Softimage 2012 announced

[Hirazi Blue]

In a thread on the Mailing List on the subject of pmG's "Dare To Share" Messiah sale Luc-Eric Rousseau made a comment (seemingly unrelated to this current discussion, but bare with me):
Speaking of certain innovations in animation...

Not likely to happen under Autodesk, where that animation design is
made by a cross-product design team, but the continued work on
expanding the SDK and ICE will allow brilliant people like StudioNEST
to build innovative new workflows and tools.

(Quoted from here)
This statement seems to point in a quite definitive direction: Softimage will continue to push the SDK and ICE to enable third-party developers to add further features to the software, implying other areas won't be pushed quite as hard.
I sincerely think the days of Softimage as a "generalist" application are over. If that's necessarily a bad thing remains to seen. Maybe the days of true "generalist" applications are numbered anyhow. But for freelancers (and hobbyists like myself) this obviously poses a problem.

[luceric]

What that comments really means is that in the softimage days, any dev like me could just decide to do whatever he wants and that becomes a feature in the product. In autodesk however, there is an actual design process where people who are specifically designers try to do things that makes sense and work across product. Imagine for example if we added stereo cameras to softimage but they would be done in a completely different way than maya: users would find that pretty stupid and ask why the team don't talk to each other. same thing for animation tools in general; if something is a good idea it should apply to multiple products.

[Pooby]

I think Softimage is becoming ever more a generalists app. It covers all areas with a fast workflow and yet allowing great depth if needed.
I understand the frustration about the apparent lack of advances in mental ray, but then I have never used it to render in.
I think that maybe softimage is constricted by having to support this one 3rd party renderer. It's a shame it doesn't have a native solution.
I would love to see ICE further expanded to the point where users could build an 'open source' full renderer with it.
After all, if I can do this in a few hours, Making it up as I go along, it shows the potential of ICE in this area. http://www.vimeo.com/20652466

[Maximus]

You know, the problem with Softimage rendering is not MR or lack of iRay..the problem is a stable solution to use, with proper features and constant updates.

The lack of tutorials is deeply annoying and makes learning Mental Ray insanely hard, since its one of the doomed engine that requires skill and optimization to be controlled.
I would love to know how many people know how to use all mental ray nodes, because honestly its not really that easy.
Sometimes i try to follow tutorials from other applications like maya and there are some stuff that are really impossible to achieve without someone telling you, wich brings us back to square 1, tutorials are really needed, proper documentation is really needed.

one example of deep frustration is just behind the corner here
http://www.xsibase.com/forum/index.php? ... 0#lastPost
viewtopic.php?f=10&p=9207#p9207
i might be stupid but i cant find same parameters and stuff, i had to drop it by desperation..

Those to me are the main problems regarding this issue:

1) lack of coherence, feature wise regarding the same engine under the same company on 3 different applications, every 3d apply has its own version of Mental Ray implemented in a total different way, with exposed or unexposed stuff, with features or not features..
this is frustrating to say the least. You cant even try to learn from the other 3d package materials and tutorials because during your tutorial you hit a roadblock since you dont have that parameter you need to use.

2) lack of tutorials, and with this i dont mean Digital Tutors level tutorials, that is the same as going through F1 help, except you waste money. Tutorials are a boost to learning, there is a limit you can reach easily alone, but after that you are totally screwed, regarding Mental Ray especially.

3) lack of updates, saying this since years now, we dont have an engine supported, updated to the current status, with the latest features wich are made to improve your work of course, at mental images they work on their engine actually, but whats the point if then end users dont get the updates?

4) lack of alternatives due to small userbase wich leads to small request wich lead to few software houses interested into tossing and (wasting) money into making 3d party renderers available for softimage.

In the end, whats gonna sink Softimage is gonna be render engine if something doesnt come up quickly, and with quickly i mean quickly.
Most people i hear are interested in Softimage, in ICE, in Lagoa, in soft workflow, performance and many nice things, but at the question "yeah but how about render? it has Mr only? what version? how is it?" everyting goes to hell.

I pray everyday VRay comes out for softimage but thats not gonna happen, its in development since what 5 years?
After all i seriously dont understand how MR got so bad updated..but again..things arent goin to change so well...cant do much more than just writing on a forum.

