PDA

View Full Version : Spontaneous Creates U2s Latest Music Video with SOFTIMAGE|FACE ROBOT


BeWare
01-23-2006, 02:10 PM
http://www.softimage.com/home/Press/PressReleases/060123_face_robot_u2.htm:thumbsup:

mental
01-23-2006, 02:52 PM
Nobody rocks the 'Extrusion' effect like NYC design houses :rolleyes:

gunslingerblack
01-23-2006, 03:09 PM
so, i saw like , 1 or 2 times he was singing...and he was in total wireframe mode, the models were nice, the female that accompanied him was freaky lookin,

overall im not really impressed still, i would like to see a better demonstration of what this technology can do.. yes i've seen the tech demos, and they look wierd too.

mr-doOo
01-23-2006, 03:26 PM
well, i found it boring as possible....

zzacmann
01-23-2006, 05:55 PM
Okay, I give up. Is there any place where I can actually WATCH this video instead of just read about it?

gunslingerblack
01-23-2006, 06:12 PM
well i had to do a search on google to find it

camposino
01-23-2006, 07:07 PM
WOW! Amazing! That is the future of CG! What, you guys didn't get it? That was all CG U2 members!

It better be...after watching that piece of crap.

oktawu
01-23-2006, 08:46 PM
well, i dunno about this one
seen it a while ago because of all the hype
artistically, i dont want to comment
but technically, i really didnt find it all that great...at all
as for the band members, if they are all cg...i give up...but i really dont think they are cg (waiting for someone to correct me)

if they are not, then the only shot of bono and face robot is that 1 second wireframe shot
which even if it looks cool it shows me once more, how money is thrown away these days in the industry (with all the mocap session and stuff....) -but if the band members are indeed full cg...i apologize

third and final...facerobot seems like a cool technology, but for us artists, currently really feels useless
all the techdemos ive seen involved mocap..so the system is built with mocap in mind...so how many of u can afford a vicon system at home raise ur hands..... (for that matter how many will afford facerobot anyway)
another thing is that even though the system seems to mimic impressively the actual facial animation, i feel it still has miles to go in terms of lipsync...i mean all of blurs movies that ive seen (even the xmen legends cinematics and the cod promos etc) even thought they had nice facial mimic, their lipsync still felt way off, even with mocap and facerobot...
of course im not taking timeframe into consideration.

just my mere thoughts

gunslingerblack
01-23-2006, 09:40 PM
the only shots of the band that were cg were the shots of them in the wierd ethereal transparent state, and the only shot i saw of face robot was the wireframe shot mentioned.

if face robot was used on the female head it made her look plain wierd...

oktawu
01-23-2006, 10:39 PM
u got a thing for the word "weird" man :)

gunslingerblack
01-23-2006, 11:15 PM
that was a constructive addition to the discussion lol


how about unsettling, zombie like, not right, uhmmmm unnatural

those are better than wierd right

even thier tech demo looked wrong, like it was supposed to be this great thing and i watched the tech demo going....whats so great about that?

polar express written all over it,

it looks as if it could have been a byproduct of the fabled "performance capture"

at this point anyways

shingo
01-24-2006, 01:03 AM
How about the fact that it was Polar Express with 10 percent of the marker data?

I agree given the little air time forht efacial anim stuff, it was dissapointing.

What looked so "wrong" abto the tech demo? That "wrong" had many of the worlds top 3D studios calling Soft to register interest.



even thier tech demo looked wrong, like it was supposed to be this great thing and i watched the tech demo going....whats so great about that?

polar express written all over it,

gunslingerblack
01-24-2006, 01:15 AM
it looked very clumsy and unnatural, for such a great system the animation and motion was very unimpressive, im not disregarding the technology but, i am saying that i haven't seen a decent implementation of it as of yet.

if i see it work the way they hype it to , then it would be a tremendous bound for facial animators to be freed from blendshapes and the constraints they imply, but right now as someone pointed out earlier, it seems like this system has been used with mocap only, so whats the point for us.

i wanna see the hands on approach and i wanna see them create a great result, not some unnatural jumpy animation that doesn't look convincing to me at all, have someone breathe real life into a character with the system and i'll acknowledge it for what people claim it to be.

spirozero
01-24-2006, 02:04 AM
Where can one actually view the video? I still haven't found what I'm looking for.

CB_3D
01-24-2006, 02:07 AM
I only found the AOL stream. Its so small that you can hardly see anything, so i wont post it here.

mental
01-24-2006, 02:16 AM
>> Right on Spontaneous' Website << (http://www.spon.com/site/reel.php?project=48)

monovich
01-24-2006, 02:20 AM
I saw it on VH1 top 20 video count down. I'd check their site or VH1.

