A buffer initially preloads a text subtitle stream recorded on a recording medium. The preloaded text subtitle stream includes a dialog style segment and a plurality of dialog presentation segments. The dialog style segment defines a group of region styles and palette information. Each dialog presentation segment contains at least one region of dialog text, where each region of dialog text is linked to a selected one of the group of region styles defined in the dialog style segment. A text subtitle decoder reads a palette update flag included in each dialog presentation segment and presents each dialog presentation segment according to an indication of the palette update flag. The palette update flag indicates whether to use the palette information defined in the dialog style segment or to use new palette information when presenting each dialog presentation segment.