A Good Art Pipe *essay 1st draft complete*
polycounter lvl 18
*EDIT*
Latest version posted here
Post comments below
Thesis
There are 3 important variables that need to be addressed when creating an art pipe:
-The artists in the pipeline
-The existing tool chain and it's engineers.
-The project timeline and the managers/producers who schedule it.
Here's the conclusion of my essay:
[ QUOTE ]
He should survey the design doc to understand the scope of the project.
Then learn the artists' development style with them in the trenches and learn to think as they do.
The next step should be a synch with the tools dept and to understand what they have to work with, how they have made changes in the past and then start the first steps towards designing the most efficient, clean and simple art pipe with supporting documentation/tutorials.
The last step would be to develop a system to monitor the pipe during pre-production to work out kinks in the system before any scheduled art asset production for the game begins.
Once a all the necessary test units has been completed exploring all the systems (respecting the schedule, workstyle of the art staff, and forseeable scope of beta tools) then productiion should begin.
Once in production the technical artist should be the go-to guy for questions and monitor the art assets, trouble shooting and maintaining the building conventions.
[/ QUOTE ]
Tell me what you guys think. I want this to be as good as it can be.
-R
Latest version posted here
Post comments below
Thesis
There are 3 important variables that need to be addressed when creating an art pipe:
-The artists in the pipeline
-The existing tool chain and it's engineers.
-The project timeline and the managers/producers who schedule it.
Here's the conclusion of my essay:
[ QUOTE ]
He should survey the design doc to understand the scope of the project.
Then learn the artists' development style with them in the trenches and learn to think as they do.
The next step should be a synch with the tools dept and to understand what they have to work with, how they have made changes in the past and then start the first steps towards designing the most efficient, clean and simple art pipe with supporting documentation/tutorials.
The last step would be to develop a system to monitor the pipe during pre-production to work out kinks in the system before any scheduled art asset production for the game begins.
Once a all the necessary test units has been completed exploring all the systems (respecting the schedule, workstyle of the art staff, and forseeable scope of beta tools) then productiion should begin.
Once in production the technical artist should be the go-to guy for questions and monitor the art assets, trouble shooting and maintaining the building conventions.
[/ QUOTE ]
Tell me what you guys think. I want this to be as good as it can be.
-R
Replies
i guess that's quite common but can be a rather expensive mistake in the long run.
It's pretty safe to say that apps like Max, Maya and even XSI are more than capable of keeping up with next-gen needs and therefore make for a solid investment. Newer apps like Zbrush focus more on complimenting existing art pipelines rather than replacing them so these would be of secondary consideration.
Future enhancements to the main package may actually reflect on its importance in the process even further. I'll use XSI as an example... writing a dotXSI exporter for characters gives us the foundation for getting them into the game right now. However, when version 5 ships, having access to tools like GATOR will streamline a lot of technical and time consuming aspects of character creation, such as UV mapping, vertex weighting, possibly morph data, etc.
Looks pretty good so far PaK. I'd love to see the rest of it. =]
The only thing I really feel is missing - and may in fact be buried somewhere in another part of the essay between Artists and Tool Chain, and just assumed to be covered here - would be the importance of using the last pipeline or previous project as a spring board for new ideas, based on what the artists felt worked, didn't work, or could be done more efficiently.
You do mention it in a roundabout way, but I wonder if it could be made to stand out more...
I don't have access to UE3 tech but from what I've seen of screenshots and descriptions, it appears that Epic has really started applying past experience and artist feedback towards their latest tools to greater effect than I've seen previously.
...the importance of using the last pipeline or previous project as a spring board for new ideas, based on what the artists felt worked, didn't work, or could be done more efficiently.
[/ QUOTE ]
I thought I highlighted that idea in my second variable,but the fact that you mentioned it means I have to make it punch harder. This is exactly what I was hoping to achieve from this thread.
Keep 'em comming. Once I have a few more ideas to address I will rewrite my essay and the publish it on my website or something.
