Spoiler Alert: NO SPOILERS ANYMORE

[Bellsey]

ah yes...Face Robot the best kept secret of Softimage and now Autodesk, imo.
Everyone that I've ever shown Face Robot too has loved it, but it was always tricky to get people to take it on. Imo, there still isn't anything like it around. Other products are around and look good, but they appear to be focused more on capturing and retargeting, but FR has some good stuff and you can actually do alot with it.

One of the many myths that always became a block to people using Face Robot was price. Though to begin with, it did have this 100k price tag, at the end when Softimage was acquired it was nothing like that. But the lower price point was never released or made public, so it always made conversations tricky.

Personally I've always wanted to see more work on Face Robot as there's still a great deal of potential. It's not too late though, so who knows.

[Bullit]

wasn't that what happened to the old (forgot the name) XSi Crowd system as well?

Yeah, Behavior.
I think there is only one path they need to do, turn ICE more friendly, less context issues with each common node to work in all situations. Then see the Icetree grow by users. If they remain with ICE only for TDs then there will be a point of not getting TD's even...They need to beat maxscript.

If they don't do that they will be bypassed by others even where they still have advantage.
I also hope the new guys lost the Hollywood fetish of Softimage old timers and stop chasing windmills in Sunset Boulevard...most of great work in Softimage are in spots not film. Besides film as we are seeing are a lot of work and pays badly.

[mirkoj]

On FaceRobot, it is amazing part of SI and yes go GREAT potential, but it needs a bit more integration before it picks up.
I mean to start with two diferent layouts, seems like separate part of SI. I understand why does it need but there is probably way around that.
Then as mentioned before, need to cut out head, work on separate scene... for keyframe animation that is big problem as for facial animation you kinda need your character in the scene and for full body interaction to work with face, it is not just separate world.
Those things are I think main reason why it never picked up that much.

But instead of wasting time on on some other things, like working better with Maya for one, more time should be spent on FaceRobot. Softimage is the best character animation tool but won't stay that for long when NOTHING is done to improve it more and push it forward.
Let's face it, just for example shape animation is FAAAR superior then Maya blend shapes. I've spent 4 days fighting with simple fish rig in Maya with simple bone structure couple dynamic bones for fins and couple blend shapes for face and it was nightmare! Adding new blend shape on finished rig is hell...
Just for fun of it did same thing in SI in 30 minutes. Fully dynamic and playback performance fully real time with all dynamics. Now I'm not saying, someone else would do it faster in Maya sure but workflow in SI is soo intuitive and great that there is really no need to fight with that in Maya.

So if suite is pushed so hard why not present SI as all around software especially better than other two in character rigging and animation, and then show how everything is done in SI in a day and full character ready rigged and animated and then button Send to render in Maya or Max. Really?

Max excelling in character animation based on that graph we saw??? Every single character animator that I know, myself included, HATE Max for that. So much that as soon as there is freelance project that asks for character animation in Max I turn it down right away. There is no money that will push me into that nightmare. It sucks. How can Max then excel in character animation and SI only in particles????? Which idiot made that graph at first place???

Sorry for some morning runt but it is not right and all SI people feels it.
It is a bit hard to understand how people that worked on Softimage before AD can live with that.. looking something great they made being presented almost as junk... I know that I would shed a tear every single night on that place...
Is it possible that no one there actually see how Softimage is great and have that much emotional detachment? Fact is people LOVE Softimage and it hurts seeing what is happening. Simple as that.

[benoit81]

Fact is people LOVE Softimage and it hurts seeing what is happening. Simple as that.

I agree.
The fact is that you spend so many hours with Softimage everyday, it become more than a tool!
I realize that everybody has it's own wishlist. But what seems to happen with Autodesk is that users get used to small improvements. I -for instance- would like that Autodesk implements standards such as PTex, OpenSubdiv, DX11 adaptive tesselation (for realtime displacement)... But I loose since Autodesk area any occasion of innovative/surprising tools such as ICE, M component tool, GATOR... All those tools that make us pride and creative.

[Kzin]

facerobot could not be a success because of two reasons.
si were not used in high profile studios for animation so i dont know which client should get fr.
second was the price itself, it was way to high. compared to other studios which could developed their own facial software it was to much for such a closed system.
and smaller studios could not afford the price, it was just a bad joke. best would be to bundle face robot in si from beginning and made some more marketing for it.
now its bundled, but i think it will make no difference in si sells.

and the crowed stuff was a half finished piece of software, a nightmare to work with, not usable without coding, limited in feature set and a pain in terms of data exchange.


that brings us to ICE.
my feelings about it, it will end like fr or si crowed.
its to technical, a plugin inside of si and not si built around ice, like it should be. until today most stuff is still not ported to ice. i thought at the beginning of ice that si will be built around, deformers based on ice, particles, cloth, dynamics, hair, all is based on ice under the hood without the need of open the ice tree for every single little piece of stuff. but nothing happens in that direction. they gave us strands and thats it. like throw a bone to a dog and say, here you have, play with it. dont understand that and it shows one more time that alot of potential is thrown away.
we dont have the plugin mentality of max, so we need a decent set of tools out of the box in si which ice cannot deliver in current stage.

and yes, i also dont understand what the whole crowd stuff should be.
an artist oriented ice based hairtool with a good featureset would be a way better addition then crowed.