[Hirazi Blue]

Guess I will have to elaborate on my earlier post:
@luceric - you're obviously right, but the fact seems to remain that the pushing of ICE & the SDK as the two things where there isn't the need to decide new features together with the sibling Autodesk products means that the future for interesting new stuff lies with ICE & the SDK, not so much with the "other stuff" that needs to be decided in conjunction with the other apps. This doesn't necessarily spell the kind of trouble I foresaw in my earlier post for all the other features, but it doesn't look very promising either.

@pooby - my remark about "generalist" apps was only partly right. While it still will be possible to use Softimage (hopefully for quite some time) as a jack-of-all-trades "generalist" application, the impression I got over the last couple of years is, that there is an ever-growing tendency in software and especially Softimage to enable doing certain stuff in other applications, while developers are not developing them further themselves. (It looks like that's what ultimately will happen on the Mental Ray front: no need to develop it, if "everybody" uses Arnold anyway.) If a feature is missing/underdeveloped in your own application, you are reminded you can easily go to some other application (preferably of the same company) to get things done. Unless you are on a tight budget, obviously.
An interesting illustration of this way of thinking seems to be the way nobody has addressed the many problems XSI Hair has to deal with, while now there's an (understandable) tendency to point to the very good third-party alternatives, freeing the developers from the burden of fixing XSI Hair themselves or providing their own new solution (or buying one of the third-party alternatives, obviously.)

Don't get me wrong: I love ICE & will even give C++ a renewed shot to get to the all the new goodies in the SDK, but I really think that exciting new features shouldn't be almost limited to ICE & the SDK. And wouldn't it be nice to have some new animation feature Max & Maya don't have? Or a renderer that works?

[Pooby]

And wouldn't it be nice to have some new animation feature Max & Maya don't have? Or a renderer that works?

I hated the pedantic nature of Mental Ray and I felt it went against XSI's user friendly philosophy, but then, I came from Fprime, in LW, which had to be the easiest renderer out at the time. If on top of that Mental Ray doesnt work properly, then I can only imagine the frustration. Luckily for me, I never have to use it and hopefully, XSI rendering issues will sort themselves out in the next year or two with new alternatives and MR fixes.

As for the animation features. XSI is leaps and bounds ahead in all the animation-related areas I need it to be, (primarily, being able to use ICE to create creature deformations) so personally, I am satisfied at the moment.
My greatest hunger in XSI now is just more and more ICE abilities and believe me, I never thought I'd say that. I was firmly in the opposite camp this time last year. ICE enables you to make you be able to do whatever you want in the areas it covers, so any non ICE 'feature' seems like more of a backward step to me.
My take is, if you use XSI and you dont use ICE, then you're missing out big time and that discrepancy is only going to get worse. THats why I made the tutorials.

[DoF]

Vector Displacement Maps

* VDMs can represent directional displacements that do not simply follow the polygon normal
* Support for render VDMs in mental ray
* New Vector Displacement Map shader node


Crosswalk

FBX

* Stereo camera I/O
* Importing of Vector Displacement Maps from Mudbox
* Port to FBXSDK 2012
* Add "Import Fbx..." under "File->import" menu and "Export Fbx..." under "File->export" menu
* Scripting commands to create the import/export options
* Import .3ds, *.dxf, *obj, *.dae formats
* Ability to select target .fbx version (2010 (Motion Builder compatible), 2011, 2012)
* multi UV in the FBX Importer/Exporter.
* Support importing CgFX materials
* Support for neutral poses
* Support for user normals
* FBX One-Click Mudbox
o Materials : phong, lambert and blinn are supported.
o Textures : normal, displacement, vector displacement map, ambient, specular, shiny, incandescence, bump, reflectivity and diffuse are supported


Suites and Interop

* A new Suites mode for the FCurve Editor
* Single Step interop for Mudbox
* Single Step interop for 3ds Max
* Single Step interop for Maya - Updated
o You can now start in Softimage



-that is so sweet!
WOW!

[Bellsey]

ok, I'll chip in abit here......

