Now let us turn to some nuts and bolts: the CG Supervisor's pipeline design work-flow. (Bear in mind this work could be lead by a studio VFX Supervisor or CG-VFX Department Manager.) In doing so, my intent is to describe my process in a very generic way. This is a process that can work if the production pipeline is for a new company or department and equally well if it is a new project or an expansion.
But first, let's look at the seminal published work of
Dane Edward Bettis, who describes a "three layer" process for designing and implementing a CG pipeline. Generally, Bettis has a good basic starting point, and without risking plagiarism, let me refer readers to Mr. Bettis's paper for a different perspective. Bettis begins with the assumption that there is an existing CG department that is re-designing the pipeline for a new project. He is also working within a specific context of CGI animated feature films. I'll summarize the main heads:
The Three Layer Pipeline Design Example
- Personnel Arrangement
(Division and Assignment of Manpower and Task Domain)
- Implementation and Managing Complexity
- Optimization of Computer Systems
Bettis, Dane Edward "Digital production pipelines: examining structures and methods in the computer effects industry", Texas A&M University, http://txspace.tamu.edu/handle/1969.1/2406?show=full, 2005;
The VFX pipeline structural form is dictated by :
the functional mission
resources available
company culture
the functional mission
First, we look at how the pipeline needs to support the overarching service mission of the company or department. This mission will be in the form of services to be offered: is the mission CG animation, motion graphics, digital effects or some other visual effects service? The more narrowly defined this mission in the designers' minds the better. For example, is the mission to provide digital effects for any producer, or is the focus on film, HDTV, DVD, internet distribution, or the latest gadget from company A.
The company's competitive strategy, which some may call pricing or marketing position is another important part of the functional mission. Is the company competing on the basis of price, service (turnaround, client relationships, volume/capacity), the work (quality, unique technologies, skills or methods) or some combination? Each of these factors may bear on how all three classes of pipeline are to be structured. For example, if the strategy is to provide a certain fixed capacity of shots with widely varying direction to a client on a recurring basis (for an episodic TV show), the pipeline will be structured differently from a pipeline designed around a fixed order of very specific shots.
The third aspect of the functional mission to determine is the actual project scope. Whereas service mission goes to intended services and competitive strategy may address general capabilities, project scope attempts to define with as much specificity as possible the parameters of the project: services needed, level of difficulty, volume of work, schedule and budget are key factors. Additionally, client interactivity and delivery needs are part of the scope. For example, if your client is hooting 12 timezones away during your schedule, this will impact your approval pipeline and that may impact how your production and material pipelines function. The 12-hour offset will tend to add a minimum of 24 hours to get feedback on each decision, which puts more pressure on your production pipeline. It also means that work on shots will proceed in spurts --so workers need to have several projects with offset schedules to reduce standby slack. Likewise, the asset and materials pipeline will have additional demands to generate approval materials and deliver them to the client. Project scope includes every possible parameter related to fulfillment of the service mission.
Use a hypothetical template project when designing without a known project scope. Suppose a client walks in the door and needs a 100 water and fire effects shots in six weeks and another walks in needing a complete branding campaign with a CG character and motion graphics in three weeks. Each of these projects have a definable functional mission. But in order to meet those deadlines your department must be pre-engineered to meet the demands. With a pre-engineered pipeline, when the actual clients walk in the door, the task becomes one of tailor-fitting rather than building from scratch. Use a template project, (or projects) together with the service mission and corporate competitive strategy to design a pipeline with general capacities.
Image by ToniVC via Flickr
of the essence,
so your company or department may need to be prepared for what walks in the door. Film studios and producers are squeezing visual effects vendors with tighter schedules and more last minute schedules to reduce time-to-theater in post. Television and advertising producers have always been tight for time. Prepare.
With functional mission defined, the pipeline designers are likely to proceed to resource assessment. Before we do, let's consider a moment some issues of company culture that may influence our decisions.
Functional Mission Considerations
- Service Mission
- Services and Products Offered
- Market sectors served
- Competitive Strategies
- price
- service
- turnaround
- client relationships
- volume or capacity
- workmanship
- quality
- unique (or rare) technologies, skills or methods
- Project Scope
- services needed
- level of difficulty
- volume of work
- schedule
- budget
- client interactivity
- delivery needs
2 comments:
I disagree with the 24-hour turn around time.
We work with O.S. studios both in and out of the timezone, on +0 to +14 hours, it is no problem on approvals.
A) They work on things approved the day before, fix, send back, we approve while they sleep, send back, ad nauseum.
B) Datastore updates happen on the wire, in real time (at least these days), so there is no down time in getting and sending updates/approvals.
Actually, I think we can agree. The point of my article is to call attention to issues that need to be considered, not to suggest certain problems are unavoidable.
Your point A) addresses what is called "approval lag time". You are correct that it should not be a problem, and it is not a problem because you've designed your pipeline to work with it. In your case it works because everyone involved understands that they have to respond by the end of their shift. This falls in the pipeline dimension of work Methods (policies, procedures).
As you state, lag time can be avoided if the work needing approval is actually approved or notes given by the start of the next work shift. Such a pipeline can be designed, I have done it.
Actually, a full 12 hour work-day offset probably is least likely to cause lag, precisely because the two shifts have no overlap. But it can cause a lag of a day if change instructions need to be clarified or there are other questions.
Lag can be worse in the same time zone because in this case, the work is sent for approval sometimes after the reviewer has ended the day or too late to be reviewed. The next day the reviewer may not get to it for some time (suppose the reviewer has a busy schedule and 40 shots to review).
In my example, I am speaking more of the client who is NOT a VFX house outsourcing to another VFX vendor. I am speaking of the client who is a director or vfx supervisor on set. In this case the lag could be exacerbated by other factors.
At any rate, lag time, for whatever reason, needs to be considered and designed around. One solution is to work hard to eliminate it.
B) I said nothing about DELIVERY time because a well-engineered pipeline will send materials for approval on the wire. However, as I said, this does mean some extra handling of the materials because they will need to be conformed to some standard and uploaded.
It may seem trivial, but the pipeline designer needs to consider it. Who is going to do the work, and how does it get done?
Thanks for getting me to clarify my thoughts.
Best success!
Post a Comment