Having established the motivations for fair payment on a "commercial free culture project" in the previous column, I'm still left with the question of what exactly "fair" means. The problem is that there's more than one way to determine fair shares on a project like this. The organization is necessarily loose, and so there's no really clear and unambiguous way to determine fairness. Nevertheless, some plan has to be chosen, and in a way that is at least defensible.
Making Movies with Free Software
This article is part of an on-going series on the challenges I've faced in producing two free-licensed movies, Marya Morevna, through the Morevna Project and Lunatics, which we are working on as Anansi Spaceworks.
One of the things that has bugged me from the outset is this business of remunerating collaborators from the earnings of the work. How do you decide, well in advance of production, just how much somebody should get out of any profits or revenue that are made? And how can you attract good professional talent if you put off that decision?
When most of the income is expected to come from ancillaries, is it fair to break that down by the same ratio for every product? Does it make sense that animators should be paid out of the income from selling a soundtrack? (Or musicians from a poster?).
Even with the knowledge that fans are going to keep us honest, there clearly are still the possibilities of differences in good faith between the artists, fans, and producer on what constitutes the "fair" division of earnings
Even with the knowledge that fans are going to keep us honest, there clearly are still the possibilities of differences in good faith between the artists, fans, and producer on what constitutes the "fair" division of earnings. In a commercial proprietary project, these decisions are all made in advance: an offer is made, and the artist accepts it by signing a contract. This provides a kind of security in the relationship that both sides can understand and stick to.
The bad side about "gift culture" is that it lacks the kinds of certainty and clarity that you get with this market economy. People can easily get their feelings hurt, especially their pride. This is a problem even on purely amateur projects, but if you add even the possibility of money in the project, you can create a real emotional powder keg, which if it goes off can blow apart your whole project. So we clearly need to think about ways to recapture some of this security by spelling out expectations in advance (before there's any money to consider and before contributions are made).
This way, everyone will feel like they know what they're getting into, and what they can expect. If the rates seem unfair, then they can opt out at the beginning. If, on the other hand, they seem reasonable, it will be harder for them to complain later if and when the virtual money becomes real money.
This is why I feel it's important, as a producer, to consider this problem now, before bringing anyone else in on the project.
Why not just split it evenly?
Perhaps the most obvious question, is why not just share it equally among all contributing artists? It sounds great. Utopian, even. But there is a problem. Who exactly is a "contributing artist"?
You can set this up artificially by collecting a group of people together, declaring them in advance to be your team, to be paid equally, and to contribute equally (or as equally as possible). If they regard the project as full-time employment (because, for example, they're being paid a stipend to cover living expenses, and they have agreed not to do other work during this time), then this approach probably makes sense. And of course, it's a pretty good description of how the Blender Open Movies have been made.
Figure 1: an "equal" distribution implies a closed group of people. Nice and simple... but anyone outside the circle is just left out in the cold
Alternatively, if you are making a live-action film, you may have a situation of very limited group of friends working together -- a "director", a few "actors", and maybe a "set dresser" or a "musician." Again, it's a small, closed group with no outsiders involved and no fuzzy boundary.
Either way, you can see some of the social dynamics:
- Each team member is under social pressure to "pull their weight" -- or seem like a free-loader
- There's pressure to keep the group very small so that profits aren't thinned too much by sharing them out
- If outside help is called for, the group has no way to pay for it -- it can only be free contributions
- If a team member has to quit or a new one joins late in the production, some policy will have to be invented for how to properly pro-rate their contribution
Anyone who can't work full time will either be shut out, or policy will have to be conceived to pro-rate their share based on some half-mythical conception of "time-fraction" they assign to the projects
It sounds fine when you say it, but the details get kind of nasty. For me, the worst part is the idea of a sharply-defined line between who is "in" and who is "out" of the production team. You can see how this looks in Figure 1.
This is very unnatural for any project that grows out of the culture of open source collaboration common to most free software projects.
Fuzzifying the perimeter
Adding a high wall around the development team means having to provide extra motivation to get people to climb it, and that's not always what you want.
The loose associational property of open development groups is part of their charm and part of their success. Since there's less performance pressure, people are more inclined to take a chance on helping out -- after all, if it doesn't work out, the consequences are minimal.
Adding a high wall around the development team means having to provide extra motivation to get people to climb it, and that's not always what you want
I also expect to do a lot of re-use: taking existing free culture works ("gifts" to the "commons") and adapting them to my project (examples include image textures and much of the music). In those cases, a lot of work was done, but it wasn't done for my project, it's just out there, able to be used.
Clearly, discounting the value of such gifts, simply because it is legally possible to do so, is not "fair."
On the other hand, it's clear that when I approach an artist and ask them to create something specifically for my production, that is a different level of involvement, and I should be able to offer more compensation for that.