[xsisupport]

that brings us to ICE.
my feelings about it, it will end like fr or si crowed.

So, if I understand correctly, you are predicting that nobody will use ICE.

[mirkoj]

ICE will not b e used nowhere near as it could be but AD never showed a single trace of interest to make it.

Having as much things in SI based on top of ICE and ICE under the hud seemed like logical way.
Example, replace all possible mesh deformers with ICE versions.
Users use them as so far with menu and if when needed they can take a peak under the hud and modify.
Wasn't that idea at any time to anyone else in AD??

Keep easy of use for artist guys and also give option to sneak peak under and modify and upgrade to more tech guys or anyone else willing to jump into. But lokos like ICE has taken max mentality and starting to be huge collection of addons that you need to install over and over again.

I HATED max for all plugins and stuff as it was alway s the feeling that you need to get plugin for anything to do in it while SI dealt with everything out of the box.
One thing is when addon or plugin strengthen your tool, completely different is if addons are MUST to be able to work at all.
99% of talk between max guys was: you got this or that plugin?

[Yanik]

facerobot could not be a success because of two reasons.
si were not used in high profile studios for animation so i dont know which client should get fr.
second was the price itself, it was way to high. compared to other studios which could developed their own facial software it was to much for such a closed system.
and smaller studios could not afford the price, it was just a bad joke. best would be to bundle face robot in si from beginning and made some more marketing for it.
now its bundled, but i think it will make no difference in si sells.

Another reason FaceRobot will never be successful is because of it's usability. It's interface is setup with synoptic views made from html image maps, css and javascript files which are extremely limited and have a huge impact on usability. I believe synoptic views don't have low level access to the objects you're manipulating so it has to go through slow XSI commands.
I hope FR doesn't get touched or updated because it will be a waste of time/resources.

I really hope Softimage gets native Qt integration in the future(not via Python) which would allow to create more usable user friendly tools. ICE is also limited by this. You can't create a custom interface for your nodes and are forced to use a limited amount of widgets. ICE could be a lot more powerful and user friendly when you could create custom Qt dialogs for nodes where you can have tabs, icons/images, html, opengl viewport, buttons, graphs or whatever your heart desires.
Imagine a hair system node dialog where manipulating a viewport in the dialog will give you direct feedback and manipulate the output of that node(see my crappy mockup below). Presentation of the nodes should be separate from the node's output. I also don't think a PPG implementation with multi-ppg editing is necessary for ICE nodes/compounds since you (almost) never manipulate more than one dialog's at a time.



One can only dream.

[Kzin]

that brings us to ICE.
my feelings about it, it will end like fr or si crowed.

So, if I understand correctly, you are predicting that nobody will use ICE.

no, alot of si users using it, but thats it. no one is getting si because of ice.
ice is a great tech, but ad missed the chance to make it a kick ass feature. i know alot of people which not using si and then i mention ice, they say, why should they learn it then they can do the same in max script or mel or max has a native plugin for doing it.

if ice would be more artist oriented instead of showing connecting vector nodes when you would have a selling point. there are so much usability thing in ice that makes the whole system to complex, especially the whole micro management of ice nodes to get you options together. sure, you can build a ui, but sorry, that took along time. its nothing to use for every little thing, also when it would be usefull. how many ppg's i had opened when working with a scatter ice tree because all the values are scattered in many different ice nodes. sounds stupid, that are points that drives me crazy when working with ice. its not a tool for everydays work, especially not for non ice artist. so we need presets, not only simple ones, presets which presents tools, not only ice compounds.

[Fer]

that brings us to ICE.
my feelings about it, it will end like fr or si crowed.

So, if I understand correctly, you are predicting that nobody will use ICE.

no, alot of si users using it, but thats it. no one is getting si because of ice.
ice is a great tech, but ad missed the chance to make it a kick ass feature. i know alot of people which not using si and then i mention ice, they say, why should they learn it then they can do the same in max script or mel or max has a native plugin for doing it.

if ice would be more artist oriented instead of showing connecting vector nodes when you would have a selling point. there are so much usability thing in ice that makes the whole system to complex, especially the whole micro management of ice nodes to get you options together. sure, you can build a ui, but sorry, that took along time. its nothing to use for every little thing, also when it would be usefull. how many ppg's i had opened when working with a scatter ice tree because all the values are scattered in many different ice nodes. sounds stupid, that are points that drives me crazy when working with ice. its not a tool for everydays work, especially not for non ice artist. so we need presets, not only simple ones, presets which presents tools, not only ice compounds.

