5 or 10 years from now, dont sell it to Autodesk.

RebelHill
03-09-2014, 10:09 AM
Well... the BIG ticket item in SI was that everything was connected under the hood... You could take any one kind of scene/object data and feed it elsewhere, which allows for maximum control and creativity. This ofc wasnt limited to SI in any way... maya is no different. The primary distinction comes from the UI, best exemplified in ICE as compared to Mayas hypergraph.

What can LW learn from this... well, I fear, not much. You can have the shiny, fancy, "artist friendly" UI all you want, but unless its supported underneath by an interconnected system... you're gonna hit walls and limitations. And the only way LW is gonna achieve this level of sophistication and flexibility is (I imagine) with a pretty much complete rewrite.

hrgiger
03-09-2014, 11:11 AM
Well... the BIG ticket item in SI was that everything was connected under the hood... You could take any one kind of scene/object data and feed it elsewhere, which allows for maximum control and creativity. This ofc wasnt limited to SI in any way... maya is no different. The primary distinction comes from the UI, best exemplified in ICE as compared to Mayas hypergraph.

What can LW learn from this... well, I fear, not much. You can have the shiny, fancy, "artist friendly" UI all you want, but unless its supported underneath by an interconnected system... you're gonna hit walls and limitations. And the only way LW is gonna achieve this level of sophistication and flexibility is (I imagine) with a pretty much complete rewrite.

agreed. But how many times does it have to be said before it sinks in? People say give us ICE in LightWave like its just a simple feature request. Which is why without a complete rewrite of the architecture, LightWave (nor Modo) will ever allow the kinds of connections and access you could have in Softimage. If you ever wonder why some people supported the "idea" of LightWave CORE, just remember Rebel Hill's post.

As a sidenote, who knows where CORE would be today had it been continued? But I imagine that a unified applcation with complete nodal access to everything in the program and with a modifier stack and the excellent poly performance it had, it might be looking pretty good right now to XSI users looking for a replacement app. As it stands now however, I doubt LightWave is even on most of those users radar. which is unfortunate, but it is what it is.

zardoz
03-09-2014, 11:48 AM
To me simply having a way to organize your scene and having render passes is a win.

robertoortiz
03-09-2014, 03:28 PM
Im am posting some videos to aid the conversation:

Softimage Face robot:
http://www.youtube.com/watch?v=sUaSSooVz8Y

Softimage ICE examples:
http://www.youtube.com/watch?v=B-rB5mIOz98

Softimage Rendering Passes

http://vimeo.com/25858011

hrgiger
03-09-2014, 04:02 PM
Here are some recent ones made by former LightWaver Pooby (Paul Smith)

Amazing deformation system he made in ICE:

http://vimeo.com/88077743
http://vimeo.com/88245138

Making a tree in ICE:

http://vimeo.com/76144838

Driving shapes with ICE:

http://vimeo.com/40524205

Then my personal favorites, he simulates a zombie attack with ICE here. These are quite humorous:

Part 1 http://vimeo.com/21059614
Part 2 http://vimeo.com/21071207
Part 3 http://vimeo.com/21105308
Part 4 http://vimeo.com/21106696
Part 5 http://vimeo.com/21113240
And the last standoff here: http://vimeo.com/21164276

jeric_synergy
03-09-2014, 05:51 PM
For over a decade we've been saying "expose everything" to the SDK-- is this STILL not the case?

Is LW still full of weird cul-de-sacs and one-way data valves?

While a modern, flexible data architecture is a good idea, CORE was a stupid idea, because it basically threw away the entire LW interface. So they set themselves TWO difficult tasks, instead of just actually accomplishing ONE, and then evolving more advanced UI tactics to take advantage of it. Swing and a miss and game over.

