Chuck Boller of the Hawaii International Film Festival, CC BY-SA 3.0, via Wikimedia Commons
Roger Ebert Voice Replacement
In 2009 we began work with Roger Ebert to replace his voice using material he had recorded as video commentary for a series of films on video cassette. Roger had seen the George Bush website and thought that if we could clone one voice we could clone his.
RusPlot, CC BY-SA 3.0, via Wikimedia Commons
Robo Jas from the Gadget Show
I’m sure everyone remembers the adrenalin fuelled crazyness of Channel 5’s Gadget Show. Not just content to report and show technology they also liked to try and build stuff to show how things worked. In one episode this involved a robot version of Jason Bradbury which they used to co-present with a holographic version of Suzi Perry. But the voice wasn’t up to scratch so they got in touch with CereProc.
Luxottica, CC BY-SA 3.0 via Wikimedia Commons
Oakley Radar Pace Sunglasses
CereProc collaborated with the Intel New Devices Group and Oakley Sunglasses to create the unique Radar Pace Sunglasses. These high end shades came equiped with a coach for cycling or running. The glasses could monitor your progress and give advice to help you improve.
Cecil Stoughton, White House, Public domain, via Wikimedia Commons
JFK Unsilenced
CereProc were approached by Rothco, a PR company in Ireland whose objective was to recreate an authentic JFK’s voice based on previous speeches that he had given so that he could finally deliver the speech he was meant to deliver in Dallas 1963. The CereProc team over the duration of 8 weeks analysed recordings of 831 speeches and built the voice by splitting them into 116,777 small phonetic units.
ITU Pictures from Geneva, Switzerland, CC BY 2.0, via Wikimedia Commons
Sophia the Robot and Jimmy Sing a Duet of "Say Something"
CereProc created the original voice for the Hanson Robot Sophia. We later worked with them to produce a singing voice for the robot and this was shown on the Tonight Show, where the robot and Jimmy Fallon sang a duet.
DonkeyHotey, CC BY-SA 2.0, via Wikimedia Commons
More Voice Cloning
Neural TTS meant it became a lot easier to clone voices. We had also developed singing synthesis so we thought why not clone the President of the USA’s voice and have him sing as well. It seemed funny - at the time.
Scottish Child Speech Synthesis
Call Scotland’s mission is to help children across Scotland overcome disabilities and barriers to learning. The project, funded by the Scottish government, was to create 2 Scottish accented natural female and male child voices. Call Scotland a close partner of CereProc, approached our speech synthesis experts with the project to give Scottish children the ability to speak in their regional accent.
daryl_mitchell from Saskatoon, Saskatchewan, Canada, CC BY-SA 2.0, via Wikimedia Commons
Haegue Yang: The Pine's Eye
CereProc supported international artist Haegue Yang for the Pine’s Eye exhibition at the Talbot Rice Gallery. Haegue presented an ensemble of sculptures from The Intermediates series on top of a nonagon shaped floor graphic with a new sound element synthesising her own voice, blending ideas of modernity, estrangement, pagan traditions and folk cultures to unsettle familiar hierarchies.
Public domain, via Wikimedia Commons
World’s First Non-binary Synthetic Voice
CereProc, collaborating with Accenture Labs created the world’s first non-binary synthetic voice that combined male and female data from transgender and non-binary individuals for the global digital assistant market. The companies released all materials to generate the voice to the Open Source community.
The Sound Voice Project
CereProc collaborated with the award winning Sound Voice Project. We met Hannah Conway while working with Peter Scott Morgan and worked with her team and collaborators to use voice cloning technology in the fascinating installation I left my Voice Behind
