This post presents some results from the English and Hungarian TV-versions of G.B. Shaw’s play “You never can tell”.
Shot sizes
Image size plays a crucial role in the viewer’s ability to perceive the state of play and the movements of individuals in it. In a extreme close-up, the actor’s face fills the entire screen, and the environment is not visible at all. As the image size grows, the environment is becoming more and more visible, and the viewer has the best chance of perceiving motion trails. The importance of image sizes as a means of highlighting the differences between language groups in the previous study (Strømnes et al., 1982) became clear fairly late. It was only when the results were analyzed using image size as a grouping variable that they started responding to the viewing experience.
The Hungarian version had more close-ups than the English. For large images, the situation is the opposite. In the English version, the three largest image sizes represent 36.4% of all cases and the Hungarian version only 26.4% (Figure 1). The difference is very significant (X2 = 46.87, ν = 5, p <0.001). Time distributions (Figure 2) are similar in style: the English used 22.5% of the time for the two largest picture sizes, while the Hungarians only 13.5%. The English style of wide-screen shooting gives the viewer a much better chance of perceiving the space of the play and the position of the actors in it. The difference is very much like that between Finnish and Swedish / Norwegian productions (Strømnes et al., 1982).
Figure 1. Distribution of shot sizes.
In the English version, there were 1039 separate shot sizes and 864 in the Hungarian version. When the durations of the plays were 7303 and 6518 seconds respectively, the average duration of one shot size in the English version is 7.0 seconds and 7.5 seconds in the Hungarian. The difference between versions is mainly due to the three largest image sizes that are longer in the Hungarian version (Figure 3). Due to the way the image sizes are measured (the change in shot size was also taken into account when the camera or the person was moving / the size of the image is human), this result would give the Hungarian version fewer camera and person movements in large shot sizes.
Figure 2. Time used for shot sizes.
Figure 3. Average duration of shot sizes.
Shot change distance indicates how many image sizes there are between two shots. As a descriptor of the impression the viewer has, it is not entirely unambiguous. A big change usually startles the viewer if the he / she is not prepared for it, but a small foresight before the cut may motivate a big change so that the viewer will not be surprised at all. Thus, for example, a conversation described in close-ups (CU) of two persons can be moved flexibly to an incoming person in full shot (FS) or to even in a long shot (LS) if one or both moments before the cut glance aside, anticipating the future change. However, this study did not specifically focus on cutting techniques.
There were 851 cuts in the English version. In addition, the image change was done with fading / opening once (between the second and third act). There were seven instances of cross-stitching between the scenes moving from a drawing to the actual environment and vice versa. The Hungarian version had 577 cuts and three fades / openings between scenes.
The change in image size during cuts was also analyzed by taking into account the zero change (Figure 4). In both versions, the most common change is that of one shot size. In the Hungarian version they comprise almost half of cuts (46.3%). The next most common is the same image size (27.7%). The last quarter is divided in this version between two, three, and four image sizes.
In the English version shot change distances are more evenly divided. In this version there are about 10 % less changes of zero or one image size than in the Hungarian version but the again more changes of two image sizes. Differences in larger changes are not great. The difference between the distributions is very significant (X2 = 42.27, v = 3, p <0.001).
The difference between the distributions is not easy to interpret based on the analyzes that have been made. It would require a separate study to find out how and where changes of angles were made. The abundance of zero changes in Hungarian material may be due to lengthy discussions that show each person in turn using the same image size.
Figure 4. Shot change distance
Both versions had no spectacular big changes in image size for the viewer, which means that the cuts were generally well motivated in both versions. One major explanation of the larger number of large (2 or more) changes in the English version relates to the above-described description of incoming persons. The English followed much more persons moving towards the camera than the Hungarians, and they had fewer omissions of parts of movements. I will return to this in the movement analysis.
Image depth
The distribution of the image depths in the image sizes is shown in Tables 1 and 2. The differences are partly due to the fact that the English version uses the full 360 degrees of space in the second act. The events of the second act take place on the hotel terrace where there is a view to the sea. The Hungarian solution is more theater-like. The viewing angle is mainly from the direction of the imaginary spectator. The first act takes place in the dentist’s reception room and the second on the outdoor terrace of the restaurant. The third and most of the fourth act is placed in a large room next to the terrace. In practice, the solutions are otherwise identical. The English version also featured other restaurant guests who were not present in the Hungarian version. For this reason, one might think that the Hungarians chose a slightly different way of shooting this act to avoid giving a picture of a deserted hotel. Actors stayed close to the wall of the terrace.
Table 1. Time distributions of depths 0-4 m and 5+ m.
| England | Hungary | |
| 0-4 m | 54.7 % | 82.2 % |
| 5- | 43.1 % | 17,8 % |
| Pictures | 2.2 % |
Since the use of the layout differed in the versions for the second (and partly the fourth) act, the distributions of depths were calculated separately for each act for 0-4 m and over 4 m. The differences are also significant for the first, the third and the fourth act that were performed in very similar one-room spaces. In the English version, the distribution of depths follows the possibilities provided by the space. The least deep images are in the first act with the least space.
Table 2. Frequencies of depths 0-4 m in each act.
| ENGLAND | HUNGARY | |||||
| Act | % | n | % | n | z | p |
| 1 | 86.6 | 217 | 92.2 | 245 | 1.965 | <0.05 |
| 2 | 65.9 | 188 | 81.0 | 290 | 3.727 | <0.001 |
| 3 | 72.4 | 232 | 90.3 | 135 | 4084 | <0.001 |
| 4 | 73.4 | 263 | 84.4 | 224 | 2.942 | <0.005 |
In the English version, the sections shot in the hotel’s interior also take advantage of the possibilities offered by the staging solution. The act shot inside the hotel often have a door leading to the terrace, where the terrace and the “sea” behind it are visible. The Hungarian version scores 55.9% of the smallest depths (0-2 m) but English version only 37.9%. Thus, in the Hungarian version, more than half of the time is used for images with a background blur or a very small image depth. The use of shallow depths and faded backgrounds greatly restricts the viewer’s ability to perceive the play’s space and the position of the actors in it. The Hungarian description thus focuses more on depicting people and their relations than on the space and the people in it .
REFERENCES
Strømnes, F.J., Johansson, A. & Hiltunen, E. (1982) The externalised image. A study showing differences correlating with language structure between pictorial structure in Ural-Altaic and Indo-European filmed versions of the same plays. Helsinki: The Finnish Broadcasting Corporation, Report No. 21