RebelHill
03-09-2014, 06:02 PM
Ofc its still the same as it ever was... And it wasnt the wrong idea to change the interface... thats entirely the right idea, because the interface is an abstraction of the code functionality... you cant just shoehorn one into the other... else u get something akin to a fat guy in a little coat. The only stupid idea was all the crying for LW to change underneath, whislt staying the same on top. Its gotta change, BUT, its also gotta stay the same. Sorry, you cant have it both ways.

hrgiger
03-09-2014, 06:16 PM
Ofc its still the same as it ever was... And it wasnt the wrong idea to change the interface... thats entirely the right idea, because the interface is an abstraction of the code functionality... you cant just shoehorn one into the other... else u get something akin to a fat guy in a little coat. The only stupid idea was all the crying for LW to change underneath, whislt staying the same on top. Its gotta change, BUT, its also gotta stay the same. Sorry, you cant have it both ways.

Exactly. There's no way you could have set out to create the design goals of CORE (unified app with any property driving any other property and with non destructive workflow) without changing the way LW works on top. How that would work, I couldn't even imagine.

But again, I put "the idea" of CORE in quotation marks above to indicate that it wasn't about lamenting CORE or what it should have been or what it turned out to be, some of us just very much supported the idea of a rewrite of LW code to eliminate the problems that we've all been struggling with for years and of course to add in the added benefits as well. NT eventually decided they could do more or less the same thing by restructuring LW's code. We're just still waiting on that outcome. LW11 while chock full of useful additions showed no signs of it. Will LW12?

Oedo 808
03-09-2014, 07:29 PM
I'm quite amused by the idea of pursuing as the holy grail, the form of a software which could not generate sufficient interest not to be sold off and then to be discontinued. Don't get me wrong, there isn't anything mentioned I wouldn't want, but it does have a certain irony. It's not good enough to say "build it, and they will come" it was built in Softimage, and they didn't come, they said "we're quite comfortable in 3dsMax and Maya, thank you very much."

As regard to changing LightWave as they did with Core, I thought it both wreckless and unnecessary, Myagi showed quite well an example of being different yet remaining the same, why on Earth you would choose any other path is beyond me. People have been asking for dockable windows and such, I think there are a good many changes the userbase would be receptive to, just so long as it doesn't transform into a Playskool themed wet dream along the way.

I could dream that LightWave Core would have been Softimage MkII by now, and not only that it was, but was pulling them in from all over, and believe me I do. But I can also imagine that it would have been Softimage 0.5. maybe not even that and people were still saying "we're quite comfortable in 3dsMax and Maya, thank you very much." Except that on top of this what share LightWave did have would now be peeking over the fence at Modo, I know that's how it came to be on my computer and it was only the want of finding tutorials that hit the mark that meant it was only a peek over the fence.

I'm going to sound like a broken record but I don't believe that the decision to integrate Core into LightWave was anything more than a way of keeping additional software releases a viable option, completely apart from having LightWave still feel like home, I'm surprised that some don't feel that the Core promotion and subsequent 10 development cycle wasn't a warning that things couldn't continue as they were, but I've flogged that horse until it has become a skeleton.

The big fear of course is that the destination has changed, or that if it hasn't technically, that it's on the back burner to such a degree we're never going to see it. I worry about this, even if everything were going as planned it's quite plausible that something might not be quite ready to unveil just yet. But here is where it's so important to make the distinction between people who weren't happy with Core, were they just whiny fanboys? Or were they happy with the direction, just not so happy with the implementation? If it's the former then LightWave continues chugging along as it has been doing, people happy in their split app, slow arse polygon handling, never gonna see decent viewport shaders bliss, but if it's the latter, then the clock is ticking, and even for those Core naysayers, time is running out.

jeric_synergy
03-09-2014, 07:37 PM
Re the interface versus the underlaying code: I disagree that because what is essentially a database structure changes that the UI has to change. The structural changes would ALLOW ui changes, but do not dictate them.

And, I trusted Splinegod's take on it.

RebelHill
03-09-2014, 08:43 PM
I disagree that because what is essentially a database structure changes that the UI has to change.

You would be correct IF the architectural change were simply a database structure...

it is not.

warmiak
03-09-2014, 09:58 PM
Of course .. what else.

