I’ve done a great deal of high profile video streaming production using just about every platform out there (e.g. Google Hangouts, Zoom, Facetime). These technologies are transformative but all lack a few key tools that would make it much easier, for pros and hobbyists alike, to use them to produce remote interviews with higher production value.
With that in mind, let’s first describe the problems which need to be solved:
Poor Camera Feeds
As a producer, I can send my subject equipment and guidance on how to set it up. This can get us close to a good visual and audio source. But more, often than not their are slight tweaks one could make to add a bit more polish to the feed and those are difficult to communicate to our mostly non-technical subjects. Some tools exist, like Webcam Settings. But again, they are relatively technical and difficult for the average person to implement well. If I could just reach out through the Internet and make those tweaks myself, everyone would be happier and we’d have a better project.
In live production, video switching is the equivalent of directing the show. The technical director (TD) chooses where the audiences attention will go. On most group-enabled live-streaming platforms, switching is done either by an inexperienced moderator (using a crude software interface at best) or worse still, the software itself which switches focus based on arbitrary cues it detects in the stream itself (like who is talking at any given moment).
Even the best technical director has a cut they’d like to take back. But before live streaming, the show was ephemeral. Unless specifically producing for concert film or something going out on to the news service bureaus, there was no expectation that sub-optimal cut was going to live forever, let alone be expected to be available for replay within seconds of the house lights going down. Live streams have changed all of that. And, more and more, that technology is being adapted by news agencies, documentary producers and even clever narrative filmmakers to tell their stories.
The solution to this problem is the camera ISO, shorthand for isolated recording. To create an ISO, the producer needs each subject to record their webcam feed alone, not the entire live stream. With powerful audio sync tools like Plural Eyes now so accessible, an editor can quickly take the ISOs, match them to the recorded live stream and produce a more polished cut to be posted to YouTube and other distribution platforms for replay.
Delivering ISOs for Post Production
OK. We have the remote ISO. Now, how do we get it to the editor to make the changes? Well, we face two primary challenges, bandwidth and, again the technical expertise of the subject. Despite modern HEVC/h265 codecs, video files are often very large, particularly if the local record is long and in 4K.
Even if the subject’s Internet connection can support a live stream, it may not be the most efficient way to migrate the local ISO from their machine to the cloud for editing. Finally, there is no assurance the subject will have the technical skill to upload the local ISO to the cloud. And, frankly, that is not something a producer should require of them.
The Best Current Solution
There are technologies like Skype TX, which can solve parts of this problem by turning a Skype call into an HDMI/SDI source that allows a TD to switch it like any other video source. Once in the local production pipeline, color adjustments can made by the TD and a remote ISO recording can be made as well. However, given the heavily compressed Skype signal, the quality of of both does not come close to matching a local camera ISO.
Ultimately, all this solution definitively allows is the ability to fix a sub-optimal cut. And, Skype TX boxes are pretty pricey and require professional expertise to configure and use.
What Might Be The Best Solution?
When I am not thinking in story, I think in systems. Below, I’ve outlined a theoretical system, let’s call it RemCUT, which I think could solve these problems.
ReMCUT: Remote Multi-Cam Capture & Upload Technology
RemCUT would be a SaaS application which, much like Skype or Zoom, works through the browser, a desktop or mobile application or some combination thereof. Based on the volume of remote interviews they conduct, a producer could subscribe to the service or book it for one-off interviews as needed. The workflow would look something like this:
- Producer schedules the interview with the subject in RemCUT, which sends an invitation to the subject.
- Subject accepts the interview time and is sent a pre-flight email, which prompts them to login to the application and test their microphone and camera at least 6 hours before the interview.
- When the mic & camera test is complete, the app sends a production profile to the producer which includes the make & model of the mic and camera.
- If that set up has been used by previous subjects, the producer can apply mic and camera presets to the event in the system.
- 30 minutes before the interview is scheduled to begin, both producer and subject log into the system.
- If the subject’s mic & camera allow it, the producer would apply the presets directly to the subject’s hardware via the application. If not, the presets would act as adjustment layer between the subjects A/V source and the live switch, which is happening in the application.
- The producer would then adjust the remote mic & camera settings as needed to get the optimal image and audio from the subject. Once finalized, the producer would save this as a final preset if needed for followup interviews with the subject.
- Included in the preset, there would be settings to allow the following steps to capture a remote ISO.
- Cloud ISO record of the subject feed, in addition to the cloud record of the switched interview.
- Local ISO record of the subject feed at the highest possible video resolution & audio bitrate.
- Using preset details, establish remote connection to cloud storage.
- Interview begins and both cloud and local ISO recording begins.
- When interview ends, the application uses the bandwidth data and the size of local ISO record to estimate upload time. It then prompts subject to leave the application open while the local ISO record uploads to the cloud in the background.
- When upload is complete, subject and producer are notified. Subject closes the application and producer, if needed, downloads the subject’s ISO recordings for use in post-production.
To be clear, this solution does NOT exist in the market. All of the components exist in different forms, but no one has put them all together into a killer app … yet. But, when someone gets around to it, producers everywhere will rejoice!
Think I am missing something or know of something that does this already? Let me know in the comments. Think I am on to something? Pass it one to an app developer and feel free to loop me into the conversation.
Image Credits: ConvertKit.