In doing another Fearless Forecast for the coming year, I re-read my previous forecast (posted exactly a year ago) just to see how I did. In a nutshell, not bad. Here are some things I predicted and how they actually turned out:
I predicted that Red’s influence would become stronger, and to some degree it did. I said they would ship some but not all of their new products “later in the year”, and I was correct, to a degree. They did deliver a number of pre-production Epics to some significant large productions in November and December, and they did ship the first of the hand built new cameras before the year was out. I predicted that Canon would unveil a competitor to Red’s line, and that did not happen, although ironically enough, some competition for Canon appeared in the form of Panasonic’s AF100 and to a lesser degree, Sony’s F3. I correctly predicted that Sony would talk about 4K but not ship a camera product in that format. I also correctly predicted that Arri would ship a new digital camera line and offer both compressed recording and uncompressed recording, which turned out to mean ProRes compression on SxS cards, with live dual link HD output via SDI. I predicted that software based post production tools would have downward price pressure, that they would be multiplatform and that lower end versions of existing high end tools would appear. I was pretty accurate on all of these things, not that it took genius to see them coming. But by the end of 2010, there were lower cost versions of Autodesk products, Avid DS, Baselight, and DaVinci Resolve, just as I had predicted. I predicted better Apple support for NVidia cards, and in particular support for the SDI daughtercards. I was wrong on that one, although companies like Autodesk and Blackmagic (via DaVinci) have successfully hacked around some of those limitations to allow successful migration of their programs to Apple’s platform.
Probably my most accurate prediction had to do with color correction. I said that one disruptive product would emerge that would force all of the others to take notice and adjust accordingly. I further said that the likely source of that product would be DaVinci, and I was right on the money with that one. I said that they would release a shrinkwrapped version of a software-only Resolve, and I was dead on with that one, too. I also predicted support for commodity control surfaces, in particular, the Tangent Wave, which was correct as well. The only thing I didn’t get quite right was the price point, which I predicted would be around $15K, in the same ballpark as Autodesk Smoke. I was taken by surprise – as was everyone else – when Blackmagic released the Resolve software at a fire sale price of $1000. I still think that’s considerably lower than it needed to be, and I still don’t know how they can make money on it at that price (or justify support), but it doesn’t matter what I think since they’re already gone down that road.
So that being said, let’s move to this years’ edition. As with the Grading the Graders series, I’m going to do this in two parts. First up will be issues and products dealing primarily with production and the front end of post production. In a second part, I’ll discuss industry trends and more specific post production issues. So, as I did last time, let’s start out again with Red.
2010 was in many ways a transitional year for Red, with much of what they were doing going on behind the scenes. Aside from Mysterium X upgrades for the existing Red One model, and many upgrades of their Redcine X software, they didn’t ship any significant new hardware product to the general public. They did show and/or talk about a lot of new things, but none proved production ready by the time the year was out. 2011 will likely be considerably different, with a lot of what was behind the scenes emerging as real product. Epic will appear in production form, followed a few months later by Scarlet. Red has reconfigured their lineup and rebranded what was originally the Scarlet 35mm camera as part of the Epic line with the name Epic S, with the Scarlet name reserved for what will, at first, be a 2/3 inch based unit. To this point, Epic has found its immediate acceptance in stereoscopic 3D production, which is a niche in which there is really nothing quite like it in terms of a combination of high quality imagery, extreme bang for the buck (including some very impressive overcranking capabilities), and perhaps most of all, form factor. Its small size is ideal for 3D rigs that have had to accommodate larger camera bodies, but can now be brought down to a much more manageable size and weight. Those using it on 3D projects will find a lot to like, and that will allow it to migrate to regular 2D projects as well. Its high resolution will make it a preferred digital platform for features, but it will likely find a more mixed audience in the television world, as will the Epic S (the rebranded Scarlet 35mm). Once it ships, Scarlet will likely become the preferred “companion” camera on Epic shows for things like additional coverage, shots that require a lot of mobility, and shots that place the camera body in danger, replacing the Canon 7D in a lot of those cases at only a slightly higher cost.
In the software end, Red will continue to improve its Redcine X program, which has already become the de facto transcoding engine for Red material. To this point, Red’s introduction of the Red Rocket card has served as a real catalyst for more mainstream acceptance of the format by eliminating a great deal of the time involved in creating files for editorial and the like. But like all hardware, the Rocket card is a proprietary and somewhat closed design, limiting its future adaptability. It was a necessary product for Red at the time of its introduction, and still serves it well, but ultimately, I think Red would be better off moving to the use of commodity GPU’s for a lot of the heavy lifting currently being done by the Rocket. This is the approach taken by Blackmagic and many others, and although a great deal of the processing burden for Redcode files is dealing with decompression, it has been shown that GPU acceleration can be applied to wavelet decompression with success. The simple fact is that commodity hardware has economies of scale that a specialty card like the Rocket can never have, and its continued development ensures a constant upgrade path. I would like to see Red embrace this and basically move to a software based, hardware accelerated transcoding engine that could ultimately take advantage of the Moore’s Law improvements that Jim often points out. This might take a new generation of CPU’s as well (which we will also see later this year), but inevitably I think Red needs to go to software so as not to limit their future growth. Red has shown prototypes of other products that will likely ship this year, one of the more interesting is the Red Ray player/server. I’m going to hold of discussion of Red Ray until part 2 of the forecast, however.
