No comments

ABSTRACT

One of the properties often identified as having an impact on the television viewing experience for subtitle users is the rate of subtitling (measured in words per minute) (1,2).

Previous studies on the subject have often restricted participants from using residual hearing or lip-reading as they naturally would when viewing television (3,4,5,6). Additionally, some studies were carried out with potentially biased participants (5,6).

No research has been done to date at a large scale on the rate of scrolling subtitles as are often used in live subtitling (5,6).

This paper presents the results of a study examining the impact of subtitle display rate on enjoyment for a representative sample of subtitle users.

Specially created and off-air material was used with both block and scrolling subtitles. Sound was available and lip-reading was possible. The results challenge previous assumptions.

INTRODUCTION

The rate of subtitles is often highlighted in subtitling guidelines as an important factor in viewer understanding and enjoyment (2), but no scientific justification is provided.

When the few papers currently available on the subject of subtitle rate are examined, it becomes apparent that the quality of previous research is poor and findings vary wildly as a result.

Furthermore, research in the field repeatedly fails to use un-biased regular subtitle users (i.e. people who use subtitles once a day or more) as participants and fails to use normal television viewing conditions (3,4,5,6).

This paper presents the findings of a new study that improves on previous work and answers some questions while questioning the validity of others.

What is the ideal rate of subtitles for subtitle users?
At what subtitle rates is enjoyment diminished for subtitle users?
How do these rates compare to the enjoyment of speech at various rates for hearing viewers?
Does the rate of subtitles even have an impact on enjoyment?

BACKGROUND

Subtitle Rate Measurement

Subtitle rate (also known as the speed of subtitles) is most often measured in Words Per Minute (WPM). This may be calculated in a number of ways. The most common method used is to take an average over a period of time by dividing the number of words in a clip or programme by its length in minutes. This method is used in much of the available academic literature.

While this is a simple method to implement, it may provide low readings for clips with long periods without speech. This may be accounted for by excluding long periods of silence from calculations. Studies generally choose their clips carefully to avoid this problem.

The measurement of rate in this study used this method. Clips in part 1 of the study had no periods without subtitles. Periods without subtitles in the clips in part 2 were excluded from calculations.

Subtitle Rate in Guidelines

Guidelines often quote optimal and maximum rates for subtitles. Figures of approximately 140WPM as the optimum subtitle rate and around 180-200WPM as the maximum rate are common.

The guidelines examined fail to cite research supporting these figures but justify them by stating that above these rates, subtitles will be difficult to follow (2).

Prior Research

The small amount of published research on subtitle rate varies wildly in quality. Participants are sometimes selected from biased or non-representative groups. These include people who work in subtitling, people who do not use subtitles and people from interest groups who may subconsciously aim to represent the standard views of their group (5,6,7).

Many studies also purposefully aim to reduce experimental influences to the subtitles alone by using footage without those speaking in shot or by using clips without audio (3,4,5,6). This has the un-desirable side effect of creating an un-natural viewing experience.

Viewers normally use visual cues such as lip-reading or facial expression to support subtitles. Most subtitle users also have some level of hearing and thus use subtitles in conjunction with audio. Viewers’ experience is a combination of these sources of information.