Phil Rhodes takes an indepth first look at the new ARRI Amira
ARRI's stated grab for the documentary market can be interpreted in two ways: as simply that, or as a slightly indirect way of offering a lower-cost Alexa without risking any loss of value to its Alexa brand. There's no cynicism inherent to that second interpretation, especially in the light of Canon's C series cameras which, at the high end, compete with Alexa, and at the low end compete with Sony's F3. There's nothing wrong with a desire to have a finger in both those pies and well-differentiated products to do so, although I suspect we shouldn't hold our breaths for Amira to compete on price with the C300 (I'd have expected it to go for about twice as much).
The key to a reasonable evaluation, then, is your interpretation of the word “documentary”, something that has already caused significant buzz. It's a difficult term because it tends to mean different things in different markets, and even among different production sectors with respect to the sort of resources and personnel available to a production. Something like the BBC's Horizon series is a documentary, and consists mainly of sit-down interviews, lab tours, and a lot of computer graphics. Amira might be entirely usable on a show like this. On the other hand, there is some concern over how suitable it would be for the most remote, least supported, most run-and-gun style shows.
Why? Well, several characteristics hang over from Alexa which conspire to mean Amira will, in all practicality, require more support than a more conventional ENG camera, which is what Arri seem to be packaging it to compete with. This may be OK, or it may not, but it's worth considering.
Concerns about weight and power consumption
First, Amira is quite big and quite heavy, especially as displayed with the Fuji Cabrio 19-90 zoom, which made it very front heavy. Your narrator is 6'3” tall and doesn't consider himself an unusually feeble specimen as cameramen go, but wouldn't want to guarantee nice camerawork off the shoulder with this configuration for more than a few minutes, especially as the Cabrio makes it very front-heavy. An assistant to throw it at between takes would be a good idea. Further, Amira is quite power hungry. At 50W, power consumption is sufficient to require changes more than once an hour with battery systems that are easy to get onto commercial airliners. This is not unreasonable as it is apparently built around the same FPGA-based electronics as Alexa. This highlights the way that differences in technical implementation can directly inform the suitability of a camera for a particular style of production. While we would not expect ARRI to build lower-power custom silicon for a niche device like Alexa, and while the increased power consumption is, within that area, not a huge issue, take the same technology out of that niche and the implications of a given design approach may not be as easy to overlook.
A boon or a curse?
The only remaining concern, and one that's already been widely discussed, is that super35 sensor is in a documentary style environment. This is either a boon or a curse, depending on the style of production, and not something I can reasonably interpret further here. Suffice to say that ultimately we know this sensor to be excellent as Alexa's images clearly demonstrate, although anyone tempted to write off the problems of self-focussing sensors this big can easily go and try it on one of the many cheaper cameras which are currently available. ARRI consider the base ISO of Amira to be a speedy 800, which will assist somewhat in maintaining a sensible shooting stop – until there's a need to drop in some of that internal ND and go for a setup with short depth of field, where that option remains, of course, entirely open. It's worth bearing in mind that the Amira viewfinder is not full resolution, which may mean users prefer to carry extra monitoring.
There is much talk of the use of B4 mount lenses intended for 2/3” video cameras, which I would consider to be a much better idea than the perhaps slightly overbuilt Fuji 19-90 that was shown at IBC. It is already possible to put 2/3” lenses on PL cameras, given the little-known HDx35 optical relay built by Abel Cine Tech in the US. Abel built this device to service their specialisation in PL-mounted high speed Phantom cameras from Vision Research on sports broadcast jobs, but it has already been used to put B4 lenses on cameras such as the Panasonic AF100 and could presumably be used on an Amira. The presumably very good imaging performance of the Amira would encourage the very best B4 lenses to be used, but this is certainly a very welcome option regardless of how Arri choose to implement it.
Will the Amira replace the Alexa?
There are a few things – of course – that you get on Alexa but not on Amira. You don't get ARRI's Raw recording format (though the high bitrate ProRes is very good, and one could presumably use an external recorder to get it completely uncompressed). You do get 200 plus frames per second, which is not something Alexa does.
Whether ARRI like it or not, it's fairly obvious that Amira risks replacing Alexa on the lower end of jobs that might have previously stretched to afford it. There is very little reason to use Alexa in a world where Amira exists if your intention was always to shoot ProRes. And I think ARRI's claims for Amira being a run-and-gun, down-and-dirty documentary camera might be reaching just a little, although I have no doubt someone will find a reason to do it at some point. What I think is more likely is that soap opera and serial drama which currently shoots 2/3” formats will use it, as a low-impact way to squeeze a little more prettiness out of their incredibly constrained timetables. Ikegami clearly thought so too, hence the presence of the HDK-97ARRI studio camera at NAB back in April.