Figure 2: For many open projects, a better model would be to try to gauge relative contributions. This way, the sharing can be more inclusive, representing the broader development model, while still recognizing the large contributions of the core team. The tradeoff is complexity
The idea here is that the sphere of contributors should be fuzzy, with a wide periphery of people who have contributed in some way, but are not core contributors. This situation is illustrated in Figure 2.
So it's probably safer to -- somehow -- judge relative contributions based on the work accomplished rather than on the perceived level of involvement. This way, the actual division of profits is decided after the fact -- but it must be predictable.
What we need to agree on in advance, then, are concrete rules for how we will do that evaluation. Thus, it should be possible, if an artist wants, to calculate what their contribution would be worth under some assumed set of circumstances.
For Lunatics, in particular, I'm concerned about loosely-connected contributors like musicians. I want to compensate the musicians who've been gracious enough to release their tracks under CC By-SA licenses so we can use them in our production (in most cases, this was not an exchange, but a gift -- they simply released their tracks, and I found them). As such, they saved me a lot of trouble and risk. These are tracks I know in advance that I can use, so I don't have to take into account the risk of them not being available or being too expensive. And I don't have to take a risk on whether I'm going to use them by paying for rights in advance of using them.
I want to compensate the musicians who've been gracious enough to release their tracks under CC By-SA licenses so we can use them in our production
By contrast, I don't like the idea of rewarding more favorably those musicians who chose not to release their work under a free license until I negotiate with them. These artists make my life harder, because each one of them has the power to hold up my project by withholding the license. So those are much riskier.
I think it's a very bad precedent to set, and would encourage withholding if it became a common practice. Whereas, if everyone is paid fairly, the findability and low-risk advantages of a free-license (I always search those first) will encourage their use.
These are artists I have not even contacted yet. Most likely, my interaction with them will be very minimal. Yet, in some cases, they may be very important to my production. A particular scene may only work with a particular piece of music. Indeed, sometimes I may rely heavily on the music, so that it seems that all I'm doing is making a "music video" to something that already exists.
Other times, of course, the music is incidental to the overall effect (barely-heard ambiance coloring the dialog in a character scene, perhaps). This music is usually easily replaceable, and so there's less risk in using it.
After a bit of thinking, I decided that a simple and easy-to-calculate way to compute the relative value of music in the production was this:
First, I decided there were basically four different "rates" for music in my production:
- "Main Title"/"Theme Song"
- "End Title"
- "Hero" (music is the centerpiece of the scene)
- Basic ambient music
In addition, there is "commissioned" music and merely "licensed" music, distinguished by whether or not it was written specifically for the production.
There are some pieces of music (notably the Main Title for a series) that will be used over and over again in the production process, while others may appear only once.
There's also length -- the shortest original track on my current pilot's soundtrack list is about 25 seconds long, while the longest is 18 minutes. Of course, the final-mix tracks are usually shorter, ranging from about 5 seconds (an "eyecatch") to about 5 minutes.
Taking all of that into account, I came up with a formula for each use of a portion of each track:
R × C × (1 + t/60)
Where: "R" is the basic "rate", "C" is 2.0 for commissioned work and 1.0 for licensed, and "t" is the duration of the track in the final mix, measured to the nearest second. So basically, the track gets a certain amount for each usage, plus an amount based on how much of the track was used, and the rate is based on how the track is used, and whether or not it was commissioned.
This is very concrete, and in fact, I can simply create a spreadsheet using it, which would be available to anyone contributing to the project, as well as to potential donors, providing complete transparency.
This is very concrete, and in fact, I can simply create a spreadsheet using it, which would be available to anyone contributing to the project, as well as to potential donors, providing complete transparency
And, though the formula doesn't capture it, I expect to pay some "advance" on commissions, based on a conservative estimate of the amount the project can be expected to make -- essentially assuming some risk on myself rather than leaving it on the artist. For licensed works, that risk is a "sunk cost" since they've already done the work.
Of course, for each production block, you just add up all the "shares", divide that from each share, and you have the correct percentage to pay for each use of each track. Add those together for each artist, and that's what you need to pay.
This is only one piece of the puzzle, of course. I'll have to come up with something that makes sense for each production block. And then, I'll have to deal with how to compare the value of different parts of the production to each other. That's what I'll address in the next part of this series.
But I think the music example does show that there is hope for a predictable system that can create a certain clarity in the process for commercial free culture projects. If so, it'll be a really good step towards making larger free culture projects possible.
This work may be distributed under the terms of the Creative Commons Attribution-ShareAlike License, version 3.0, with attribution to "Terry Hancock, first published in Free Software Magazine". Illustrations and modifications to illustrations are under the same license and attribution, except as noted in their captions (all images in this article are CC By-SA 3.0 compatible).