Greger Wikstrand, Gene Hughson and I recorded Fail fast, learn fast last week, and this episode is a very good example of the topic we discussed in the show.

So how did we fail, or was it a failure and what can we learn?

We don't think we lived up to our ambitions in this episode, but we tried to make up for the issues in recording during post production.

I.e. replacing remote video with a still image of Gene, from a short video sequence and adding subtitling for the Skype call.

We have been recording Architecture Corner for a year now and we are constantly trying to improve our shows, learning from each episode.

The new thing with this episode was that we had a remote guest via Skype in US and that we recorded our part in Malmö, Sweden. It was not the first time we did Skype calls or Hangout. We tried the basic approach before, but not in this environment.

We used a projector to show the remote guest live on a canvas behind us and using two videocameras to record the session. One medium shot and one close shot in cinematic terms. We also used a webcam to the computer running Skype so that the remote guest could see us.

We learned in the early days of Architecture Corner that sound is important and have since then invested in semi-professional sound equipment on our side. But a chain is not stronger than it's weakest link. In this case, the sound from the guest recording.

You can't use a mobile phone and get a decent recording with a built-in microphone when making a Skype call. On top of that, you need to assure that the rooms on both locations are quiet, no fans from equipment or air-conditioning.

Second issue was video from remote guest. Filming a live call on a projector and get a good quality footage is not the easiest thing to do.

Third issue was the length of the recording. We cut down 36 minutes interview to 12 minute of video. We have to be focused on what we say and how we say.

Lessons learned from this experiment are:

prepare the guest of the requirements on his side
sound-check with new guests and/or new equipment a day before
evaluate the possibility to use a large TV instead of a projector
use an analog mixer to avoid echo in recordings
need for more rehearsalbefore shooting
don't film yourself as you loose focus, you must have a dedicated camera man

What were the consequences with this failed production? We did survive without damages to ourselves or the equipment. It took much more time in post production than anticipated, but there was no extra cost associated to this.

Will we loose any viewers due to this session? We don't thing so, as we have the texting as a complement to a louse sound on the phone line. We can also look at the statistics on Youtube in a few week to follow up the number of views, how much the viewed and if the number of subscribers increases or decreases.

What is your example of fail fast, learn fast?

Disruptive architecture