Those same people who gonna jump in joy when a third party comparable ICE system will be available to them. I have the same people in the Maya side, doesn't stop them from up voting "ICE for Maya" in the feedback site.

[Kzin]

Those same people who gonna jump in joy when a third party comparable ICE system will be available to them. I have the same people in the Maya side, doesn't stop them from up voting "ICE for Maya" in the feedback site.

wait, wait, wait. ;)

[Pooby]

I agree that ICE is underused in the industry in general. I am pushing to get even our Maya team trained up in it.
It reminds me of the feeling I had about using Fprime in 2003, or whenever it was released. At the time, it was the only 'live' progressive raytracer, and still the most interactive I have ever used. but now, that kind of live rendering preview is the norm.
For me, I know that I can workflow with ICE to do stuff that Maya artists would have to stop and get a coder on for a week before continuing. This gives ICE users a gigantic advantage, that one day will be considered to be the norm.
Personally, until there is an ICE alternative, (Fabric?) I'll keep using Softimage to at least prototype stuff. Even if Softimage stopped right now. I'd keep using it for years as there is no way I could operate without it. It has a very comprehensive fast vector manipulation system in ICE, which is at the heart of making the deformers you need for creature work.
When something like ICE comes out in Maya. Watch as 'the industry', no doubt, will hail it as something revolutionary. We have that right now..

[adrencg]

that brings us to ICE.
my feelings about it, it will end like fr or si crowed.

So, if I understand correctly, you are predicting that nobody will use ICE.

no, alot of si users using it, but thats it. no one is getting si because of ice.
ice is a great tech, but ad missed the chance to make it a kick ass feature. i know alot of people which not using si and then i mention ice, they say, why should they learn it then they can do the same in max script or mel or max has a native plugin for doing it.

if ice would be more artist oriented instead of showing connecting vector nodes when you would have a selling point. there are so much usability thing in ice that makes the whole system to complex, especially the whole micro management of ice nodes to get you options together. sure, you can build a ui, but sorry, that took along time. its nothing to use for every little thing, also when it would be usefull. how many ppg's i had opened when working with a scatter ice tree because all the values are scattered in many different ice nodes. sounds stupid, that are points that drives me crazy when working with ice. its not a tool for everydays work, especially not for non ice artist. so we need presets, not only simple ones, presets which presents tools, not only ice compounds.

ICE may not be the easiest thing to wrap your head around if you're a non TD user, but it's a thousand times easier than doing actual programming. I can't write any code. When I'm connecting nodes to make simple custom effects for my own projects, I'm writing hundreds of lines of codes in seconds. I think that's pretty amazing for an artist.

I'm never going to be able to grasp it like Eric Thivierge or Eric Mootz, but ICE can be useful even for the artist brain, if you spend a little time to understand a few basic things. I'm glad it's there.

But, it wouldn't bother me if some genius used ICE to create a point and click hair system.

[angus_davidson]

I think where ADSK has missed the boat on ICE is not on the tech side which is solid, but rather on the learning side of it. Its not a system (for the average user)that you can just look at and figure out from first principles. There should be much more tutorials , videos, howtos etc. Since ICE has been out we have had 30 Postgrad students go through our course. Only one of them has had the determination to push ahead and do something with it. Everyone else has had a look and gone , oh hell no, just don't have the time. I am not disputing what a great system ICE is. Its just the bar to entry is pretty high for the average artist and would be greatly helped by better and more interactive learning materials.

[bottleofram]

When something like ICE comes out in Maya. Watch as 'the industry', no doubt, will hail it as something revolutionary.

http://download.autodesk.com/global/doc ... r2014.html

Well, knowing autodesk, i whink its fair to assume this is how it will go: there you have an editor - all nuts and bolts will be exposed so that an enthusiastic third party developers can sit for several years in their dark rooms cooking an elaborate system to replicate most of the ICE functionality directly in maya. Than, Autodesk will buy them out and *poof* a whole new feature! This will generate quite few braingasms but soon enough, they'll start complaining about how restrictive the thing is, how badly interconnected and so on.

Meanwhile, ICE itself will stay pretty much the same because-- Well, maybe not exactly the same... But not much different either.

If you're hoping something ELSE will come along i think you'll be waiting a loooong long time for that...

It's just what happens when someone owns the industry.

[Pooby]

It's just what happens when someone owns the industry.

Todays perspective can be a compelling convincer, but Autodesk don't OWN the industry.. All it takes is for something to come along and truly show that a system can be employed that enables work to be done 10 times faster and for those studios using that to start making headway, and things will change.