Mainichi Broadcasting (MBS) | Improving the Efficiency of Video Editing with AI

Mainichi Broadcasting System, Inc., or MBS, is a radio and television broadcasting company serving the Kansai region. Monstarlab was asked to improve the efficiency of video editing processes for their annual concert "Suntory presents Beethoven 9th with a cast of 10,000". The event brings professional and amateur singers from around the world creating a "Choir of over 10,000 people that transcends Generation, Gender, and Nationality."

Putting 10,000 Voices Together

"Suntory presets Beethoven 9th with a cast of 10,000" is an annual celebration of the Symphony No. 9 in D minor, Op. 125, a choral symphony composed by Ludwig van Beethoven.

In 2020, to prevent coronavirus spread during the pandemic the choirs were not allowed to participate in the event. Instead, 10,000 videos were gathered from the public to be played on screens at the venue.

The challenge was that due to different environments, lighting conditions and footage quality the time it took to edit the material was excessively long.

AI-Powered Solution

To address the problem Monstarlab first divided the submissions into video and audio components. First, we sought a solution to align multiple audio signals using cross-correlation technology to determine which part of the waveform corresponds to which part of the choral symphony. At this stage, we also removed noise and artefacts, and also normalised volume, to bring public submissions in line with the expected quality level for the event.

40% Faster Editing

Monstarlab analysed and adjusted 14,215 video submissions and used AI to help synchronise and cut the videos, as well as align the subject of each video. This helped greatly simplify and automate pre-editing work which allowed MBS editors to reduce overall editing time by 40% compared to the previous year.