New Features
First a question, what is a new feature? I ask this because when you look at all three apps, they're already pretty full of every feature you can think of, what else is there?..really. Sure there's new tech and tools being done all the time, but they all generally fall under something that has already been done. What's new about them is the way user can use that feature. In the last few years, the term 'new feature' has started to mean something else to what it was originally labelled for.

Another question.....do you really know what ICE actually is? An odd question I know but if you do know what ICE is then you'll know that adding more ICE, and SDK work, to Softimage is by no means a bad thing, it's a great thing. If we are adding 'new features' or enhancing/replacing older ones, then it makes perfect sense to try and leverage as much of ICE as is possible. Unless there's a technical/logistical reason, why shouldn't we do that? It's also not limiting or restrictive, quite the opposite in fact.
Luc-Eric is right, you only have to see from tools like Momentum, whats if you want to add a new tool/feature into Softimage and you use ICE to do it.


Rendering
Mental Ray is by no means a bad renderer, it's very very good. People maybe be spitting their coffee out after reading that, but honestly it's true. However like any renderer, the more work you put into it, the better results you will get out of it.
Also many users point to us regarding Mental rays problems, but remember Autodesk don't make Mental ray, it's Mental Images. There's issues we can deal with and issues that we can't.

The other issue with rendering in general is that it can be made to look very easy, when in fact it's pretty hard. You do have to know what you're doing, and renderers like Vray are good are making things easy for the user so they can get a good result very quickly. Mental Ray can produce a just as good result, but need more love and attention, which to some industries is fine, to others perhaps not.

On the subject of iRay, I'd ask people do they really need it right now for something, or do they want it because Max has it and Soft/Maya doesn't. For me iRay is smething that suits some users more than others, depending on the work they do.
I see and talk to alot of Soft and Maya customers and though they see and wouldn't mind iRay, its not a deal breaker for them and they don't have any immediate use for it. For them, they have other priorities When it comes to Max and the amount of usage that package has in the Arch/Design-Viz sector iRay makes perfect sense.

[SreckoM]

What about progressive rendering, is this tied to iRay or this can be implemented as separate feature?
@Bellsey
However, there are several important things that needs to be fixed on Softimage side before sending ball in Mental Images courtyard. Most of them are listed in several Maximus post on this forum. Not that he or anyone else didn't give a love to mental ray, but this is just masochistic sometime

[Bellsey]

However, there are several important things that needs to be fixed on Softimage side before sending ball in Mental Images courtyard. Most of them are listed in several Maximus post on this forum. Not that he or anyone else didn't give a love to mental ray, but this is just masochistic sometime

Precisely, which is why I said that there are things we can sort out, but also things we can't. I wasn't trying to pass the buck.
Also, there's been alot of work go on to ensure that the MR versions are alot more consistent across our main apps and also MR Standalone, which again falls into that collabrative approach that Luc-eric has mentioned. We're also very much aware of how important stability is to people, so we need to ensure that before a new version of MR is implemented, it doesn't result in being more trouble than its worth. That doesn't help anyone.

[Kzin]

Rendering
Mental Ray is by no means a bad renderer, it's very very good. People maybe be spitting their coffee out after reading that, but honestly it's true. However like any renderer, the more work you put into it, the better results you will get out of it.
Also many users point to us regarding Mental rays problems, but remember Autodesk don't make Mental ray, it's Mental Images. There's issues we can deal with and issues that we can't.

The other issue with rendering in general is that it can be made to look very easy, when in fact it's pretty hard. You do have to know what you're doing, and renderers like Vray are good are making things easy for the user so they can get a good result very quickly. Mental Ray can produce a just as good result, but need more love and attention, which to some industries is fine, to others perhaps not.

