Ok, one last post to this topic ;) The point I wanted to mention in my last posts wasn't 'the one software solution' is best but the 'major application solution' is a lot easier to handle than 4 or 5 packages. You never will get all the tools you need in one software but if you get most of the tools in one package and only use secondary tools for special tasks I think that's a better way than to split every step of a production into a special software tool.
A simple case for example: You model your character in silo, set up the UV coords in UVmapper, animate it in messiah and set up the materials and render it in Lightwave. Now the client told you that he wants other shoes on the character. You again model in Silo, create the UVs in UVmapper and so on. I think if it is possible it's easier to do all this changes in Lightwave instead of import and export 4 times, reconnect properties, shader, lights and so on. Surely, if you need a special tool in silo to model this shoes it's no question, do it there. But often you can do it in another software also, maybe not so easy but faster than 4 import/export steps including all the additional actions after importing the new object. It all depends on the purpose. In an ideal linear production this changes won't happen. The modeling is finished before animation and rendering. But a perfect linear production is really rare. More often you work on different steps of the production at the same time and then the import/export steps (including all the addtional steps after import) becomes a real nightmare. Two or three years ago I worked as a modeler on a production where XSI was the main application. At this time I was much faster in modeling in Realsoft and Lightwave than in XSI. So I started modeling in my favorit applications and export the objects to XSI. During the production I recognize that the export and import steps (and sorting the objects into the scenes, applying basic materials, deleting unused clusters and so on) slow down my workflow so much that I tried to model more and more in XSI instead of the other applications, in particular when there were changes to objects. Although the workflow was new to me, and I had to search for the tools I need for modeling I did the last objects in XSI only. Why? I simply more and more learned the 'new' tools and afterwards was definitely faster in modeling in XSI than in modeling in LW, exporting and resorting. One thing you are defintily right is that a software should concentrate development on their strength and not on fancy new features. But the really important thing behind this is that the software should be production stable. The workflow have to be production stable. A tool with a good and fast feedback is mostly more worth than some tricky features. Take a look on other packages but don't let you blind by some special features or nice example images (you know it's mostly the artist and not the software that makes the image). Best example in near past was the hype around maxwell. Yes, it has some interesting concepts and nice looking features and example images, but it's totally useless in production in my opinion. Or, to be more concrete, for productions like I do. I can't wait for a final rendering in screenresolution(!) 10 hours to be noise free. What is if the client has some changes? And I'm working in printresolution in most cases. Even the preview feedback when setting up materials and light is to slow for my workflow. Sure, it's nice to have such features like 'physical correct' lights and materials and a workflow like photography. But don't forget it's not the realism of an image but more often the mood that counts. Realism is a subjective sense in many cases. One last note: I think to learn a tool in most of it's aspects is totally important. You have to know the strength and the weekness of the software, the possible bugs and workarounds if you want to do a production with it. And so I think it's better to learn your tool in depth before switching to a new software because it has the button you search for. If you realize the other tool is better for your purpose, no question, change to this tool, but first take a closer look if it is really not possible with the tools you already use. I personaly think to have a major or key application in your productionpipeline is great. This key application may change from job to job, depending on the tasks, but it's a center in the production where the other tools are some kind of satelites. And this major tool surely can't be a modeler or renderer only. Best Tim