Thx...keep the crits flowing plz.
-R
While you could expand on any of these points easily, your tight editing says a lot. I'd like to see when you put the final together.
Artists needs to estimate their time in a realistic manner, which is sometimes very difficult for them to do.
We also have been using a spec sheet that lists what the asset is, where in the production tree it should go, how many polys, textures, materials, etc it should have on it. This helps immensely and reduces the amount of questioning at the start.
Also, if some part of the process has failed there should be some sort of post-mortem on it. This is extremely helpful.
Exactly what do you mean by that, and by an artists' workstyle, in detail?
Perhaps I'm misunderstanding it but it sounded sort of like the AD should be subservient to the artists' personalities, when it should be the other way around to some extent. Ideally I see them as cogs driving the machine, rather than the machine trying to keep up with the cogs.
If I'm understanding it correctly then I think you're leaning in this direction as well, but thought you might be able to benefit from the thoughts I've had on the subject.
Based on my experience and studying I've done on systems, process and even franchised businesses, my top priority in developing a pipeline would be maximizing efficiency and minimizing variability by establishing a clear, ordered series of steps for each task. This is the machine.
In practice I see this as the AD (or whoever is developing the pipeline) quantifying each individual step of the asset development process from start to finish and documenting every concrete bit of it down to the lowest acceptable level of sophistication. This is so practically anyone with, say, Maya experience can walk right up, read the doc and start generating fully functional content the right way, the first time, every time.
This way you can minimize the pain of adding and removing artists from a project and let them have a process bible to work from and cut down ramp-up time. For an art pipeline, this strikes me as an way to spread out your risk in such a way that if you lost one, two, or ten members of your team, the pipeline will survive because all your processes are so well-defined, even below average artists could follow the directions and fill the gaps adequately. Seems smarter than the pipeline's continued functioning relying on top shelf talent, which is rare.
The machine can quickly change out cogs without breaking, if it needs to.
The point of creating all this is so that the artists use this as an absolute baseline for everything they do, and work within these guidelines. From there the AD should absolutely know his artists, their strengths and weaknesses and unleash their ability and passion on tasks, but keep them working within these constraints.
So from within these constraints, individual artists' personalities work within the boundaries provided them, work on improving the processes, and streamline the pipe over time. If individual artists need replacing, ramp-up time is minimized.
That's just addressing the steps up to the system to monitor the pipeline. Overall, sounds pretty good overall, but could use a little more meaty detailing, I think.
Still, I suspect I'm just expanding upon what you already had.
http://www.easel3d.net/docs/TA_pipe_v1.html
Post comments below plz. I wanna rewrite this for a final draft Thursday night.
-R
This thing is done. 6 revisions *phew*
-R
The TA has to assess the type of artists in the pipe and accommodate them. Too many efficiency procedures for non-technical / individualistic developers stifle creativity. The strive towards efficiency on the TA’s part can go too far. <snip> turning your art staff into button pushing, procedure-driven drones will sap your game of creativity.
[/ QUOTE ]
The trick to that is to define where creativity occurs. Concepting a character is creativity-intensive, and constructing an efficiently deforming joint should be a clearly defined, reproducible process. This also deals with setting expectations with your artists and letting them know at what points in the pipeline that you depend on their creativity.
Good stuff. A few typos here and there still, but overall, not bad at all.
Jon:
You ahve pointed out a very real flaw in my pipe doc, as it doesn't make room for the concepting phase pipes. To be honest I have never particiapted in the concept phase as such. yes I have been a 'part' of concept but not integral enough to say I understand howto make the systems work better. I have ideas about how I see concept, and I certainly have good input for pre=production (which io covered in my doc)
I would love to discuss cocept pipes with you guys...I wonder if it has to be as robust as this. IDEO has great concept generations and nurturing ideas, and I will share them as soon as I fully understand them. My roommate works for IDEO and I am learning so much about it.
-R
I know it's a bit late, but you might want to check it out. He was very thorough.