The Foundry released their first public beta of STORM earlier this week and the initial feedback from the community has been positive. In the first 24 hours it was available, over 2000 users registered the software…that’s actual registrations, not downloads. The software will be free for use until the 1st of March, at which point it will be available for £250 / $375 / €300.
The news about the release came out several days ago and has, from what we’ve seen around the web, been well-received. We were in snowy London at The Foundry’s offices the day after its release and the offices were buzzing with the overwhelming response to the product. We sat down with Richard Shackleton and got an update on where STORM was at, the beta process, and some new insight into what The Foundry has next in store for the technology behind STORM.
The software was originally targeted for an early November release, but they found this date to be a bit agressive for the functionality and stability they wanted to provide. There was some hope that they might be able to provide more post workflow paths than Final Cut Pro for the initial release, but decided to concentrate on making sure the workflow was solid on a single path instead of shipping something that was only somewhat working for multiple paths (such as Avid).
When The Foundry announced STORM publicly earlier this year, they saw a need for an R3D workflow tool in the market. What they were surprised by is the broad interest in the software throughout the community. “From the DITs to data wranglers through the camera operator to DP through to supervisors on set. And then even through to post production and people in data labs,” people saw interest in STORM says The Foundry’s Richard Shackleton. “It is a much broader range of people all wanting to solve problems related to digital camera workflow. I think where we’re seeing some success is by building one tool that’s accessible by them all — even though the cinematographer and data wrangler might want to be doing something different — having one platform to do that has had a very positive response.”
That being said, one problem with broad acceptance is that software can become too generalist and not be as effective for the target, becoming a jack of all trades and master of none. The Foundry’s developers have been very cognizant of this and for now are focusing on the on-set/near-set data workflow for getting material into post production and data delivery while maintaining metadata.
What should one’s expectations be regarding the beta? It’s the first time The Foundry has done such a wide range public beta, but they wanted to expose the software to the broad range of users expected to find it useful and see what problems they ran into. “We have a proper QA team behind STORM here at The Foundry and we’re doing a very robust QA process but we can only cover so much ground. And I think we’re kind of where we expected to be. The first reaction has been for some people it hasn’t run, some people had media rendering problems, and for many people it’s been working great.” There are a great deal of variables, from notebooks to desktops, systems with and without Red Rocket cards, and various RED camera software versions being used on set. By widening their testing base The Foundry hopes to be able to create an official shipping version of software which is more robust than a more limited beta.
Some software beta releases have what’s called “feature lock” which means operational features won’t change and software developers are only looking for bugs in the software. While obviously interested in fixing bugs, according to Shackleton they’ll “also be listening very closely on the feedback that comes in, especially from people who are are well versed and experienced in the workflows STORM is trying to address.” They’re open to making changes to the software should users find it necessary. The Foundry will be actively monitoring and participating in the RedUser STORM workflow forum as the main information source for feedback and new releases.
In STORM, The Foundry has tried to build a tool that will be useful on both laptops and desktop machines. The application uses software decoding and playback, so faster processors will have an impact on responsiveness — it is not dependent upon the GPU. One particular area which has helped performance is the use of memory caching, which is something new added since showing the product at IBC. Even on a slower laptop, if it is loaded up with RAM, STORM will use this effectively to cache decoded imagery and provide realtime playback. Current generation MacBook Pros can take up to 8GB in RAM, so a $150 upgrade to max out the RAM would be a great investment for on-set work. On the desktop side, The Foundry focused on making sure that the Red Rocket card worked efficiently as well as driving decodes through the SDI output. “If you’re staying inside the RED color toolset, then having a Red Rocket card will be a big advantage,” says Shackleton. Additional creative tools from The Foundry would not be accelerated by the card, since the card is intended to speed up native RED decoding.
Now that beta is out and plans to ship with support for R3D fotage and Final Cut Pro conforms, what does the future hold? The Foundry has previously shown a multi-track workflow which they turned off for this release in order to concentrate on the core feature set and robustness. That’s planned to reappear again next year. In addition to this, supporting post workflows other than Final Cut is high on the list for next year’s planned 1.1 and 1.2 releases. “We’ve had a lot of feedback and Avid is ranked high on peoples’ requests. We’re also listening to other exchange formats….CDL has come up quite a bit for exchanging look data with other systems.” The Avid request isn’t all that surprising, since with the recent feature-filled Media Composer releases there seems to have been an increased interest in the software. I’d expect to see support for both sometime next year (my words, not theirs).
STORM will also be expanding the various camera formats that we’re being asked to work with in the production and post production community. They’ve obviously had the complete spectrum of cameras mentioned to them, but the two primary requests at this point in time are Arri and 5D support. As an example, in the case of the 5D H.264 footage, the lack of native timecode has been a hitch in the workflow. Even though there are tools out there such as Magic Bullet Grinder which can add timecode to ProRes clips created from 5D originals, it’s shockingly commonplace to have offline edits reach the conform or grading stage without timecode. The Foundry is looking at addingtimecode metadata to the clips *before* offline so that once the offline is completed, timecode can be used to match back for conforms along with all the other metadata one might add using STORM.
As far as adding of creative features, The Foundry wants to to make sure that STORM maintains its focus as being an on-set or near-set workflow tool. “We’ve spoken about STORM X in the past, which has been a working name (we don’t know what we’ll call it) for what could be a big brother to STORM and we’re just finalizing those plans at the moment,” says Shackleton. “There will be a bigger version of STORM built on the same technology platform that starts to introduce some of the tools that are more applicable to post production and the short form project finishing workflow.” This includes things such as file-based playback of DPX files and “of course, the ability for STORM and Nuke to work together in an efficient way.”
The Foundry are seeing many facilities working on short form projects or commercials adopting Nuke. The software is used in a variety of ways…from an assistant supporting a flame or smoke suite with roto and fixes to comping final scenes for inclusion in masters. “I think the thing that is missing from Nuke is the conform from the offline and timeline viewpoint. How does the client see the shot in context? How do they make adjustments to positioning or color? Or doing a quick mockup of a key to sign off on the intent while the actual visual effects will be done slightly offline,” Shackleton relates.
A tandem of Nuke’s shot-based effects approach and STORM’s timeline editing, clip management, GPU grading pipeline, and formatting/delivery is sure to make artists salivate as to the possibilities. Eyeon’s Generation and Autodesk’s smoke/flame timeline + batch, while quite different in approaches, do provide some similarities in workflow. Each of these applications have an integrated editing timeline made up of a series of shots, with easy access to effects compositing scene-based tasks. Generation has solid collaboration and versioning tools in a multi-seat workflow and Autodesk’s smoke/flame combo set the standard in the industry for fast and efficient creative tools and finishing in interactive sessions.
Nuke’s roadmap includes new features which will help this duo. Nuke 6.2, which releases soon, has a more of a dope-sheet style interface to allow artists to slide and line up clips within Nuke in a more visual way. Version 6.3, due next year, has a new particle system and a reworked spline/grid warper as well as a focus on efficiency of caching and playback. These are “all tools that are heading towards a more interactive experience in Nuke, allowing it to start to play more heavily in the short-form/shorter-time-turnaround type of workflows,” says Shackleton. “Having a really close relationship as to how the products hand metadata off to each other and how they can start to share media, I think we’re pretty confident we’ll end up with a quite powerful solution for people that today aren’t perhaps looking at Nuke for the type of work they’re doing.”
Should make for an interesting year to come.