Slashdot Mirror

Like this one, maybe?

Autodesk bought Softimage XSI for cheap, and just killed it to remove competition from their flagship products 3DMAX and MAYA. There is a huge thread about this over on cgsociety.org: http://forums.cgsociety.org/sh... ...Basically, anybody who built their studio pipeline around Softimage XSI, including many indy game developers, is royally screwed. Softimage's most powerful feature "ICE" (a multithreaded, node-based visual programming language that lets even non-programmers build custom tools and functions inside Softimage) is being migrated to Autodesk Maya instead. Its going to be called "Bifrost", as it is the "second coming" of Softimage ICE. Many Softimage users are wondering what other 3D software they can migrate to. Many are considering migrating to SideFX's "Houdini" (http://www.sidefx.com/), which is a very powerful procedural-animation software used extensively in some of the most complex VFX shots you see in Hollywood films, like the character shatter effects in TRON LEGACY. Some are considering moving to the open-source Blender 3D software, to escape from Autodesk's business policies completely. Basically, Autodesk bought Softimage, slowly killed it, ripped out the best bits, and is now forcing Softimage users to migrate to either 3DMAX or Maya, which are Autodesk's cash cows in the Media & Entertainment division. A lot of people are very pissed off about this. But this is hardly the first time Autodesk has killed a successful product (e.g. the once-excellent Autodesk Combustion), because it didn't make enough money for Autodesk's profit hungry shareholders. A sad day for Softimage XSI users. It has powered films ranging from the first Jurrassic Park to the recent LEGO movie. It was particularly strong at pulling off complex character animation, including complex muscle-and-sliding-skin simulations (e.g. the all-CG primates in "Dawn Of The Planet Of The APES"). XSI was a good CG software. It will be sorely missed by many... If Blender can get its UI overhaul right in the next release, some XSI users may migrate to the open-source software.

Here you go. http://wiki.cgsociety.org/index.php/Comparison_of_3d_tools
Blender compares extremely well with other sweets, especially now with it's new dynamic topology sculpting
http://wiki.blender.org/index.php/Dev:Ref/Release_Notes/2.66/Dynamic_Topology_Sculpting

Reposting this at a higher level, since I posted as a response to an AC below you.

It's a very recent example of exactly what you mention:

http://peregrinelabs.com/2012/04/to-our-us-customers/ and here is a long forum thread discussing it, including posts by the patent holder: http://forums.cgsociety.org/showthread.php?f=59&t=1048283&page=1&pp=15

In my view it's clear to see how this kind of thing has a stifling effect on innovation, business, and the economy of the US, but I suppose some do not share that view.

Here's one that just happened a couple of months ago: http://peregrinelabs.com/2012/04/to-our-us-customers/ - it's an European 3d hair and fur program that is no longer going to be sold in the US because of a patent dispute. You can also read a whole 140+ posts long forum thread about this situation including the patent holder defending himself, here: http://forums.cgsociety.org/showthread.php?f=59&t=1048283&page=1&pp=15

I just happened to know this example of exactly what the GP post mentioned because I frequent the CGSociety forum. I imagine from what GP said that there are probably many other examples.

And not just in terms of hiring low-wage people in 3rd world countries. Ultimately there is no magical American advantage in smarts and intelligence that says we're going to have an advantage in inventing stuff.

Starting a manufacturing plant requires a huge investment, cooperation of the local government, etc., etc., but the next big thing in software can still come out of someone's garage, and that garage can be anywhere, even some "backwater" country where they don't even have garages. A couple motivated developers and a couple mangy PCs are all you need, and that's rapidly becoming available to a large percentage of our 7,000,000,000 people.

As an example of the internationalization of creative talent, take a look at CGSociety's GCChoice gallery on the front page at:

http://www.cgsociety.org/

which shows the latest images voted to be some of the best submitted by artists using computer graphics software. Just like software development, the tools to produce these images are available to anyone in the world with a computer, and that is reflected by the international nature of these images. Just a quick look today shows the most recent top images coming from: Slovakia, the UK, Sweden (x2), Mexico, Iran, China (x3), the USA (x3), Turkey, Korea, and Singapore.

Maybe we can use our big money to quickly buy up the talent when it appears, but the "next big sotware things" are ultimately going to come from all over the world.

G.

The benefit is that raytracing is a more natural way to do 3D rendering (by simulating light). It basically gives you a unified model of rendering. For programmers, this means almost every effect can be done very simply, and in a more physically realistic manner. Shiny surfaces, shadows, mirrors, are actually trivial to do with raytracing, but with current hardware (which performs rasterization), those effects are all hard to do, usually involving multiple rendering passes and dirty hacks. The end result is also less realistic. One problem that is *very hard* to tackle (I find) using rasterization, is real-time lighting, and again, it's very easy to do with raytracing.