On the subject of iRay, I'd ask people do they really need it right now for something, or do they want it because Max has it and Soft/Maya doesn't. For me iRay is smething that suits some users more than others, depending on the work they do.
I see and talk to alot of Soft and Maya customers and though they see and wouldn't mind iRay, its not a deal breaker for them and they don't have any immediate use for it. For them, they have other priorities When it comes to Max and the amount of usage that package has in the Arch/Design-Viz sector iRay makes perfect sense.

i think in terms of iray ad is going the right way. the same people who wants iray in soft are the first which would crying about missing feature of the renderer.

and you have 2 other rendering options in xsi. one, of course unofficial, is arnold, and the other one is 3delight (octane and other obj export based renderer are another one). i dont know why people dont use 3delight. its free! for 2 cores, renders fast, stable like hell compared to mr, but it looks like people are not willing to use it. and i think i know that the problem ist, also for mr. this might be offensive, but more and more poeple are not willing to learn anymore. they start a thread and crying about how bad mr is, missing feature (which are in mr) and the overall bad performance. some points are really a problem, stability and the out of the box state to name two of them (besides the biggest, what maximus says, the integration). but more and more threads are showing that the users dont read the docs, search the forums or simply insert their problems into google search. for example, cgtalk is such a huge knowledge base, keep your eyes not only in the soft section, maya ones are far bigger and more frequent, the same for max. you will find many many informations, tricks and optimizations, example szenes and alot of help from other users. if you have a question then open a new thread and ask, you will get an awnser here. and after some time you will see that mr can do alot more then you think. ;)

and i know its a problem to read faq's are threads to solve a rendering problem, but you will always find a method to render your shot (besides buggy or unsupported shader).

[ActionArt]

Mental Ray is by no means a bad renderer, it's very very good.

I'm one of the few that would agree with this. I've posted a few times before that I like MR and if you take a bit of time to really learn it, it's very capable and flexible. Others might be easier but not necessarily better. Arnold may be an exception but I have no experience with that.

On the subject of iRay, I'd ask people do they really need it right now for something, or do they want it because Max has it and Soft/Maya doesn't.

My wish for Iray has nothing to do with Max. I do a lot of hard surface mechanical modeling which I guess is unusual for a SI user. I suppose I should be a Max user but I just like SI better I think I would find Iray very useful and this seems to be a major direction MR is going. I was hoping SI would stay right on top of it. Iray and the render region tool are made for each other!!

Just my 2 cents.

[Hirazi Blue]

do you really know what ICE actually is? An odd question I know but if you do know what ICE is then you'll know that adding more ICE, and SDK work, to Softimage is by no means a bad thing, it's a great thing. If we are adding 'new features' or enhancing/replacing older ones, then it makes perfect sense to try and leverage as much of ICE as is possible.

Strangely enough, no contest from me , there is however such a thing as usable tools and less usable ones. I really wouldn't mind if Softimage developed an alternative to the XSI Hair for instance, with lots of ICE-goodies under the hood, that can be accessed when need be, but also with the clean tools and the ease-of-use of the old, otherwise hardly usable, XSI Hair. Sometimes you just want an easy toolset, not all the control from some clunky ICE Tree.
That doesn't mean I don't know what ICE is, BTW. The whole idea of a "core-rewrite" as it was mentioned around XSI 7, however, doesn't mean that ICE has to show in everything you do within Softimage. ICE should be accessible, not necessarily "in your face" all of the time.
But hey, that's just my opinion...

[jamination]

People talk a lot about 3delight as a viable rendering alternative, I own (and am glad I do), but I find it craps out early in a project and I have to go back mental ray (especially where ICE is involved).

[Maximus]

We kinda need easier render engines, not even more hard and poorly documented..
Find me another render engine that is so helpfull and clear on documentation: http://www.spot3d.com/vray/help/200R1/
(and you can actually learn vray once, and use it in every 3D package in the world, since the parameters are all supported and all the same unlike Mental Ray AD integration)

3Delight is horrible regarding tutorials, its a tech engine, and i guess we got enough of those..
You see yourself coding and writing renderman shaders? Well i prefer to stick with MR so, 3Delight is not really a viable solultion, add to that the fact 3Delight is dead slow Raytraced compared to MR, maybe its good for animation/dof/displacement but meh i dont consider 3delight a raytracer.

All the times i tried to learn and move to 3delight result in heaps of crashes, disconnected shaders, not supported shaders, freezes, and a total desert in terms of documentation and help.

[jamination]

I agree Maximus, I would be happy if they just had some examples. Mootzoid and Helge Mathee are wonderful when it comes to examples.