5 - COMBINING FILM AND MUSIC

[ 5 ]
COMBINING FILM AND MUSIC

5.1 Aim of experiment

In the second part of the research, we were able to select the visual and auditory material that best met our criteria. Then each film excerpt was associated with two different pieces of music, resulting in two sets of six audiovisual combinations each. These sets were presented to different groups of participants that, for each combination, rated how well film and music fit together.

We hypothesized there could be a relationship between the above mentioned ëfití ratings and the differences between film and music on the Pleasure/Arousal dimensions.The audiovisual pairings created and tested in this phase of the research have then been used in the final experiment.

5.2 Stimuli selection

From the initial bulk of twenty-two soundtracks and twelve film sequences, six pieces of music and six film sequences were selected and then used in the final experiment.

This selection was based on a number of different criteria:

- Stimuli with lowest variance values on the arousal and pleasure dimensions were preferred.

- The final selection of soundtracks had to be as heterogeneous as possible and cover a variety of different musical genres.

- Selected film sequences had to be somewhat ambiguous, without a definite conclusion, so that they could be accountable for different interpretations rather than having a clear and unique meaning.

Please note that compared to the initial bulk of 22 pieces of music and 12 film excerpts, some stimuli have been removed.

This is because in the previous phase of the research the pleasure and arousal ratings of that material were found to be significantly different between female and male or between musically trained and musically untrained participants.

We calculated the overall variance for each of the remaining musical and visual stimuli, by means of averaging its Pleasure and Arousal variance scores. Stimuli have then been ordered according to their overall variance.

Results for music excerpts are displayed in the following chart:

Chart 5.2.1 - Musical tracks sorted by ascending overall variance scores.
Items in brackets are those selected for the final experiment.

This is a list of brief descriptions for the selected music material:

#01. Lydiom

Slow ballad. Trumpet plays theme.

#04. Sunshine

Up-tempo piece with a Latin (samba) rhythm.

#06. Chili

Fast jazz (be-bop) tempo. A brass section (trumpet, saxophone) plays the melody.

#13. Throw

Classic rock song. Uptempo.

#16. Duncan

Downtempo ballad. Full orchestra with strings playing a major role.

#17. Grace

Slow tempo. Piano soloing, low strings accompaniment.

(Item number refers to the initial selection of 22 pieces of music)

This chart displays overall variance scores for film excerpts:

Chart 5.2.2 - Film scenes sorted by ascending overall variance scores.
Items in brackets are those selected for the final experiment

Here is a list with brief descriptions for the selected film scenes:

#01. Phone

Daytime. A man and a woman talk over the phone. The sequence features alternate close-ups of the womanís and the manís faces. From their expression it is clear that the conversation is emotion-laden.

#03. Couple

Daytime. A young couple is driving a convertible sports car. The trees and other vegetation suggest they are some place out of town. The young man is behind the wheel while the woman talks to him.

#04. Walking

Evening. Downtown street. A young woman is walking alone. She approaches a parking lot, and then moves on to the sidewalk, passing the window of a bar.

#05. Cars

Night. A car is running on the highway. A young couple is in the car. The man is driving and keeping his eyes on the road, while the woman often looks back. A second car approaches the first car, then moves to the fast lane and passes it.

#06. Cheers

Living room. Two man and two women are there. The older man (in his sixties) is opening a bottle of champagne with a corkscrew. He pours the wine into everyoneís glass. The young man (in his twenties) refuses the glass and says something. Then the older man and the two women have a toast.

#11. Flowers

Daytime. Flower shop. A young woman walks in and asks some questions to the old lady that is selling flowers.

(Item number refers to the initial selection of 12 film excerpts)

5.3 The appropriateness of film/music combinations

In the subsequent step of the research we put together film and music, so that each film sequence was paired with two different soundtracks. The result was twelve different combinations of music and film, which were divided into two groups.Both groups featured all film sequences in the same order, but the soundtrack for each sequence differed between groups. In this way participants in each group would watch all six sequences and listen to all six soundtracks, but each individual stimulus, whether music or film, was presented only once to the same person. The order of presentation for audiovisual stimuli was initially randomized, then kept the same for the subsequent experiments.