The way this stuff works (3d) has been based on well known algorithms and designs from decades ago.

Everything thing else is just a matter of properly design representation of the domain ( the database) and the UI on top of it.

hrgiger
03-10-2014, 08:09 AM
I'm quite amused by the idea of pursuing as the holy grail, the form of a software which could not generate sufficient interest not to be sold off and then to be discontinued. Don't get me wrong, there isn't anything mentioned I wouldn't want, but it does have a certain irony. It's not good enough to say "build it, and they will come" it was built in Softimage, and they didn't come, they said "we're quite comfortable in 3dsMax and Maya, thank you very much."



There are a couple problems with this line of thinking. First, while SI did not have numbers that Max or Maya have, it still has a considerably larger userbase then LightWave. More then LightWave and Modo put together. But when you stack even that number up to the numbers of Max or Maya numbers. that number probably looked insignificant in comparison. But LW3DG would be thrilled to have that many users. Second, Avid went dark for a considerable time while they re-wrote XSI. I imagine they dug themselves quite a hole that they never quite dug their way out of hence eventually selling to Autodesk. But its just not accurate to view XSI as not appealing enough software that 2 different companies couldnt make it work. it sounds to me more like Avid made some bad decisions and Autodesk never had any desire to keep it going.

And concerning where it would be now, even today CORE wouldnt have been a replacement for XSI, not even close. But with its architecture and desgin goals, it would have been well placed for it to be adopted by XSI users hesitant to use another Autodesk product but wanting eventually what CORE set out to do. And many XSI users will continue to use XSI for some time so there would have been more time for CORE to mature.

Im just hoping LW3DG can make the necessary changes to bring similar results.

jeric_synergy
03-10-2014, 01:16 PM
I'm still not sure the thread title has been answered: what is it that make SOME animators so thrilled about SI, and why can't the other apps (esp Maya but hopefully one day LW) accomplish the same thing?

IOW: I'm told it's intrinsically IMPOSSIBLE to accomplish certain tasks in LW because of the data architecture, that the data Just Isn't Accessible. And that in MAYA this is not so. So, was it even MORE not so in SI??

(After seeing what 3D Powers did in 9.6, NINE POINT SIX!!!, I have some doubts about what is and is not possible within LW.)

(poster ToeKnee got closest, but I didn't understand his post, I think he was using some SI specific jargon.)

hrgiger
03-10-2014, 02:17 PM
I'm still not sure the thread title has been answered: what is it that make SOME animators so thrilled about SI, and why can't the other apps (esp Maya but hopefully one day LW) accomplish the same thing?

IOW: I'm told it's intrinsically IMPOSSIBLE to accomplish certain tasks in LW because of the data architecture, that the data Just Isn't Accessible. And that in MAYA this is not so. So, was it even MORE not so in SI??

(After seeing what 3D Powers did in 9.6, NINE POINT SIX!!!, I have some doubts about what is and is not possible within LW.)

(poster ToeKnee got closest, but I didn't understand his post, I think he was using some SI specific jargon.)

Well aside from the fact that if you're knowledgeable about such things, you could use ICE in XSI to create your own tools. Don't like the way the bevel tool works in XSI? You can create your own via an ICE tree. But that's a very simple example. Pooby created a deformation system, a fur system, his own renderer, custom simulations, driving animation shapes.... The thing is, I think you are talking mainly about access via the SDK for 3rd party developers. But that's just one part of the open architecture. Users also have complete access inside the application. 3rd powers made some useful plug-ins for LightWave for sure but for an example, you do notice that it has to use its own timeline because apparently it must not be able to function with the use of LW's timeline.

But speaking for myself and my experience with XSI... The U.I. is very clean and very consistent across the application, context aware menus, many great selection options, everything was animatable inside of XSI including modeling operations (even down to primitive creation), Viewports could be anything including the explorer (sort of like scene editor in LW), gigapolygon core, quick rig creation including spline rigs (actually reminds me a bit of Rhiggit 2), modifier stack allowing non destructive workflow.. if you generate geometry from a curve (like a lathe or loft perhaps), if you edit the curve later, the geometry moves with it.... Mind you this was version 7 that I used and I only used it for a short time so I'm sure others could give you a more thorough summary of its advantages.

