I'm going to be vague about my employer and the event because I don't want to publicly give away too many secrets but if you know me you'll know who I work for or you can easily find out.
The station recently put on one of its biggest events of the year and the challenge was to visualise the radio output. We were kitted up with a vision mixer, digital recorders, three Panasonic remote control cameras mounted in the OB studio, a roaming handheld Canon camera for capturing any action at the event which didn't happen in our studio and a Mac Mini with a BlackMagic box and Wirecast installed.
We were getting a ton of guests in to the studio and the kit would allow us to vision mix the interview (taking the audio feed from the desk) and either stream it live via Wirecast or record it on to a memory card for play out later while Wirecast streamed holding slides alongside station output.
In theory it's was great, Wirecast is a powerful bit of software, a virtual vision mixer which is practically a virtual TV station. Wirecast took care of the encoding, allowed me to stream the live vision mixed output from the studio, play back pre-recorded video from the studio or the roaming Canon footage. It also allowed me to control the audio which audio was routed through to the stream, play pre-rendered videos which looped and cycle through holding images which displayed station messages.
Event coverage like this has been done before, most notably by BBC Radio 1 but there's a very big difference between the BBC and commercial broadcasters, especially my employer and that is slickness and polish. Radio is always the main concern so to make the radio show sound as good as possible most of it is pre-recorded and heavily edited.
I'm not going to debate live vs pre-rec right now, I'm sure many of you will have strong feelings either way but the fact is it does allow for a more polished broadcast, the problem it causes when visualising content is that you can't stream.
My plan was to record video alongside all the pre-recorded interviews and play it out on the stream to roughly coincide with the play out on air, the rest of the time I was going to stream the station output with the rotating slides during songs and ads and cut live to the studio cameras for the presenters' links.
I knew that the video versions of the interviews wouldn't sync up with on air because it requires much more time to edit video and there wasn't the time to match it to the audio at a live event with a team of just two. The idea was to tell the same story, the same conversation so the transitions between the station output and the video inserts was a little clunky, it meant fading down the the audio feed and dropping in the video then fading back the radio feed after, there was the risk that I might opt back in to the show during a link, ads, halfway through some production but was hoping in most cases I'd opt back in during the start of the next song.
In reality this proved to be really difficult at an event of this scale to follow a show plan, I ran the stream for six hours and in all that time there were less than 10 live presenter links. A lot of interviews went out which we didn't have video for because on air had run off to record the audio at short notice.
Interviews got dropped due to time restrictions with no notice, at times I had already opted out to play the video before I realised the content had been dropped from on air. Where the on air team had two interviews of some artists they ditched one interview in favour of another without notice which meant the wrong interview played out on the visual stream.
The main thing I have learnt from this first attempt at live event visualisation is that working together is the most important thing without a doubt. The visual team need to work with on air at all times to make sure that any content that it recorded for on air also has a visual element, this of course comes down to communication so that we know what the on air producers and planning and we can ensure we go along with them.
Once content has been capture in audio and visual form it then needs the same treatment, the video needs to be edited alongside the audio which might mean adding extra time in to show logs to allow for the more complex video edits and rendering. It's much quicker and easier to turn around audio.
Communications is a massive part of this, things can get hectic in a live studio and decision have to be made with seconds notice but as the director of the visual output I need to know those changes, in most case I can react as quickly, I just need to know what's happening.
As this was our first real venture in to visual radio at an event things were done on a bit of a budget, to pull something like this off we really needed more crew. This time round there was someone manning the vision mixer mixing the studio content and using the roaming camera. And there was me, directing the output, doing the live mixing and dropping in pre-recorded content and holding content.
Technically the stream should have worked although could never be as slick as visualising live radio content. Visual radio is something which is only going to get more important within the industry so we need to find a way to make this kind of digital coverage work without compromising the on air output and I do believe that pre-recorded radio content can form an engaging visual stream, this attempt was in no way a failure, just difficult.
Last year Josh & Kenny from Crush Radio, the University of Hertfordshire's student station produced a visual show all with pre-recorded elements. They took a different approach, the two third year media students recorded video alongside all of their on air links and uploaded them to YouTube. On YouTube they used the playlist tool to add in music videos and build the show. It meant listeners had the option of either listening to the pre-recorded links as they were played out on the station or watch the YouTube playlist like as "as live" show. This also meant the whole show is available to re-watch online afterwards.