Slashdot Mirror
GPL'ed 3D Modeler And Renderer
hardgeus writes: "A lot of people (including me) have said that what the Open Source world needs is a 3D counterpart to GIMP. Well, it looks like it's finally here: The OpenFX Open-Source 3D modeling, animation and rendering suite . It has a renderer and raytrace engine, NURBS support, kinematics-based animation, morphing, a plugin API - and it's under the GPL. Currently only for Windows, but they're working on a Linux and FreeBSD port." There's this and Blender - what other options are there?
There's Renderman Interface Bytestream. RIB is not like SVG, however. It's more like PostScript, in that it's intended to be an interchange format between modellers and renderers, rather than between modellers and other modellers.
sub f{($f)=@_;print"$f(q{$f});";}f(q{sub f{($f)=@_;print"$f(q{$f});";}f});
A good GPL'd geometric kernel with support for offsetting, rendering and a bunch of other gooodies. It's supported on Linux, Windows NT, SGI Irix, IBM AIX, and Sun Solaris. Check it out at: <A=href http://www.opencascade.org/>Open Cascade</A>
MMORPG fan-boy? Prove your worth
I actually don't know if they intend to release an opensourced version of the modeller, but since the render engine was developed by Mental Images and not by them they won't be allowed to release the code.
This is also the same render engine the Softimage|XSI uses, tho in a more powerful version.
Linux 3D Graphics Programming has some links to other Linux 3D modelers. Invention 3D comes to mind as a promising modeler.
What's more, those features are probably going to be things that the average person would never suspect. Like "the renderer must be able to handle more geometry than you have RAM for without thrashing", or "you must be able to specify the coordinate system that shading happens in separately from the coordinate system of the geometry", or even "you must be able to tune the shading sampling rate separately from the image sampling rate".
I might add that all of the above companies (ILM is an exception with respect to renderers because Pixar used to be part of Lucasfilm, so they have good licensing terms for Pixar's renderer) have written their own animation systems and renderers precisely because nothing on the marked did what they wanted.
That's true, but the cost of hardware and software is (generally) not as important as the cost of people. Look at Pixar's render farm, for example. It would have been cheaper to buy PCs and slap a free Unix on them all than buy the farm of 100 Sun E450s that they have. However, Sun E450s can fit 14 CPUs in each box. That's 1400 CPUs in 100 physical machines. Compare to number of physical machines you'd need if you bought PCs instead. Fewer machines require fewer sysadmins. And since sysadmins are more expensive than machines in the long run, it turns out cheaper. Similarly, the cost of a Maya licence is not nearly as much as the cost of a talented modeller or animator to use it.
sub f{($f)=@_;print"$f(q{$f});";}f(q{sub f{($f)=@_;print"$f(q{$f});";}f});
Looked at the website, and it is apparently an application called SoftFX, which has been on the market for years. It was never much of a commercial success. I looked at it a few times over the years and it's OK, but pales in comparison to apps like Maya, Soft, Max, etc... Seems like the author's given up on selling the thing and has decided to give it away. Maybe some clever people can add enough cool features to make it production worthy. Hook it into BMRT for a start. I think this is a great idea. Maybe authors of other failed commercial ventures can do the same thing - give away the code via GPL. Some of it might find life elsewhere.
Forget it, it doesn't exist.
3D programs take many different approaches to representing objects and their interactions, from simple polygonal meshes to surfaces defined in terms of spline patches to volumetric representations to parametrically defined solids, to implicit surfaces (Meta-Blobs), with a plethora of texturing, partitioning and animation systems. You'd also have to include any physical and dynamical parameters - weight, mass etc. since this is very important in Engineering/FEA fields
X3D, VRML, DXF, IGES, OBJ, 3DS, OpenNURBs etc. only deal with a certain subset of these representations and you'd have a huge job to come up with a file format that could efficiently represent all of them.
The only way you could really do this would be to define a standard API for access to all this information, and let vendors simply implement the parts that their apps need.. i.e. we don't need a standard 'file format', we need a standard 3D codec system.
Just like any program n Windows can manipulate DivX or MPEG video on Windows without implementing it's own DivX loader/saver, as long as the necessary codec is available.
Still, 3D is changing fast, and the breadth of the field makes creating any kind of 'standard' very difficult.
I gots ta ding a ding dang my dang a long ling long
Don't forget the POVLAB development effort, which is beginning to pick up the pace. POVLAB has been around for years as a modeller for POV-Ray, and is just now being rewritten in C++ and made cross-platform with wxWindows.
See the POVLAB development site for details and to find out if you would like to help.
~ Give me 101 plastic soldiers, and I will conquer the world.
An open source project where open source operating systems are the second-class citizens.
Traditionally, open source stuff was taken care of on the Unices, especially Linux and BSD, first. Windows was the second-class citizen simply due to it's lack of portability.
As time went, open source software appeared that made software developed under the Unices easy to port to Windows. SDL and GTK for Win32 come immediately to mind. The Gimp doesn't seem to lag very much, and the OS SDL games out there don't seem to lag at all.