RebelHill
03-10-2014, 02:22 PM
I'm told it's intrinsically IMPOSSIBLE to accomplish certain tasks in LW because of the data architecture, that the data Just Isn't Accessible. And that in MAYA this is not so. So, was it even MORE not so in SI??

No... the difference between maya and si in this regard was the interface.

In maya, you can do all the same fancy stuff... but largely you need to do it in mel (or the hypergraph if you're really brave). In SI, there was ICE which allowed this in a simple, click n drag interface.

Oedo 808
03-10-2014, 05:05 PM
And concerning where it would be now, even today CORE wouldnt have been a replacement for XSI, not even close. But with its architecture and desgin goals, it would have been well placed for it to be adopted by XSI users hesitant to use another Autodesk product but wanting eventually what CORE set out to do. And many XSI users will continue to use XSI for some time so there would have been more time for CORE to mature.

Im just hoping LW3DG can make the necessary changes to bring similar results.

I've tried to chop this down as it was getting a bit, too long, didn't read. I can see why you lament the demise of Core, but do you think people would have to have been happy with not even 10.1, just more and more Core Betas for the HardCORE members?

You seem to be putting a lot of faith in the belief that Core would be showing, well a lot more than I think it would have at this stage, and just as there would have been time for Core to mature, there is time for LightWave as it is to mature. There is also the cost of not having anything meaningful added to LightWave in the mean time if Core had been continued as the focus.

I think that it won't be too long before XSI users start to at least consider their options even if they will be sticking to XSI for the time being, if the LightWave group can convince them that they can look at what LightWave has to offer now and see that moving forward will be what we saw a glimpse of with Core, that would be ideal. And although I don't think the 'drop everything for Core' approach would have got us anywhere, I wouldn't mind being convinced that this approach is indeed getting us somewhere.

But you know the LWG, they don't like to say much.

Blessed is he who expects nothing, for he shall never be disappointed.

Blessed is the developer who does not raise expectations, for they shall have no customer to disappoint.

And before I get blah blah Pixologic/ZBrush from someone, they raise my expectations whether they intend to or not, as could be said of the 11 development cycle for LW, except ZBrush doesn't have such fundamental questions hanging over its head which need resolving. I guess they could say that's already been answered with the 'same destination, different vehicle', but it wouldn't hurt to reassure me of that, from time to time, maybe there is some pitfall in doing it more widely that I don't see.

robertoortiz
03-10-2014, 06:35 PM
This article might be of interest for those familiar with Maya:
The truth about Softimage and Autodesk. Did Autodesk try?
Quote:
"At that very first view from a Maya user, seems that Softimage lacks a lot of stuff, without stupid icons everywhere. Maya artists that actually went to start using Softimage, and not overwhelmed by such a pristine UI never went back to Maya. I personally know several of them."

http://e-roja.com/the-truth-about-softimage-and-autodesk/

hrgiger
03-10-2014, 06:39 PM
I've tried to chop this down as it was getting a bit, too long, didn't read. I can see why you lament the demise of Core, but do you think people would have to have been happy with not even 10.1, just more and more Core Betas for the HardCORE members?

You seem to be putting a lot of faith in the belief that Core would be showing, well a lot more than I think it would have at this stage, and just as there would have been time for Core to mature, there is time for LightWave as it is to mature. There is also the cost of not having anything meaningful added to LightWave in the mean time if Core had been continued as the focus.

I think that it won't be too long before XSI users start to at least consider their options even if they will be sticking to XSI for the time being, if the LightWave group can convince them that they can look at what LightWave has to offer now and see that moving forward will be what we saw a glimpse of with Core, that would be ideal. And although I don't think the 'drop everything for Core' approach would have got us anywhere, I wouldn't mind being convinced that this approach is indeed getting us somewhere.



