This is my feeble, semi-automatic, hand corrected translation of the reply that Diego García Huerta (Senior Pipeline & Tools Developer at Blur Studios) gave me in this cg-node.com forum
I asked him to help me with my research, with:
(B) Software (programs)
(C) References (interviews, webs, manuals, whatever)
1.- A general approach to analyzing a business workflow. A recommended generic reference book ("The bible of...") that allows to settle a proper theoretical background to build on top of it.
2.- Detailed analisys of 3D production in different companies. I try to define a general workflow, based on real cases.
3.- Programs that allow to manage tasks, assets, processes, files, etc. Specially GPL ones.
4.- Nomenclature conventions.
This is an interesting subject for a thesis, for sure, and indeed more research is needed in this field.
Obviously in order to organize a studio with over 40 persons you need some sort of help in order to speed up the search for files, know in real time what is going on in the production, automate processes as asset creation, assignment of working hours, semiautomated budget calculation, etc.
To begin with, a good directory structure avoids the production growing as a wild plant out of control.
We have a typical structure of directories based on Project / Sequence / Shot. In each shot you can find all the necessary files for layout, animation and scene assembly that are normally referenced by character´s meshes and rigs.
Oher studios follow the same methodology, but it is difficult to find out exactly how a CG/VFX company pipeline, besides some glimpses in the DVDs extra of some films, mainly CG.
Perhaps the best way is interviewing people from those studies, and check wether this is an open subject, as you are doing with me!
After the Directory Structure, you need a Naming Convention (NC)
I have heard terror histories in Disney where a long time ago it simply consisted in a few letters and numbers.
Here in Blur, the name of the 3D objects, files, etc. are more or less indicative of their content (which complicates their management from the point of view of programming, and is more vulnerable to human error)
Other studies decide to put everything in the name of the file, from the person that uses it, to the day that was recorded.
I personally plead for clarity, neither too cryptic, nor "everything in the name", although it means more problems for the programmers
Here in Blur we have our own tools that put tags to the 3D objects, or files, so that with said tags, we can automate processes in a simple way, or identify different types of objects, like for example the meshes that have to be exported as a cloud of points for scene assembly.
An example of NC of an object:
With this name you can tell it is a "Templar" character asset, with probably other two duplicates of this asset on the scene, and that this specific piece is the left leg, ready to be exported as a Point Cache from an animation file.
I know this sounds like madness, and it could be of course if artits didn´t have tools that add/remove or automatize tags in the names of the objects, but if this is followed throughout the whole pipeline it takes care of many of the typical problems, and the possibilities are multiplied when you can create tools that automate processes very easily.
Then comes the Asset Management System. This came out two or three years ago, and each studio deviced its own system, including us here in Blur. It´s interesting to see how different and unconnected studios come up with the same ideas over and over, we are still far away from a DAM ( Digital Asset Management ) industry standard
It is also interesting to point out that the videogame industry stepped into this subject long ago, maybe this is related to the fact that they are more programming-oriented, they started with check-in/out systems and then evolved to DAM systems.
Web contenct publishing is also very advanced in this subject, there are many apps, even GPL, look for MAM (Media asset management), DAM, or CMS (Content Management System).
Looking at your blog I see you have some interesting links, mainly the article on Final Fantasy, a clear example of how easy it is to start a DAM and how hard it is to finish it.
Digital C.O.R.E started their own one for the film “The Wild” (not there as they called it in Spain) and when the film was over they still hadn´t finished, it turned out that in the end it was easier for them to redo it from the ground up with all the experience that they had acquired.
Another recent example is “Barnyard”, where Omation studio began with their own system simultaneously whith the production, and as you will imagine it was a little chaos. At the end of the production they had something decent that I guess they will be able to use if they are make some other CG film
As for bibliography, I can recommend these two books, specailly the first one is very useful as an introduction:
Implementing a Digital Asset Management System: For Animation, Computer Games, and Web Development
Focal Press (August 26, 2005)
The Game Asset Pipeline (Game Development Series)
Charles River Media; 1 edition (September 2004)
About commercial software, the best known is Alienbrain VFX, I can´t tell you how good or bad it is because I have never used it, but it has integration with XSI, MAX, Maya.
I would look for in the Animation Industry Database to see if you find something useful
Good luck with the thesis, and keep me informed, it is a subject in which very I am interested!