Game Development Reference
In-Depth Information
8.5C
INEMATIC
L
OCALIZATION
Cinematics arevideosequencesthatareusuallyusedtotellthegame'sstory.Cinematicsare
usually encoded using video formats that will vary in quality, performance and cost. There
are some special considerations to have when preparing cinematic assets for localization.
Mostvideoformatshavebuiltinsupportformultipleaudiotrackstobeencodedintoasingle
file, some may also provide support for encoding text subtitles into them. The advantage of
using this support, in particular for localized audio is that the audio and video will generally
have very good synchronization, the video may be played using the audio as the driver of
the streaming such that if there is a lack of synchronization the audio will have higher prior-
ity while the video may drop some frames to compensate. Other configurations may opt for
using the video as the master, unfortunately this approach may result in audible skips. Our
ears are far less forgiving than our eyes, we may not notice if a video interpolates between
two or more frames depending on the visual coherence of the cinematic, however and audio
skip is hard to ignore.
Video encoding may be a time consuming process. And videos may need to be encoded
multiple times during a game's production, especially closer to the end of the development
cycle. Consider that every time a new audio track is delivered by the localization team the
video will be need to be encoded to include the new track. If subtitles are encoded within
the video, fixing a small typographical error in the text would require a full re-encoding of
the video. The problem becomes more severe when the game is being released in multiple
platforms, videos are generally encoded using different quality settings depending on the
platform they will run on.