Now we have open source developed under Windows that people want under the Unices. I bet that's going to be fun to deal with. That's going to end up having to be completely rewritten under SDL/GTK/Qt or something before being useful.
There are a bunch of people here saying, "Yay, 3d software for Linux."
Not for a while...
Is there a standard out there like SVG(a nice XML based format that I'm counting on getting popular) for 3D models? I'd love to make a bunch of 3D models and such for fun and be able to use it in one edior or the other.
Well, lets face it, blender's a bitch to learn. the UI plain out sucks. The only real decent UI i've seen for a linux modeler/renderer is Moonlight3d, which has all but died (not opensource, but based on opensource code. and there'r sounds of life commin from globfx, rumored to be the creators of moonlight3d)
~~~ They call me Little John, but don't let the name fool you...in real life I'm very big.
anyone got a mirror of this? looks like they've been slashdotted...
~~~ They call me Little John, but don't let the name fool you...in real life I'm very big.
---
if you want something that has a sharp UI and runs on Linux, wait a few months for Maya to come out.
Between hot-boxes and hot keys, the menu is never needed and most commands are a small mouse move away.
Sure it will cost a bundle, but CG is one of the places where quality does not come cheap.
-I just work here... how am I supposed to know?
Although I generally agree with you, I have to say that I laughed out loud at 'The programs are utterly immense, yet need an interface free of ad-hoc additions' when talking about Max. That's Max the poly-modeller, Max the broken-bezier-patch-modeller and Max the NURBS modeller.
:-)
:-)
:-)
For the record, I didn't mean to suggest that Max is entirely successful along those lines, allthough I was thinking of Max from a different point of view - 3DS R4 was getting to the point where the only way to introduce the new features made possible by the new technology was going to be seriously ad-hoc. 3DS Max re-did the interface from the ground up to incorporate those things consistantly. Regardless of how successful it was, it was far more successful than trying to patch the features into 3DS R4 would have been. That was years ago, and while I haven't used a recent version of Max, it wouldn't surprise me if 3DS Max is now where 3ds R4 was (ie the interface is starting to break down under the weight of features that could not have been anticipated, and so could greatly benefit from an interface overhaul). I don't know if that's the case or not, but it sounds like at least you would describe it along those lines
I've been out of the high-end 3d loop for a while
For me anyway, this is a pretty big problem - these packages can take years to truly master (partly due to their complexity) so you don't want to throw half of that investment away by throwing out the UI every year, but the UI breaks down pretty quickly due to the speed of innovation, so it's a choice between two evils. If only another way could be found. Direct neural connection or something
What about using winelib? Isn't this exactly what winelib is supposed to help with? Easy porting of code from windows to *nix machines? Especially with open code...
Power to the Free Software folk! Using Winelib I bet they could get a port in a few weeks...
-- Erich
Slashdot reader since 1997
Ok, I admit from the start that it's unfair to compare this to the likes of 3ds (though it's hard for me not too, as I haven't done much work with lower-end stuff), but at first glance, this doesn't look like it will be useful for much for a long time (other than really basic stuff, like simple web animations etc). And I suspect it might even look worse at second glance.
:-) with the expertise to make a non-raytracing renderer that produces raytracer-quality (or better) results at a fraction of the render time.
:-)
Gimp, while not photoshop, is still at a level where it offers a viable alternative (for a fair amount of uses) to one of the leading packages. Even mid-range 3d apps dwarf photoshop in complexity (and usually price), so it's no surprise that this (great) start in that direction has a long way to go, but I don't think this thing can be called the Gimp of 3d - it just doesn't offer a serious alternative, and doesn't look like it will for a very long time (if ever).
But for a very long time now, I've been of the view that open source simply cannot produce a 3ds MAX or SoftImage, or whatever (or at least not with the current methods of production). The programs are utterly immense, yet need an interface free of ad-hoc additions and localised revisions to be a fast production tool, not to mention there being so few people in the world (virtually none
Compounding that is that the 3d apps evolve much faster than the like of photoshop (which already seems to evolve faster than Gimp, (but I haven't compared the latest versions, I might be wrong)), so the successful open-source model of gradual accumulated improvements - great for a word processor - just won't work.
Such a project would have to be so full on that it would have to be full time for a lot of people for a very long time. I can envisage some business models which could allow this, but I'm not going to hold my breath. (Besides, holding no hope can allow for nice surprises
Hmmm. I realise this entire post sounds like a petulant bashing of what is a praiseworthy and excellent piece of work, but I have this sinking feeling that the next thing we know, people will be touting this as almost up there with production-level apps - as viable an alternative as Gimp to photoshop, and I just don't want to see that happen. I can so imagine some rabid zealot defending open-source along these lines and thus teaching people who are unfamiliar with open source that o/source is a "viable alternative" in the same sense that amputation is a viable alternative to antibiotics.
Prevent email address forgery. Publish SPF records for y
2 things:
First:
Blender isn't open source.
Second:
The submitter, whose comments are in italics, stated that this is the first open source 3D program that he knows of. Michael (of slashdot who posted the storry) then stated after this "There's this and Blender - what other options are there?"
That's why it was a revelation to the poster. I suggest you work on your reading comprehension before rushing to post.