Arri introduced the Alexa in 2010, one of the most significant products it’s ever developed. It found almost immediate acceptance and praise, in part because it represented something that Red was not, in part because it displayed some very impressive imagery right out of the gate, in part because of its increased dynamic range compared to existing products (including Red), and perhaps most significantly, because it was, well, Arri. There are many in the industry who sincerely want Arri to succeed, for many reasons, some of which are simply sentimental, but many that are not. Arri has demonstrated a commitment to very solid build quality in its camera products, and many have come to trust that. They have also demonstrated a deep understanding of color science and how best to yield extremely pleasing images for both electronic display and film recording. So expectations were high, and arms were wide open. However, as successful as they’ve been with the Alexa, especially in television, it is clear that there are real limitations in this first version of the camera that put it a quite a bit behind Red’s roadmap, particularly for feature work. The great images coming out of it, along with the convenient ProRes recording capability, have been the primary selling points for television, but that only takes you so far when you’re looking at a market that is producing images to be shown on 40 foot screens. My feeling is that Arri knows this quite well, and will unveil a higher resolution “Alexa 2″ this year. That higher resolution could be 4K, or it could be higher. It could involve a Vista Vision sized imaging device or a 35mm based one. And it will likely involve an on-board recording solution that will involve some kind of data compression. Some might say this would be copying Red’s lead, and to a great degree that’s correct. But just as Kodak and Fuji have always competed in film stocks on the basis of the images they produce and not the scanning resolution they’re capable of, in a digital future the high end competitors will likely not be competing on who has the higher resolution or the better compression. They will compete on the basis of the images they produce, with post paths evolving that will be more generalized and able to handle images in a common way (I’ll talk about that a bit more in Part 2 as well). It might take a year or two to get to this point, but I think the migration will start this year with an Arri product introduction (or at least an announcement..) that will put the industry’s two primary file based, high end camera suppliers on more equal footing. And I think that’s a good thing. Once you let go of the tech, you can concentrate on the art. And that’s what all of these things are ultimately built for.
Sony would appear to be the loser in the area of digital cinema cameras, but most people in our end of the industry seem to overlook the fact that Sony is a giant in other camera markets – consumer, broadcast, and industrial. I feel that in the face of an industry that seems to moving more towards higher resolution capture and file based recording and post production, Sony might actually retreat somewhat from the digital cinema camera market. I wouldn’t be surprised to see the F35 be the last camera of its kind, with the larger chip technology migrating to far less expensive products intended more for the industrial and high end consumer markets. They’ve already done this with the F3, which I feel will ultimately undergo a price adjustment to put it more in line with the current EX3. And they’ve also done it with the replacement for the venerable F900, by announcing a modular upgrade for the new HDCam SR based camera that replaces it. Their expertise and high profile in the broadcast market allows these products to be practical, but my prediction is that in the television market, 2011 will be to videotape as 2009 and 2010 were to film – namely, the beginning of the end. It is clear that file based recording is the future, something Sony already dabbles in with the XDCam and XDCam EX products. They have announced a file based implementation of the HDCam SR codec as well, but I feel this is a bit too little too late, as the SR format just doesn’t fit in a higher resolution world. Sony will also face competition in the 4K exhibition market for the first time this year as well, with the introduction of 4K DLP projectors. So my fearless forecast for Sony in 2011 is a return to its roots in broadcast and consumer, and a retreat from its digital cinema efforts. That may be a pretty bold prediction, or it may be a bit premature, but let’s check back a year from now and see if it pans out. The wild card here would be if Sony were to acquire Imax, a rumor that has recently started making the rounds. If that happens, Sony’s priorities could quickly change.
Last up for this part of the forecast is another wild card, Canon. Nobody really seems to know if they are serious about a digital cinema presence or not. They took a back door into the industry by creating a product for a specific purpose when they put HD video capture on the 5D Mk II for the Associated Press, only to see the device adopted by a much wider audience of indie filmmakers looking for an ultra cheap camera that could deliver decent images using 35mm optics and yielding shallow depth of field. Since Canon, like Sony, is a large company that basically requires large potential volume in order to justify production of a product, it is hard to say whether they feel that kind of volume is possible with a targeted digital cinema product that would evolve from the still camera technology, especially at anywhere near the price points they can hit with their digital SLR’s. My prediction is that they will show something beyond the design prototype they’ve shown already, that will likely be targeted at the potential market for Red’s Scarlet camera – that is, high end consumer, low end digital cinema, and special use on productions using things like Alexa and Epic as their “A” camera. The user base they have with the 5D and 7D will be very open to such a product, particularly as it would allow them to continue use of their current lenses, and even more particularly if it eliminated most of the current compromises on image quality by going to a higher clock speed for the imager and a far less limited recording codec. So I hope they do go down that path, but I’m not definitively predicting that they will. As I said, they’re a wild card. At least as of January 1st, 2011.
That’s it for part 1. Within the next few days, I’ll post part 2, in which we’ll look at general industry trends and where they might take us this year.