But I think this is where you misunderstand me. I don't lament the demise of CORE. What I lament is the wasted time. My time I spent collaborating, testing, making suggestions, making videos to contribute to the rewrite of LightWave. Also, the wasted time by the now LW3DG building an application only for them to discard it a few years later.

As far as what people would have been happy with, well people had their choice. They were told that this application was the future of the software. That it would continue to ship with Modeler and Layout until such time CORE could replace them. Obviously a bold decision that not everyone would be happy with but then again, it was a bold decision to go dark and recreate XSI into the application it is now. But its clear that a lot of the whole thing wasn't handled well from the start.

I do tend to think that a new architecture would have been quicker to mature then trying to mature current LW. LW needs big changes to do a lot of the things they would like to do with it. And that means breaking a lot. And that means long periods of time putting in new systems, and then going back and rewriting almost every tool/feature to work with that new system. But hey, that's just my impression of the situation and I could be wrong. I know that When Viktor put NURBS in LWCAD, he broke every single tool and had to rewrite them all to work with them. But I would argue in all this time since LW10 was released, that LW has matured very little if at all. I don't see any changes to performance or the way systems work in LW, only new features added on top. Of course, the hope is that they have been working on new systems and that it will just take a long time to make those new systems readily obvious in future releases.

But concerning XSI users, I fear that LW is not very visible on their radar as a suitable replacement. Maybe LW12 will attract more attention but LW12 seems to still be a long way off. And LW3DG isn't giving any indication of what to expect even after an XSI user asked for some indication of where LW might be headed. But LW3DG is so skittish because of broken promises in the past, they've gone completely tight lipped about future direction. Which I can somewhat understand but on the same token, they're not generating much interest in the software by being so tight lipped. I say throw out a treat every now and again. I made the analogy the other day, give someone an oreo and they just want another one. But starve someone and they'll go somewhere else to look for food. I think they need to find a balance between keeping their plans discreet while still showing people that they are hard at work behind the scenes with some sneak peaks/teasers.

Of course my one big fear in all of this is that they won't change enough and we'll still be struggling with some of the same issues. People cried so much during CORE that it was so 'un-LightWave' and too different but then ask why we can't have things like ICE.:screwy: Obviously where it makes sense, keep it familiar. But change what you need to reach a goal.

erikals
03-10-2014, 08:46 PM
from production i've seen that TAFA did a better job than FaceRobot,
so didn't find FR that big of a deal, unfortunately. maybe FR just needs too many tweaks to get right.

i guess that says a lot about TAFA though... \:]

i asked the TAFA developer earlier on if he was willing to sell it to NT,
but he liked to program it himself, which i completely understand.

maybe NT could cut some kind of deal with him though? before AD comes around... :O

------

i can't say too much about ICE, other than several ICE effects can be done without it, but it requires time.
i've seen Pooby do some damn cool tricks with muscle-sims though...
Lagoa was cool, but i never saw how the lowres results could be of use > http://www.youtube.com/watch?v=OJn3sP5yKK8

------

some people just don't like the UI, personally i like it
some say the GUI is slow... (can't say)



What can LW learn from this... well, I fear, not much. You can have the shiny, fancy, "artist friendly" UI all you want, but unless its supported underneath by an interconnected system... you're gonna hit walls and limitations. And the only way LW is gonna achieve this level of sophistication and flexibility is (I imagine) with a pretty much complete rewrite.

yep.

personally, i'm pretty happy with LightWave these days, i just wish L/M was faster...
maybe Hydra can change some things...


People cried so much during CORE that it was so 'un-LightWave' and too different but then ask why we can't have things like ICE

it was probably more a matter of the Core GUI functions being only 50% done, + the other issues.
but agree, one can't get changes in a snap...

cresshead
03-15-2014, 12:12 PM
http://www.youtube.com/watch?feature=player_embedded&v=DO33ckLkcms

as requested