Screen seeks 3D subtitle standard

Currently, there is no standardised means of distributing this type of material, which would facilitate the safe insertion of burnt-in or optional graphics such as subtitles in 3D TV content.

“A lot of people are placing subtitles at a fixed depth out of the screen, which induces significant visual fatigue because the viewer’s eyes are constantly being forced to focus and refocus unnaturally on objects,” warned Screen Subtitling’s 3D product specialist Jonathan Jenkyn.

“The continued mismatch of convergence between 3D scenes and inserted graphics will induce headaches, and will sabotage the 3D TV business.”

Jenkyn is the co-author of a white paper that proposes that a ‘disparity map’ be generated at the mastering facility and maintained as metadata throughout the production chain and beyond.

Using an infrared camera attached to a 3D rig, he suggested that the depth map could record the distance from the camera of every object in a scene.

Provided that this was updated as the shot was tweaked during production, a second set of data, the disparity map, could be derived and used to automatically place subtitles according to the relevant depths in each scene so
that a viewer’s experience of reading captions is complementary to the 3D material.

“Disparity maps are much more valuable to us as they represent the actual perceived disparity within a piece of stereo material, rather than a depth map, which may not,”said Jenkyn.

The company is currently liaising with standards bodies the DVB consortia and SMPTE to incorporate its ideas.

For example, in order to correctly position graphics and captions in live 3D broadcasts, Jenkyn suggested that an additional delay in transmission of up to 4.5 seconds may be necessary.

His white paper, entitled ‘Safe graphic insertion for stereoscopic media’ and written in partnership with chief technology officer Simon Hailes, is available via Screen Subtitling Systems’ website.