personally when I read Zaps comments, I thought they were slightly off and perhaps abit short sighted and/or narrow minded. But being as I have only seen him at mostly Max events and his comments and stuff tend to be more leaned towards Max, it could be that's where his focus as been and he perhaps wasn't aware of the bigger picture or some other internal Autodesk workings.

Now that Zap is at Autodesk, I don't know what his remit or role is as I don't work within the development teams. But the work done with Mental Ray for Maya, to make it a normal plugin that can be updated seperately makes sense to me, especially if all our other 3rd Party renderers work the same way. It kinda makes both parties lives easier really.

The MR integration in our products has got better, gradually over time, especially in Max and I've seen Max guys rave about it. But for me, I still hold the Softimage one as the best by far.

I was not pointing to general Zap comments, but instead, to Luc-Erics take on how Mental Images used to deal with things.



Hope the whole ARC move has some impact on this... but dont feel confident about it yet.

For one company to interface a renderer that's been around for so many years and with all the features mental ray has and which is used in uniquely ways and different ways over time by completely different users and markets, is a huge undertaking. Actually I believe its impossible.

How about Autodesk separate mental ray render interface module and related components such as GUI from Softimage and then open source it. A community of developers, including Autodesk, mental images/nVidia, studios and individuals who now put equivalent energy into work a rounds anyway, could work on the actual XSI mental ray interface module and GUI. Autodesk can then bundle it with Softimage. This way it can have its own release cycle, be kept up to date with the latest mental ray features and people are happier with Autodesk and the Softimage mental ray integration. Autodesk can put their energy into improving the overall rendering interface for all renderers more generally. There are developers with requirement's and motivations to tackle individual issues and features that are important to them and each effort combined leads to a more complete implementation.

At the moment Autodesk has to appeal to the lowest common denominator in its rendering interface but a community of developers has no such requirement. Some would choose a branch/switch targeting only the most modern features, others the branch/switch with the most features surfaced. If Autodesk chose to bundle the GUI targeting only the most modern features, the software would be easier to learn and encourage best practise. You could also reduce the GUI complexity per scene or project with switches in other ways such as your approach to rendering ie "global illumination" vs "simulated global illumination using direct lighting". Those who need backwards render compatibility with an old project but who are now using latest version of XSI, could use Softimage 2013 but with the rendering interface from the earlier version and a shiv/shim to decouple it that makes the new 2013 mental ray compatible, ie use the old options, with the old mental ray that the project started in for compatibility. Its only power users who would chose to use something other than the "most modern" or "most features surfaced" options and contribute to their development. The other specialised options would be invisible but still available. This powerful switch customisation system is an additional option that, could, be developed by the whole community.

very well said by this guy at XSIBase.com:

Sorry, lengthy post ahead.


With all due respect, i think there's a little bit of rose tinted glasses effect in that assessment. Max and Maya have to leverage a wider range of fields they get utilized in. Plus, maya is fully multi-platform... They have a tougher crowd to please. Its not my field and i might be wrong, but softimage does not get too often a foot in architecture, manufacturing, automotive etc. When those two get updates, sometimes they look like they are nothing updates because they spread a lot. Granted, most of the new features are stuff they bought and than just blatantly thrown on top of the pile, but soft is not exactly innocent of having alien code either (polygonizer, lagoa, craft, facerobot). I do agree that something really impressive happens only once in a while and the development takes time, but i do get the feeling that softimage gets far less resources for RnD than it should. What it gets, goes into ICE or, more recently, a viewport. Yes, little things are always added here and there, but can you tell me, with a straight face, that those are really something more than tweaks and adjustments of existing technology?

Im more familiar with maya so thats what i use as a straight comparison. Disregarding the quality, AD has developed it more evenly. Every corner was given some attention at any given time. Only modeling has been stagnant and the little work they've done on it has been poor, but thats about it. I will acknowledge that im looking at this from my own subjective perspective, but if maya can go through major changes in animation, rendering, interface, data handling, file referencing, interoperability, dynamics... and at the same time target game developers, animators, vfx artists, mocap guys, TD's, light&rendering guys etc. I would expect softimage to be able to do the same seeing how it is priced the same.