Table 5.3.1 contains the actual list of combinations as well as their order of presentation for both group A and group B.

Combination	Film sequence	Music for group A	Music for group B
#1	Walking (03)	Lydiom (01)	Sunshine (04)
#2	Couple (02)	Duncan (16)	Chili (06)
#3	Cheers (05)	Chili (06)	Grace (17)
#4	Phone (01)	Throw (13)	Duncan (16)
#5	Cars (04)	Grace (17)	Throw (13)
#6	Flowers (06)	Sunshine (04)	Lydiom (01)

Table 5.3.1 - Audiovisual combinations used in the experiment.
The number in brackets refers to the order used in the preliminary appraisal of stimuli.

5.4 Participants and experimental design

Participants (f=18, m=16), once again students of University of Padova, were split into two groups of 17 and assigned to combinations A and B. They all were in the same room while the film was projected on a large screen and music was played through the speakers.

This time the only rating to perform was on the following scale:

Film and music are:

contrasting coherent

( 4 )___ ( 3 )___ ( 2 )___ ( 1 )___( 0 )___( 1 )___ ( 2 )___ ( 3 )___ ( 4 )

5.5 Results

As we did for all other rating scales, also scores for the coherent/contrasting scale were translated to positive values ranging from 0 to 8. Results in average fit for group A and group B were:

Combination	Fit (group A)	Fit (group B)
#1	7.06 (1.80)	1.53 (1.64)
#2	2.53 (3.39)	5.41 (3.88)
#3	4.59 (3.75)	4.24 (3.56)
#4	1.12 (1.61)	6.42 (2.06)
#5	6.24 (1.56)	5.82 (2.40)
#6	2.24 (3.06)	3.53 (3.01)

Table 5.5.1 - Film/music fit values (mean and variance) for group A and B.

We compared the judgments of males and females using independent samples t-tests. All differences were non-significant.As mentioned above, we suspected that fitness values could be somewhat related to the proximity vs. distance of the Pleasure and Arousal score for the film and music track used in a particular combination.

To verify this we calculated two indexes named Distance-in-Pleasure (DP) and Distance-in-Arousal (DA) respectively, defining two new variables as:

Distance in Pleasure = Pleasure of Film(x) Pleasure of Music(y)

Distance in Arousal = Arousal of Film(x) Arousal of Music(y)

Where film(x) and music(y) are the visual and auditory stimuli used in a particular combination. We considered the absolute value of the result (i.e. the score regardless of its positive or negative sign).

DP and DA values were calculated for all 12 combinations (6 for group A and 6 for group B)

Combination	Fit	Distance in Pleasure (DP)	Distance in Arousal (DA)
#1 (group A)	7.06	.58	1.85
#2 (group A)	2.53	2.47	.76
#3 (group A)	4.59	1.14	2.02
#4 (group A)	1.12	3.34	3.38
#5 (group A)	6.24	.35	3.76
#6 (group A)	2.24	1.50	2.05
#1 (group B)	1.53	1.58	.91
#2 (group B)	5.41	.84	2.12
#3 (group B)	4.42	1.59	1.65
#4 (group B)	6.24	.10	.44
#5 (group B)	5.82	2.51	.85
#6 (group B)	3.53	.66	.71

Table 5.5.2 - DP and DA scores for all combinations of film and music

Calculating Pearson's r coefficient, Fit and DP proved to be significantly correlated:

r (Fit, DP) = -.645*, n=12, p<.05

Instead there seemed to be no relation between Fit and DA values:

r (Fit, DA) = -.009, n=12, n.s.

Therefore, when used together, music and film with similar Pleasure scores (that is to say a low DP value) will get significantly higher fit scores. This however will not occur if music and film are given similar scores on the Arousal dimension.

[ next ]-[ index ]