IT for Moving Images

This translation was generated automatically.

We talked about digital video in detail on the special episode Bits und so #503 (A Tape a Day). In about three hours, the show surveys software and hardware for digital video production, and covers some of the problems, solutions, and workflows.

I want to revisit one aspect that came up briefly on the show, but perhaps didn’t get enough attention.

I think there’s an interesting gap at the intersection of video production and IT. Systems are no longer black, immutable Sony boxes; they can be integrated into larger IT environments.

There are many examples of successful companies reducing friction at these boundaries. For consumers, obviously, YouTube, Periscope, and Snapchat have made video distribution much easier to access, bringing it to a broad audience.

In the B2B space, a number of companies have succeeded with relatively simple applications. A few examples, not necessarily representative:

• SnappyTV makes it easy to share clips from your own video on social media. It raised almost $3 million and was acquired by Twitter.
• Bitmovin is a cloud service that re-encodes video into streaming formats. Almost regardless of the volume and length, a video is processed in ten minutes. It raised $10 million.
• Kasper Skaarhoj builds control-surface hardware for video production.
• LiveU seems to have successfully moved video contribution—traditionally done via satellite—onto (wireless) IP networks.

Almost all major cloud providers (AWS, Google, Azure, etc.) offer off-the-shelf video services that can provide great value to end users with a little additional packaging. They also offer scalable capacity that you can program yourself however you like.

This even extends to little things that are trivial for IT people. On the show, Nils mentioned a small tool that “only” reorganizes preference files for video-editing software. It saves him endless headaches and time, so it’s worth money too.

I know of a one-person project providing professional video-stream training in a tiny vertical market.

Other ideas:

• Gyroscope-based stabilization in post, as with Hyperlapse.
• AI-assisted subtitles and transcript timing, as on YouTube.
• Database-backed, collaborative management of on-screen graphics.
• Automated recording and distribution of conferences/meetings, like Apple’s discontinued Podcast Producer.
• Something with VR™.
• Asset management and integration with NLE XML formats.
• Versioning for project data.
• Skype TX without Skype; video conferencing; WebRTC to SDI and back.

Or, more generally: video producers and processors need a nerd in the cupboard. Not that nothing is happening in this space, but there’s room to improve. Once the needs of video production hit a wall, there’s probably a software solution.

Over the last few years, I’ve produced around 800 hours of live-to-tape video. That wouldn’t have been possible without the custom software I developed to automate the process.

If you’re a computer scientist looking for your next challenge, or a video professional who wants your everyday problems solved, you might want to talk to each other.

If you have interesting projects in this area for me, get in touch.