It is weird to complain about this when it clearly took time to develop that viewport, but as a comparison, the year maya announced VP2 they've done a new skinning method, rewrote the interface and even went full editorial among other things.

If Sofitmage guys are to tell me that they do not wish to develop half-baked features for the sake of having something shiny every year, or that this was a major bug fixing year, than i would be fine with that. But there are areas in softimage that are really strong and one should play to those strengths (its a plural), not ignore them.

Instead, i am consistently wondering where is softimage headed? Is it going to be a vfx tool for autodesk? Is it a companion destined to fill in the gaping holes of max and maya? Is it trying to compete with houdini?... Or is it a full fledged 3d modeling, animation and vfx package as i believe it is?

I have lots of praise in store for softimage developers, i do not wish to offend anyone. I am fully aware of softimage's quality, but that does not make you excluded from critique.

If anything, im going to link here a video that Chinny made in 2010. It doesnt show that many features, but they are all very interesting/impressive and he took the time to present them really well. Now i look at the area page and they seem to think softimage is a vfx tool (this is not your corporate site, this is AREA). I also took a look at max page and they embedded every little thing they could to make up for the lack of features. Softimage doesnt even bother... Would it make me feel better? Nope. But at least it would stop me, and other people complain about it.

Btw, this is how they announced VP 2.0 in 2010 and how HQV is presented today:



Just as a counterweight, and to show that i see the other side of the argument, this is what i stumbled upon while searching for that video:

Thats a max guy and ive seen more positive reactions so at least the plan to attract the users from other applications might be working. I just hope you do not take it too far.

Cheers,
Alex

Presentation nowadays plays a huge role in selling a product or we would not have seen flashy advertisements every where. So guys at Autodesk please find/hire a person knowledgeable in Softimage to produce some fine demo videos. Not asking for a tv commercial but at least make some videos that are on par with the other products from your suite. I am a Maya user and a Softimage well wisher, so please don't take it the wrong way.

And people complaining about how slow the viewport is. Wait till you get your copy. Screen Recorder Softwares tend to slow down the viewport while capturing so no matter how many videos we see, we can't guess the performance of the actual viewport. Also this is the first version of this new HQ VP so we should expect more refinements and updates in future releases.

Cheers

@bottleofram - You make some valid points, but still I feel the urge to comment (you know I'm funny that way!).


Splitting hairs here: Saying facerobot represents alien code is a bit strange, it might not have gotten the best possible integration, but alien that code obviously isn't. But then, you forgot to mention XSI Hair, which apparently is mostly alien code.


Personally I was surprised by the incorporation of the new viewport, as it is most definitely isn't ICE or a SDK enhancement. And as it already has been stated that the Softimage resources for RnD are marginal at best, we sadly shouldn't expect too much development in one release anymore. But, that said, even if the new viewport and the CrowdFX aren't what most have been waiting for, they seem to represent a fair bit of "new" development (at least for Softimage).


The battle for full acceptance as a full fledged 3d modeling package has been lost and we probably should learn to live with it. I count myself lucky that they haven't started taking out parts of the older code that isn't geared towards vfx already... (Oops, I hope I didn't give them an idea!)


And you're right: this whole discussion is technically all about shades. Some are wearing the rose-tinted variety, while others wear the black ones, while I myself switch between the rose-tinted ones and the Joo Janta 200 Super-Chromatic Peril Sensitive Sunglasses to get me through release season.

@Bellsey

And sorry Mr.Bellsey about the comment I made on one of the SAT videos in the AREA before you took it off your blog. I couldn't properly express my feelings through my words and I think I might have hurt your feelings sorry again

Ok, I quickly tested SI 2013 for bugs! Sorry, I couldn't resist! But most of them are fixed, except that I know and which is not mentioned anywhere

1) EdgeIsHard attribute works fine! More than that, if you set attribute before subdivide topo node smoothness is removed automatically! Many thanks to Guil apparently!
2) Merging sliced polygon topos is not a problem anymore!
3) And Lagoa Mesh Advection compound is working now!

Thank you very much everyone @ AD team!

any chance the fixed the geo shader? Please check it

btw how did you check it? i checked my sub account and its not available yet