The Elements of a Visual Effects Set Extension: Evaluating a spectators level of Believability for VFX shots through a synthesized Model and Methodology framework

Rasmussen, Jonas ; Andersen, Asbjoern

4. term

Medialogy, Master

2012

2012-05-23

112 pages

The following report outlines a 10th Semester Medialogy Thesis at Aalborg University Ballerup. The thesis undertook the process of synthesizing a theoretical model and test methodology to evaluate upon a spectators level of Believability in regards to Fore- and Background elements in a Visual Effects Set Extension. The relationship was hypothesized to be relevant as valuable work and time could be saved if a high amount of background elements were found to be unnecessary when the foreground presented an element of interest. Relevant perception topics were addressed extensively as a means to create the model that would allow us to evaluate the otherwise very subjective topic of Believability. Through the analysis we created a qualitative method for evaluation involving a questionnaire with respective guidelines for interpreting a subject’s answer. The end result was a per-case evaluation with one of two answers; believable or unbelievable. In addition to the model of Believability we created a test methodology that strived to examine whether the Fore- and Background elements in the shot were related and to what extent. Various significance tests including the Chi-square, Factorial ANOVA for Mixed Designs and McNemar were used to test for these relations. Even if numerous considerations in terms of the results must be reflected upon the thesis showed that the Fore- and Background Elements were indeed related. When the Foreground has an element of interest, the level of detail of detail in the Background is less important. Thus the thesis addressed that valuable work/time may be saved in terms of VFX Background details when the foreground presents an element.

Download
View record in AAU Student Projects