Avatars

Researchers wearing VR headsets can explore the spatial dynamics of the theatre models from the perspective of both actors and spectators. The Queen’s model has been further enhanced with avatars to create a performance laboratory. A similar development of the Komediehuset model is currently in production.

To accurately create the three-dimensional acoustic of the theatre, actors and spectators must be represented inside the model as avatars.

Displaying large numbers of avatars animated by motion capture files in the VR Queen’s is technically complex. An initial test was conducted using a draft male avatar.

Adhering to the archival documentation, a male and female avatar were created that represented the majority of the Queen’s spectators. In 1841, they were aged between 17 and 35 years and came from London and its surrounding boroughs. Costume colours, hairstyles, and simple modifications of the male clothing were used to create variety.

Darren Pack from Ortelia compressed the physical characteristics, costumes, and hairstyles of the avatars into single digital objects to prevent the animation slowing down. We have not created avatars to represent any child spectators, nor the one or two First Nations elders who may have been in the theatre on the opening night.

Digital Costuming

Historically accurate costumes for the avatar spectators were realized by Finnish digital costume designer, Maarit Kalmakurki.

A male avatar dressed in a collarless shirt was used to represent working men in the lower auditorium, a waistcoat was added to this figure to signify the wealthier young men in the gallery, and the same avatar dressed in a full evening coat depicted the colonial establishment in the dress circle and boxes. A female avatar in a dress was used for both the working women and the wealthier spectators, with class differences signified through hairstyle, colour, and cloth texture.

Throughout the process, the overriding concern was to minimize the underlying polygon structure of the avatars while achieving an historically accurate silhouette.

Animating the Avatars

Motion capture technology was used to animate the audience of avatars and the avatar actor. 

The animation of the audience was subject to the same file size limitations as the costumes and the spectatorial responses were restricted to five basic variants of the same movements: focusing on the actor on the stage, looking around the auditorium, and applauding. Additional physical movements representing a variety of spectatorial responses were captured, but not used, as they overloaded the VR model.

The British actor, Gerrard McArthur, animated an avatar representing the colonial actor, Gustavus Arabin, performing the prologue that opened the Queen’s Theatre in 1841. He worked to a soundtrack of the prologue that he had pre-recorded in an anechoic chamber.

Putting it all together

When all the elements for the performance of the prologue had been completed, Darren Pack put everything together in the Queen’s VR model. More complex audience reactions could have been visualised with a substantially reduced number of avatars. Equally, if the focus of the work had been limited to the stage, one or more actor-avatar could have been created with character-based costumes, and with facial movement added to the body animation.  However, the aim of this first experiment was to reproduce an interaction between the avatar actor and the auditorium filled with the avatar spectators. Finally, the soundscape of the performance was added using the acoustically enhanced VR Queen’s.