Current efforts at real-time raytracing are limited because there isn't that much research into the topic, compared to the billions of R&D that went into rasterizing 3D accelerators. Lots of shiny surfaces is also a combination of poor artistic choices and trying too hard to demo the technical capabilities available. What I've seen involved either doing all the work on the CPU, or designing custom FPGAs to do hardware acceleration with a fraction of the computational power that a modern GPU has.

Still, in the long run, raytracing is *the way* to get the most realistic graphics possible. Raytracing is what's used by the the most realistic software rendering packages available today. Not saying you'll have graphics like these in real-time super soon, but it's worth looking forward to:

Rendered using VRay
Mental Ray
Mental Ray
VRay

The benefit is that raytracing is a more natural way to do 3D rendering (by simulating light). It basically gives you a unified model of rendering. For programmers, this means almost every effect can be done very simply, and in a more physically realistic manner. Shiny surfaces, shadows, mirrors, are actually trivial to do with raytracing, but with current hardware (which performs rasterization), those effects are all hard to do, usually involving multiple rendering passes and dirty hacks. The end result is also less realistic. One problem that is *very hard* to tackle (I find) using rasterization, is real-time lighting, and again, it's very easy to do with raytracing.

Current efforts at real-time raytracing are limited because there isn't that much research into the topic, compared to the billions of R&D that went into rasterizing 3D accelerators. Lots of shiny surfaces is also a combination of poor artistic choices and trying too hard to demo the technical capabilities available. What I've seen involved either doing all the work on the CPU, or designing custom FPGAs to do hardware acceleration with a fraction of the computational power that a modern GPU has.

Still, in the long run, raytracing is *the way* to get the most realistic graphics possible. Raytracing is what's used by the the most realistic software rendering packages available today. Not saying you'll have graphics like these in real-time super soon, but it's worth looking forward to:

Rendered using VRay
Mental Ray
Mental Ray
VRay

The benefit is that raytracing is a more natural way to do 3D rendering (by simulating light). It basically gives you a unified model of rendering. For programmers, this means almost every effect can be done very simply, and in a more physically realistic manner. Shiny surfaces, shadows, mirrors, are actually trivial to do with raytracing, but with current hardware (which performs rasterization), those effects are all hard to do, usually involving multiple rendering passes and dirty hacks. The end result is also less realistic. One problem that is *very hard* to tackle (I find) using rasterization, is real-time lighting, and again, it's very easy to do with raytracing.

Current efforts at real-time raytracing are limited because there isn't that much research into the topic, compared to the billions of R&D that went into rasterizing 3D accelerators. Lots of shiny surfaces is also a combination of poor artistic choices and trying too hard to demo the technical capabilities available. What I've seen involved either doing all the work on the CPU, or designing custom FPGAs to do hardware acceleration with a fraction of the computational power that a modern GPU has.

Still, in the long run, raytracing is *the way* to get the most realistic graphics possible. Raytracing is what's used by the the most realistic software rendering packages available today. Not saying you'll have graphics like these in real-time super soon, but it's worth looking forward to:

Rendered using VRay
Mental Ray
Mental Ray
VRay

The benefit is that raytracing is a more natural way to do 3D rendering (by simulating light). It basically gives you a unified model of rendering. For programmers, this means almost every effect can be done very simply, and in a more physically realistic manner. Shiny surfaces, shadows, mirrors, are actually trivial to do with raytracing, but with current hardware (which performs rasterization), those effects are all hard to do, usually involving multiple rendering passes and dirty hacks. The end result is also less realistic. One problem that is *very hard* to tackle (I find) using rasterization, is real-time lighting, and again, it's very easy to do with raytracing.

Current efforts at real-time raytracing are limited because there isn't that much research into the topic, compared to the billions of R&D that went into rasterizing 3D accelerators. Lots of shiny surfaces is also a combination of poor artistic choices and trying too hard to demo the technical capabilities available. What I've seen involved either doing all the work on the CPU, or designing custom FPGAs to do hardware acceleration with a fraction of the computational power that a modern GPU has.

Still, in the long run, raytracing is *the way* to get the most realistic graphics possible. Raytracing is what's used by the the most realistic software rendering packages available today. Not saying you'll have graphics like these in real-time super soon, but it's worth looking forward to:

Rendered using VRay
Mental Ray
Mental Ray
VRay

I work in the visual FX business, and "Praxis rings" have been mocked as cliche for well over a decade now.

http://forums.cgsociety.org/ http://www.cgpad.org/forum/

That must be why I couldn't register it.