I was puzzled by it. there were small interesting bits in it, but I didn't even get the narrative. What was the point of all of that? It was like those old Mind's Eye videos.

Not trying to jump on the critical bandwagon...

phyle
01-24-2006, 03:12 AM
I would like to jump on the critical bandwagon.

I think face robot has heaps of potential though.

opus13
01-24-2006, 03:41 AM
hmm... in a cg laden media world, any moderately sized company can pick and choose their press release material. im surprised that softimage considered this worthy of a release because of the U2 name.

all in all the video is pretty uninspired. grab any frame of it and it looks like a garden variety renderosity image.

SheepFactory
01-24-2006, 03:46 AM
Lol , that looked more like a Face Robot technical reel than a music video.

zzacmann
01-24-2006, 04:03 AM
Lol , that looked more like a Face Robot technical reel than a music video.

If I were the team behind Face Robot, I'd be embarassed to mention this on my reel.

It looks like someone couldnt figure out how to use Face Robot and decided to use Poser instead.

CB_3D
01-24-2006, 04:35 AM
I suppose only the bono wireframe singing behind the girls head is facerobot, and that little piece looked VERY good and natural to me. Much better than anything ive seen so far in FF or that strange trainmovie with Tom Hanks.

gunslingerblack
01-24-2006, 05:24 AM
haha i still love that people bash the final fantasy movie,

i agree that there wasn't alot of life in those characters but nowhere near polar express level of zombies.

i personally liked final fantasy, i accepted it for what it was an action movie that attempted photoreal humans


as for the 1 second of wireframe we see of bono's face, it's 1 second...and it's a wireframe, and it's behind the wierd ass girl face that was prolly animated using face robot too, and she looked damn freaky.

i do acknowledge the potential of the system, i just wanna see it being used by hand and have a character actually look like its alive, instead of motion captured.

EpShot
01-24-2006, 06:08 AM
the cinematics for X-men Legends 2 uses a beta version of Face Robot.

gunslingerblack
01-24-2006, 06:28 AM
thank you epshot for bringing that to my attention, i do not own the game but i just took a look at the hi def game intro

the intro had some pretty cool parts, but i want to comment on the face robot implementation , the faces seemed very stiff to me, as if the parts of the face were almost disconnected, and poses that should have been loosened were held, (which could be a result of the face robot system or an animation choice) a key example of this is when wolverine and sabertooth get in each other's face after mystique points the way down the hallway, wolverine grimaces and peels his upper lip back, but the rest of his face remains pretty static his brow doesn't furrow any more and his sneer doesn't increase at all (i was looking for a gentle pulling on the skin of the nose particularly on the bridge and around the nostrils) its seems like his lips moved independent of the face when the whole face should have struck a more extreme pose with the lips coming up.

however i also saw the potential of this system as well, when sabretooths expression went from tense to relaxed as he reacted to the xmen showing up it was quite a nice transition. still tho perhaps you can shed alittle light on how the face robot system works, all ive seen of it is the locators and it's implementation with motion capture, im assuming that it works in a similar way with no motion capture, where you would assign an influence value to those locators, and may even be able to place them on a per mesh basis for individualization in characters? i know it's still in it's beta phase so it's probable that you wont be able to say much, but the system is interesting to say the least, i just wish i knew exactly what it did.

ps, how does modelling in xsi compare to max or maya in your opinion?

EpShot
01-24-2006, 07:57 AM
you aren't mentioning technical limitations, merely artistic (or in our case, limited time) decisions. you can clearly see what you ask for in othe rcharacters(tho i don't remember if nessisarily in the opening sequence.

I t bascialy is motion capture. Tho because it uses very few markers (i don't know the numbers, i think its like.. 15, 20. *shurg*) It is 'relatively' easy for an animator to go in and adjust it. given a time frame. I have no idea if this was done during the X-Men cinematics. tho you should realise that i was being programmed at the time. I'm don't even think its finished yet. I have however seen the setup, and it is really simple. the big selling point is the ease of use, flexibility and speed. basicaly theres nothing like it out there, which is why we talked with XSI in the first place (i should ad a disclaimer that i am neither a spokesman nor even very knowledgable in this, i've just had a quick look and have heard what basicaly has been going on) this system doesn't rely on weighted points or driving morphs like old systems, it actauly moves the mesh around, basicaly a set of muscles, fat and bone, not so much in a technical sence, but thats how it basicaly works out, i think ;)

face robot is the only thing we use XSI for. (unfortunately :P )
i actualy have grown to enjoy modeling in max more than i did maya, tho maya could push way more polygons. tho since we got new computers, its doesn't really matter =D


hope that shed some light.


