With a strange mix of bewilderment and confusion I’ve been following the MOX file format discussion ever since it started out on the After Effects Mailing List. Let me start off by saying that in fact it would be great to have something that matches what MXF and AAF once aspired to be in a form that actually works and solve all those problems with Gamma shifts, different marker and timecode handling, audio track assignments and whatnot. A replacement for the ageing Quicktime has long been overdue, too. So far, so good. But something doesn’t feel right. So far we only have quite sketchy information and well, to me none of it sounds convincing. After reading this interview, let me run you through some of my thoughts.
- The promise of only relying on OpenSource or patent-free formats is already prematurely limiting the scope of what could possibly achieved. The point is that if you use properly licensed commercial programs that support e.g. MPEG formats or use them natively in their inner workings, there would be nothing wrong putting them into a MOX container and it could still improve workflows by fixing what the others get wrong.
- Similar to the previous point, hardware devices such as cameras do not record to these free formats (not at all or not as default, at least), meaning that conversions have to happen. That by all means should occur automatically on ingest or import/export or otherwise people will just keep working with what they have and trust.
- OpenSource does not necessarily mean everything is free (forever) nor that it’s completely without any patent claims. You never know when someone might decide that a specific algorithm or implementation is not hurting another patent and files respective complaints or imposes licensing fees. Or someone could just revoke the use of his stuff completely and then everything collapses from top to bottom.
- Conversely, even if the MOX spec itself is published as OpenSource and developers are encouraged to let their experiences flow into it, that doesn’t mean it will actually happen. Companies could implement proprietary stuff all the time and just like today, other companies will have to reverse-engineer these parts and implement their own compatibility fixes. Some of the really good stuff may in fact never even become known to the public because someone uses it to have an edge over a competitor in an internal workflow.
- Just because something is OpenSource or a managed commercial standard doesn’t mean that people get implementations right. Every 3D artist knows how all over the place different OpenEXR, FBX, Alembic and Collada routines are in different programs and what trouble and headaches they can cause. Ergo, long before MOX can work its magic, it’s probably safe to say that software vendors will have to straighten out such basic stuff first.
- Even if the previous point were less of an issue, you still have to know how formats work, what their shortcomings and limitations are. You have to implement pertinent workarounds and safeguards to make up for that.
- As formats and programs evolve, so MOX will have to evolve. In this industry, nothing lasts forever and today’s “standard format” could be old news tomorrow. Different development levels will further contribute to fragmentation across versions and ironically at some point 10 years from now people may be discussing all the same frustrations about compatibility and performance issues, the software being bloated, the source code poorly documented and you know what…. SNAFU – situation normal, all fucked up.
- Especially with the two previous points the whole thing may actually be too ambitious and crumble under its own weight, which again isn’t that untypical for many such projects – they all start out with great, noble goals and in the end the horse becomes too big too ride when only a handful of developers work on it and the rest of the world doesn’t care and loses interest. You just need to visit the “graveyards” of dead projects on GitHub, SourceForge, Kickstarter and elsewhere to have that verified.
- Finally, and perhaps that’s what gets me most, it seems to me while the people involved seem to have an idea about the technical aspects, they don’t have worked out how to make it user friendly. There’s simply too much talking about specific technical aspects like CoDecs, compression rates, bit-depths and so on, but to tell you the truth, that, to me, is just the same Quicktime mess all over again. It’s 2014 and I really do not even want to care about what bit-depth or the finer points of the Pros and Cons of different OpenEXR compressions. Why should it even matter to me? If the format uses 32bit internally and has to inflate my 8bit stuff for transfer only to deflate it back when importing in product X then so be it. as long as the conversion is robust enough, and ultimately that should be the point, then there is nothing wrong with that.
As is obvious, in summary I’m a little more than skeptical. Perhaps it simply sounds too good to be true, perhaps seeing things come and go in almost 20 years in the media business has told me otherwise. Either way, even if it takes it will be quite a while before we will see the fruits of the whole endeavor…