[Clam-devel] Formats & Protocols for 3D scenes...

Pau Arumí parumi at iua.upf.edu
Wed May 14 10:32:29 PDT 2008


Wow, very interesting collection of standards and projects.

Luckily, for the Blender integration project we just need a) a 3D
geometry file format and a b) OSC based source/listener positioning
protocol.

I don't have an opinion on a). Natanael, could you collect a list of
Blender-friendly alternatives?
For b) we need a very trivial format, since it's just a matter of
positioning sources and listener using OSC. So, starting with a (subset)
of SpatDIF i think makes sense. 

Pau





El dl 12 de 05 del 2008 a les 02:57 +0200, en/na Xavier Amatriain va
escriure:
> Hi Natael,
> 
> Interesting question but I am afraid there is no easy answer.
> 
> First (and just to add some more confussion :-), another possibility 
> would be using VML from the Verse project [1]. This project is very much 
> related to Blender (one of the partners) and although using VML would 
> tie us into their data model, Verse is released with a BSD license.
> 
> As for SPATDIF I am not sure this is currently anything more than 
> wishful thinking. I know they are organizing a panel for this year's 
> ICMC and I know some of the people involved. Knowing the fate of SDIF I 
> doubt this will ever become a standard. Another downside is that I don't 
> think it will have integration with 3D graphics anytime soon. That said, 
> we have always tried to use SDIF in CLAM as an open interchange format 
> for synthesis data so using SPATDIF would make sense... if it is in any 
> usable form now.
> 
> Using BIFS and AudioBIFS [2] would seem a good idea, especially knowing 
> that SDIF can be transcoded into SAOL and this can be used inside 
> AudioBIFS [3]. Note that BIFS is not combined with VRML, it is simply an 
> extension of it. The downside is that I think there is no open reference 
> implementation of this that I know of (although Ross has implemented 
> this in audiomulch [4]).
> 
> Finally, I think X3D shares a lot with BIFS. As a matter of fact the 
> MPEG standard already includes a link to X3D [5]
> 
> So, probably if I had to chose out of the box I would go with BIFS. 
> Still, I think the issue requires some more time to study and maybe even 
> some preliminary tests to see how things would look one way or the other.
> 
> Hope it helps!
> 
> [1] http://verse.blender.org/vml/
> [2] http://sound.media.mit.edu/mpeg4/sa-bifs.html
> [3] 
> http://archive.cnmat.berkeley.edu/ICMC99/papers/saol+sdif/icmc99-saol+sdif.pdf
> [4] http://www.audiomulch.com/~rossb/code/sa/SAQuickref.htm
> [5] http://en.wikipedia.org/wiki/X3D
> 
> Natanael Olaiz wrote:
> > Hi!
> >
> > For my SoC proposal, I'm wondering about the formats and protocols to 
> > use in the Blender/CLAM integration. So, I want to ask you for your 
> > opinions.
> >
> > The basic idea is to import 3D Blender scene objects parameters onto 
> > CLAM, on RT and offline. For RT I talked to Pau to use Open Sound 
> > Control, but anyway I need to define some 3D Scene descriptor format 
> > to use over it. For the offline implementation I need to define also 
> > the intermediate file format.
> >
> > I think a great option could be to use SpatDIF [1], since it claims to 
> > be a standard for this kind of data. It uses OSC for RT, and SDIF as a 
> > intermediate file format for non-RT purposes. The only problem that I 
> > see with it is that is a new proposal, and I didn't found so many 
> > implementations yet (so, I don't know if it would be a real standard).
> > I also saw that the MPEG4 standard uses BIFS [2] + VRML. As I 
> > understand, this option is more closed, but could be interesting to 
> > make a MPEG4 render system within CLAM. Plus, I saw that X3D [3] is 
> > some kind of sucessor of VRML (with the advantage of using an new XML 
> > format instead the binary of the VRML+BIFS), in which the sound is 
> > integrated too [4]. But as I saw, the free implementations (of 
> > browsers and editors) of X3D are not so good right now, and the major 
> > ones are in Java (language that I don't know well -and neither 
> > like...-). BTW, there is an available Blender X3D exporter plugin 
> > right now...
> >
> > What the CLAM gurues think about it? I saw that there is some 
> > mpeg-spectral-descriptors code in CLAM... so, do you want to improve 
> > the mpeg support with X3D, or just use a new proposed open standard 
> > (SpatDIF)?
> >
> > Anyway, it wouldn't be so difficult to make one using the other as a 
> > base...
> >
> >
> > Best regards,
> > Natanael.
> >
> > [1] Spatial Sound Description Interchange Format: 
> > http://www.jamoma.org/wiki/SpatDIFSpatialSoundDescriptionInterchangeFormat 
> >
> > [2] BInary Format for Scene: 
> > http://www.chiariglione.org/mpeg/technologies/mp04-bifs/index.htm
> > [3] X3D specs: http://www.web3d.org/x3d/specifications/
> > [4] X3D and VRML Sound Components: 
> > http://www.tml.tkk.fi/Opinnot/Tik-111.590/2002s/Paperit/pohja_x3d_sound_OK.pdf 
> >
> >
> > _______________________________________________
> > Clam-devel mailing list
> > Clam-devel at llistes.projectes.lafarga.org
> > https://llistes.projectes.lafarga.org/cgi-bin/mailman/listinfo/clam-devel
> 
> 
> _______________________________________________
> Clam-devel mailing list
> Clam-devel at llistes.projectes.lafarga.org
> https://llistes.projectes.lafarga.org/cgi-bin/mailman/listinfo/clam-devel





More information about the clam-devel mailing list