On Helping Hand, we provided 4D facial performance capture and processing services to Axis Studios for the animated short. Both DI4D PRO & DI4D HMC systems combined to produce over 6 minutes of high-fidelity facial animation.
March 15th, 2019 on Netflix
Axis Studios, one of the UK’s leading animation and VFX studios, created the episode titled, Helping Hand, which focuses on the emotional journey of a single on-screen character called Alexandria, played by actress Elly Condron. To achieve a high degree of realism, Axis created Alexandria as a digital double of Elly. The success of the piece, which features several close-up shots of Alexandria, relies heavily on being able to transfer the emotional performance of Elly onto Alexandria. We used 4D capture to record Elly’s facial performance with the highest possible level of detail and realism.
We used an early version of our proprietary DI4D HMC system to record Elly’s facial performance during a motion capture session at Audiomotion’s mocap stage in Oxford, England. This allowed Elly’s facial performance and full-body motion all to be recorded simultaneously, which is important to achieving realistic synchronisation, especially for scenes involving greater physical activity.
We provided a full DI4D PRO shoot at Savalas Post’s audio studio in Glasgow. This allowed us to capture Elly’s facial performance for the close-up shots at an even higher detail than the DI4D HMC system. Setting up the DI4D PRO system in the audio studio allowed production quality audio of Elly’s dialogue to be recorded simultaneously with her facial performance.
The DI4D PRO system was also used to capture detailed facial expressions and shape data from Elly. Axis used this DI4D PRO data to help them build a facial animation rig for Alexandria with animator-friendly controls in order to tweak the performance later if necessary. As a final stage of processing, we solved all of the facial pointcache animation we captured from Elly onto the Alexandria rig in such a way that the rig-solved animations were as faithful as possible to the original pointcache animation.
“The facial performance became such a key thing as half of it is just her expression. We worked with a company called DI4D and the process that DI4D has created is a 3D capture of the entire performance so you know exactly the shape of her face for every second of the performance captured.”
“It was a multi-cam set up along with a body performance set up. The detail of the capture was amazing, we had every little nuance, eye movement and blink. The footage they capture is applied in such a way that it is supplied back and we can animate on top of it.”
The raw intensity of facial emotions portrayed during Helping Hand makes it one of the most memorable episodes in the series and demonstrates the effectiveness of combining our DI4D Pro and HMC systems.
Love, Death & Robots
Created by Tim Miller and David Fincher, Love, Death and Robots is an 18-part series of animated shorts commissioned by Blur Studio and published by Netflix. DI4D provided 4D facial performance capture and processing services for four episodes of Netflix’s animated series Love, Death and Robots, these episodes were Helping Hand, Aquila Rift, Sonnie’s Edge and Shape Shifters. Helping Hand went on to win best Animation at the British Academy Scotland Awards in 2019.