Prevent email address forgery. Publish SPF records for y
It needs a GPL modeller, and I've put a bit of work into Giram. It's not bad, but it needs more work by better GTK+ programmers than I. Runs on Linux & BSD. Help. Please.
Vik :v)
On the other hand, universities won't be paying list price for software of this type. Engineering departments typically get huge discounts for real high-end software, so that they produce graduates who know how to use it, who go onto to work for companies that buy it. Compare the student to professional costs of a package like MATLAB. And universities will get all the support they could want, plus access to the full, very specialized software package which, like another poster said, will contain features that wouldn't even make sense to people outside the niche market. Compare MATLAB to Octave. The core engines might be comparable, but Octave doesn't offer the specialist tools.
Another blessing is that it will force the cost of the professional version of these programs down.
Unlikely, for the reasons above.
. Especially as people start copying the features in the top end programs and add them to the GPL'd stuff.
Again, this doesn't help. Why wait for an imitation when the commercial product will deliver return on investment so quickly? (And on a separate note, why doesn't the Open Source camp innovate?).
PVMPOV has supported parallel rendering of a single frame on a cluster (or a massively parallel supercomputer, or a network of workstations) for years.
For a cross-platform, GPL'd multitrack sound editor, check out Audacity:
http://www.cs.cmu.edu/~music/audacity/
I'm suprised that no one has mentioned K-3d. I haven't tried it myself... But I've always been fond of screen shots and they have some cool ones. Also k3-d is gpl and runs on linux and windows.
BTW their website looks like crap. It used to look good. Brilliant even compared to how it looks now. Perhaps setting a background color would help?
There's also SART, the renderer I'm working on. Its current focus is rendering, but there'll be a modeller in a, well, not too near future. Unfortunately I don't have the time to work on it as much as I'd like.
Current features are full programability (using guile), support for NURBS, blobs, parametric and implicit surfaces, volume rendering (including nonuniform textured volumes), radiosity, postprocessing. Check http://petra.zesoi.fer.hr/~silovic/sart for more info.
---
I refuse to use
So, if you know about Blender, then you realize that there has been an OpenSource 3D modeling/rendering package available. So why is this new one such a revelation to you?
And you forgot one. Kinetix/Discreet/Autodesk/(whoever they are this week) will be releasing an OpenSource version of the Max modeller. I don't think it will have rendering capabilities, though. They're putting it out there so people can create content for their favorite game mods.
There's only three features I know of not in GIMP.
1) CMYK color. This is the biggest thing. There is a proprietary plugin that implements this, though
2) Recordable "Actions". This are _much_ simpler than scripts.
3) The Editable Text utility is more advanced
Then, of course, GIMP has a feature that Photoshop doesn't have - scriptability in multiple languages, currently Scheme and Perl. And, it has a batch mode so you can use these to do web-enabled scripts. Check out www.cooltext.com to see this in action.
Also, remember that free software itself is a feature, because you can pay someone to make any sort of modification you need for your purposes.
Engineering and the Ultimate
Universities will love things like this. There is a need for three dimensional rendering tools in engineering and art classes, and schools do like to trim dollars. These classes often don't require the fancy bells and whistles that are required in the top end rendering programs. Likewise people who don't use these tools profesionally (and have a decent budget) will gravitate to these things.
Another blessing is that it will force the cost of the professional version of these programs down. As I have observed before, the bright side about open source programs is that they raise the bottom line. All of a sudden te functionality of OpenFX becomes the baseline standard and people have to look at the other features and ask themselves if they really need that other stuff. Especially as people start copying the features in the top end programs and add them to the GPL'd stuff.
I don't know how many different OSS 3D renderers there will be after a time. I suspect that there's really only going to be mindshare (given the resources required to create a program like this) in one program. There may be major rearchitectures over time, but I think there will be consolidation on that point between this and any other GPL'd renderers out there.
What will be interesting is any evolution towards cross fertilization with software like Crystal Space, the GPL'd 3D engine. Sooner or later people will think it might not be a bad idea to make sure stuff created in the modeller and renderer works directly well with an engine to use such things in games. Open source makes such things possible.
If you're interested in some other 3D software for Linux (some GPL, some not), there's 3dom, 3dpm, Behemot, G3D, Giram, 3delight, AC3D, and of course Blender as mentioned above.
Come on Karma, don't fail me now! The Linux Pimp
--It's Pimptastic!--
These guys suddenly disappeared? Google says that the website, www.moonlight3d.org, had a "please come back later" message on it. But now, it appears, the site itself is gone. I remember looking at this package for a while, but suddenly the web site shut down, with some nebulous reason cited. Weird.
I haven't had much luck with the free/open alternatives out there. I used to use Ray Dream on Windows to goof around with 3D, but it had a lot of bugs and many limitations. But still I'm not thrilled about having to shell out $500 or more for an OK package, or several thousand for a professional one.
It seems though, that all the bells and whistles that go into making a professional 3D package like Lightwave would be a daunting task for an open/free project. Probably similar in scope to duplicating PhotoShop. Anyone care to comment on what percentage of PhotoShop functionality that the Gimp has implemented?