I tried to use this as my profile pic but they said they had a prohibition on pictures of pets. I haven't been back since.

performance: 20x speed-up ("from what" is unanswered at this time) to 200x speed-up down the line
limits: limited more by your machine than the card
dynamic scenes: it's an accelerator - if the renderer can, then it still can with this card
sorting (accelerations structure building, I think you mean?): wouldn't know but seeing as it's supposed to accelerate the ray tracing process, I would imagine it's either on the card or via their own algorithms in software
photon mapping/MLT/etc.: it's an accelerator. If the method traces rays, then the card can accelerate it. This applies to most of the methods you mention.
performance comparison: should be coming up later - but presumably much better than software-only methods and better than GPU-assisted methods
image quality: it's an accelerator - image quality would depend mostly on the renderer that invokes the card, not the card itself.
geometric primitives: I believe they had at least a sphere thing going on.. presumably, again, that means that other mathematically-defined surfaces could be calculated 'as is' as well. If not, cast them to a mesh - doesn't hurt that much.
textures: n/a
algorithms: their own
acceleration structure: probably a closely guarded secret

if you want actual answers, try...
http://forums.cgsociety.org/showthread.php?f=59&t=739494
and
http://twitter.com/causticgraphics

OK, I have to show off - I did something similar four years ago.
Post
Webpage
I'm a much better programmer now, just for the record. :)

It's really not that difficult to write the software. I managed to follow this paper and get it working in a few evenings....

http://www1.cs.columbia.edu/~ravir/cvpr07.pdf

I only had a go after seeing that a guy on CG talk managed to do it (and he's an artist - not a programmer).... http://forums.cgsociety.org/showthread.php?f=109&t=636851

Not a matter of time, done already. See: CGTalk, in particular Weird Science.

Not a matter of time, done already. See: CGTalk, in particular Weird Science.

I'll take ray tracing, TYVM. I think if ray tracing had near the amount of resources invested in specialized hardware, I think we'd see things like this, and this in our games.

I'll take ray tracing, TYVM. I think if ray tracing had near the amount of resources invested in specialized hardware, I think we'd see things like this, and this in our games.

Maybe in the past this was true, not anymore though. The quaddro/firegl cards just dont offer enough bang for your buck so the majority of cg studios have opted for the highend gaming cards. The 8 series cards are causing problems though... http://forums.cgsociety.org/showthread.php?f=7&t=510244

Not a film, but Babylon5 opened the floodgates to full on CGI use in TV series. The large highly dynamic space battles were also a first generally (afaik).

B5 was the death knell for miniature work in TV (Trek, etc), Foundation went on to do the effects for DS9 and Voyager. A lot of other shows, such as BSG & SG, owe a great deal to the pioneering work done back then - and it was all done on a shoestring budget (read: networked Amigas in an apartment with cables running through rabbit cages!).

Until B5 most TV producers didn't think good CGI was a possibility with the budgets/timescales limitations they had to work with. A lot of jaws dropped when B5 first aired, mine certainly did (it's what inspired me to became a 3D artist).

CG Society has quite a nice writeup on the work: http://features.cgsociety.org/story_custom.php?sto ry_id=3971

OK. For newbies, the Blender interface is difficult because you have to learn it all. There isn't much hand-holding. But, when you ask professional modelers, they scream hard at anyone who wants to touch the U.I. as its *PROFOUNDLY* much faster than (pick your professional $100,000/per 10 minute license 3D application here). Blender is starting to rule over some other 3D applications (high end special effects), etc. As for the 'old codebase', some of it is old, but a lot of it has been replaced within the last 2 years, and more gets replaced with every new version. Certainly, all of the Google summer-of-code stuff is new. I wouldn't call it an 'also-ran' or 'its open source, so its ok' either. Compare it against commercial applications. There isn't a lot that isn't 'there', and in many cases, it bests a lot of them (sometimes all of them). See for yourself here:.

Hope they don't fuck it up. I mean, how can you 'actionize' quantum physics?

(You could have a car chase, but by observing the chase you collapse the wave-form and the cat dies. Hmmmm, maybe Lynch could pull it off.)

Here's some amateur-created trailers:

http://features.cgsociety.org/story_custom.php?sto ry_id=3987&page=2

There is a chart created by the 3d graphics society that does a reasonably good job doing a side by side comparison, found here. Blender doesn't ususally include as many pre-built pieces, but compares very favorably in advanced algorithms and high-end features. It leads other 3D applications in a few areas too (yes, I know some of them cost upwards of US $17000). More than one PhD has contributed part or all of their PhD thesis to Blender (in terms of software), and many others through papers. I would also bet that at roughly 10 megabytes, its the smallest 3D app. that does all that it does (compare Maya at 170 MB).

Was reading EON - that's a pretty amazing book when it comes to imagination. Anyway - I was looking up some illustrations for the book and found this site - the CG Society's competition (cleared by Bear) to create your own visuals for the book.