"It's an honor, more than anything else, when Apple chooses to single out an application like this ... it's pretty amazing, especially to win two," says Paul. "We worked really hard on the user interface, we worked really hard on the adoption of technologies ... we don't jump up and down and flaunt the technology. The technology is driving the innovation, I think that's something that Apple appreciates. When you hold an iPhone, you don't see all the technology that's jam-packed in there, it just feels right and looks beautiful, and that's what we were trying to do. We said, 'Okay, we have this vision of how we want it to work, let's build the technology and basically hide it.'"
When I asked if he had expected a win, he said, "we entered into the competition, but we'd see other apps come out and [say] 'Oh that's a pretty app,' 'Oh that's nice.' You forget what yours is like because your head's so into it all the time."
If you've used it, you know that ScreenFlow is capable of recording for extended lengths of time with barely noticeable impact on CPU performance. And the amount of information it compiles is pretty astounding considering the unusually small file size of a raw recording. I've been curious about this magic since the first time I launched it. Paul explained, "ScreenFlow records to an intermediate codec, [...] it's not a codec the end user will ever see. It's a way of compressing everything that's going on into one file and then decompressing that. We spent a great deal of effort creating this codec, it's what allows ScreenFlow to do what it does. The best way to describe it is when you're running ScreenFlow, the CPU's not getting killed, the GPU's not getting killed, it just feels like your machine is normal because we're running this internal codec that's saving everything to disk as efficiently as we could conceive of. [...] The codec is designed to do its best to capture at 30 fps while having as little impact on the machine as possible, [...] that's what we saw as the Holy Grail." A little vague, I suppose, but I can respect a secret.
The first reaction of many Screenflow users is annoyance with not being able to record a small part of the screen. Paul says that the technology driving ScreenFlow allows a new philosophy: "record everything and don't think about it. And then later, go ahead and edit." He continues, "we felt that the userbase has been told for years and years that the only way to do screen capture with any reasonable FPS is to define the smallest region that you can deal with on your screen, and then record that. We felt that was extremely restrictive. [The] paradigm change of 'crop after' as opposed to 'crop before' ... a lot of people resisted that until they came around to 'wait a minute, actually it's better that way."
Given what those of us who weren't sitting inside the WWDC sessions were able to pick up about Snow Leopard, I was curious if (and maybe how) the new OS would benefit a graphics-oriented application like ScreenFlow. Paul responded, "Yes, there are things we can take advantage of. xxxxx xxxxxxxxxx xxxxxxxx xxxxxxxxxx xxxxxxxxxxxxxxx NDA xxxxxxx xxxxxxxx xxxxxxxxx xxxxxxxxxxxxxxxxx xxxx xxxxxxxxxxx x xxxxxxxxxxxx xxxxx."
We talked a little about some of Vara's other software, as well. VideoCue, an application that adds a teleprompter to basic video recording, is an indispensable asset for those of us who don't do so well with on-the-spot speech. These days I stutter and blank out a lot when faced with any type of microphone -- some odd form of stage fright that I never had in my youth -- so I was an early adopter of this application. It's not their flagship product, but it's an example of the creative problem solving that Vara has built their reputation on. "We want to provide products that allow people to create media easily. We saw videoblogs coming down the pike [...], and we said 'OK, if you're going to do a videoblog, what's one of the things you're going to have [the most] trouble with?'. For me, it's trying to remember exactly what I'm going to say. So, we said 'okay, let's write a teleprompter system that records your video and allows you to add a couple effects'. It was actually somewhat inspired by [player pianos]. I'd seen those before and we talked about it and said 'that's it.' It's the words next to the action."
We also talked about Wirecast, their application for live webcast production. Specifically, we got into the details of Wirecast's chroma key capabilities. "It's our own engine" said Paul, "a realtime chroma-keying engine, and it uses the GPU quite heavily to do the heavy lifting. Generally chroma keying is quite computationally expensive. We're proud that we can do what we're able to do in realtime. Obviously there are professional solutions out there that are going to be better, but if you put ours up against any realtime system out there that's not hardware based, I think you'll be pleasantly surprised. It's also [almost] zero configuration. You just fire up your greenscreen, turn it on, and that's it." I briefly shared some of my own greenscreen frustrations, with which he readily sympathized, going on to state "we really did try to make it zero configuration. There are a couple of twiddles you can do, but generally we say 'if it doesn't work, you've got lighting problems.'"
Thanks to Paul for the chat, and a hearty TUAW congratulations to Vara on their ADA awards. I'm looking forward to the further development of some outstanding applications.