ATV
Closed Captioning
CHALLENGES & CHOICES
1. The Videotape
The first challenge was to determine an effective research tool. A demonstration of actual Advanced Television captioning was not possible, because it was not yet technically feasible. The goal was to simulate Advanced Television captioning as realistically as possible, while staying within the budget and timelines for this project.
The original strategy was to use the ATVCC Simulator, a computer program being designed by the Working Group, as the research tool. However, the capabilities of the simulator needed to be scaled back because of budget restrictions. The scaled back version of the simulator could not produce captions over moving video, which we decided was essential to successful audience testing.
We agreed that the most effective test would be to use real captions over an actual program. By real captions, we mean captions that are directly related to the video. So instead of arbitrary video with self-descriptive captions (e.g. "Here is an example of a Helvetica font"), the captions would be faithful to the audio.
The Project ultimately decided to produce a videotape in NTSC format and use an AVID editing system to create captions that reflected the different styles possible with ATV. In order to maintain a high image quality, Betacam tape was used.
The Project selected a program from WGBH's NOVA series in order to simplify many of the logistical issues involved in videotape production (e.g. clearing rights, obtaining a copy of the master). Since this program had already been captioned by The Caption Center at WGBH for television broadcast, the Project was able to use existing caption files, eliminating the time-consuming task of creating the captions from scratch, and ensuring that the text and timing were accurate.
Since the videotape could not be demonstrated on an ATV set (since none exist yet), another question was whether to simulate ATV's 16 x 9 aspect ratio by presenting the video in "letterbox" format. The Project decided against letterboxing because it felt that it would not be an accurate representation of ATV's aspect ration. Whereas the shape of the ATV sets will reflect this new aspect ration and enable the image to fill the screen, letterboxing on an NTSC set maintains the correct proportions, but reduces the overall size of the image, resulting in a black band above and beneath the image. The Project felt that this would be distracting and confusing to the participants and that they might not be able to separate this from what they were being asked to evaluate.
2. The Participants
One concern when conducting market research is that participants tend to be biased or predisposed toward things with which they are already familiar. That was one reason we elected to include in the research hearing participants who do not watch captions. However, the question arose as to whether or not to leave the sound on when hearing participants viewed the tape. We ultimately elected to leave the sound on for two reasons: (1) Experience shows that viewers who are unaccustomed to watching captions have a difficult time keeping up, and we felt that if they were struggling with this, they would not be able to evaluate the features; and (2) We had left the sound on during the interviews with deaf and hard-of-hearing participants and wanted to be consistent. However, the fact that the hearing viewers were not as "dependent" on the captions may have influenced their assessment. If there had been time for additional testing, we believe it would have been interesting and worthwhile to include a group of hearing participants who watched the tape with the sound off.
We had also hoped to include in the research people who identified themselves as "hearing" but who were beginning to lose their hearing as part of the aging process. We contacted several Senior Centers to try and recruit participants, but did not get any response.