the brothers in arms commercial we did used it breifly. it wasn't perfect, but it was a lot nicer than if we ahd to use morph targets given the time frame. (tho we still don't use it for everything.. unfortunately, (working on morphs right now.. but at least you'll never see them :D)

gunslingerblack
01-24-2006, 02:16 PM
thank for takin the time to give us that answer my friend,

it's unfortunate that limitations are imposed but thats the nature of the biz i guess.

i myself am a student of the osipa system of facial setup, which right now provides me with a very rich set of facial morphs to choose from. i've only been doin facial setup for a short time but, i credit the osipa system with landing me a job, it definitely made me look good.

thank you for the explanation on face robot, it does seem to have quite a bit of potential.

ps

quaaaaaaaaaaaaaaaaaaaaake!

oktawu
01-24-2006, 02:25 PM
i thank u also for the insight
that was really cool of u

though there is one thing i've always wanted to ask about the process ever since i heard about it, not having worked in a production environment.
How exactly (not necesarily in blurs case) do u split ur facial pipeline from ur animation pipeline
i mean, isnt it the same object at rendertime
i just dont get it...u animate the face,export baked anim, bring it into whatever software renders (max i guess in ur casE) and then just connect it to the animated body?
if im talking stupid, correct me...

pS. i know its a bit offtopic, but im really curios

elvis75k
01-24-2006, 02:57 PM
If I were the team behind Face Robot, I'd be embarassed to mention this on my reel.

It looks like someone couldnt figure out how to use Face Robot and decided to use Poser instead.

hahahahaa.. right!!

gunslingerblack
01-24-2006, 03:08 PM
Originally Posted by zzacmann
If I were the team behind Face Robot, I'd be embarassed to mention this on my reel.

It looks like someone couldnt figure out how to use Face Robot and decided to use Poser instead.


ouch...but....it's somewhat true

and zzacmann the team behind face robot is softimage and blur studio incase u didnt know

but the video could be bad artistic choice on behalf of the director and individual artist or artists, or a bad combination of both

shingo
01-24-2006, 10:33 PM
What is the OSIPA System you are referrring to?

thank for takin the time to give us that answer my friend,

it's unfortunate that limitations are imposed but thats the nature of the biz i guess.

i myself am a student of the osipa system of facial setup, which right now provides me with a very rich set of facial morphs to choose from. i've only been doin facial setup for a short time but, i credit the osipa system with landing me a job, it definitely made me look good.

thank you for the explanation on face robot, it does seem to have quite a bit of potential.

ps

quaaaaaaaaaaaaaaaaaaaaake!

gunslingerblack
01-24-2006, 10:35 PM
well i only call it the osipa system because i learned it in the book on facial animation written by jason osipa "stop staring" u can check google or amazon to find it.

shingo
01-24-2006, 10:42 PM
Soft are keeping Face Robot pretty close to their chest for now and working with just a handful of high profile clients. They probably want more feedback and suggestions to get it to a point where it's more suitable for the wider market.

It's got some pretty mazing stuff in it and a huge amount of control. It works around a picking session and calculates where the bones of the face lie and the soft tissue areas. When you move the control markers, it respects areas like the teeth and cheeks. Once it's set up, it produces all kinds of maps, tension and stress maps that you can use to control the coloration of the skin as it stretches and use the maps to control displacement/bump detail. You can also paint creased and lines on the face and control how the creased respond to stretching with yet more maps.

It seems very mo-cap centric, but maybe that has changed. The retargetting tools are also very lfexible from what I saw in a demo.

As far as modelling goes, XSI is rocks and is at least as good as MAX if not btter (for organic stuff). Certainyl it;s easier to poly model than Maya, but that's just MO.

EpShot
01-24-2006, 11:44 PM
ps

quaaaaaaaaaaaaaaaaaaaaake!

much appreciated, btu i'm one of the battlefield geeks ;)

gunslingerblack
01-25-2006, 12:03 AM
i prefer battlefield to quake myself

its more fun to work in a team.

jason-slab
01-25-2006, 07:10 AM
yes battlefield rocks! just got SF, its pretty darn cool

jason

skunk184
01-25-2006, 09:40 AM
I wonder what bono thinks of face robot ...I am sure he has an opinion as he has one about everything else .... save this save that blah blah blah

huydoan.com
01-25-2006, 08:59 PM
http://img96.imageshack.us/img96/2314/u22ts.th.jpg (http://img96.imageshack.us/my.php?image=u22ts.jpg)

here's an index of the video. the entire video was posted in this newsgroup recently:

alt.binaries.mpeg.video.music

NUKE-CG
01-26-2006, 08:19 AM
I feel sorry for the artists, listening to U2 over and over again to sync.

/me cringes.

CGTalk Moderation
01-26-2006, 08:19 AM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.