Academia.eduAcademia.edu

Figure 7 – uploaded by Nick Redfern

See full PDFdownloadDownload figure
Figure 4: The parameters Ao, a = wf, and 6 of a Hawkes process with exponential kernel fitted to 134 Hollywood films released from 1935 to 2005, with fitted 2D-plane fitted by the linear model 6 = 0.018A9 + 1.315a@ + 0.006. An interactive version of this plot is available as part of the supplementary material for this article.

Figure 4 The parameters Ao, a = wf, and 6 of a Hawkes process with exponential kernel fitted to 134 Hollywood films released from 1935 to 2005, with fitted 2D-plane fitted by the linear model 6 = 0.018A9 + 1.315a@ + 0.006. An interactive version of this plot is available as part of the supplementary material for this article.

Related Figures (8)

Figure 1: Timeline-based editing in DaVinci Resolve. Clips are arranged in the timeline chrono- logical order.
Figure 1: Timeline-based editing in DaVinci Resolve. Clips are arranged in the timeline chrono- logical order.
Table 1: Extract from an edit decision list (EDL) for The Butcher Boy (1917) produced by DaVinci Resolve. Each edit in the EDL appears as a numbered event that contains the shot number, reel number, edit type (V = video, C = cut), source timecode in and out points (hours:minutes:seconds:frames), and record timecode in and out points. cuts: s; = t; —t;_1. The counting process is simply the number of cuts that have occurred to time t. It is natural to think of editing as a point process because this reflects how films are edited. Non-linear editing systems such as DaVinci Resolve or Adobe Premiere Pro are timeline-based, with video and audio clips, effects, and transitions arranged in chronological order horizontally across the screen (Figure 1). The ordered timecodes of the cuts in a film can be exported as an edit decision list (EDL) for use in post-production (see Table 1). The timeline visualises and the EDL describes the editing in a film as points on a line and although filmmakers do not describe their work in terms point processes, the marking of points on a line representing the running time of a film is nevertheless a key concept in how editors think about time in the cinema.
Table 1: Extract from an edit decision list (EDL) for The Butcher Boy (1917) produced by DaVinci Resolve. Each edit in the EDL appears as a numbered event that contains the shot number, reel number, edit type (V = video, C = cut), source timecode in and out points (hours:minutes:seconds:frames), and record timecode in and out points. cuts: s; = t; —t;_1. The counting process is simply the number of cuts that have occurred to time t. It is natural to think of editing as a point process because this reflects how films are edited. Non-linear editing systems such as DaVinci Resolve or Adobe Premiere Pro are timeline-based, with video and audio clips, effects, and transitions arranged in chronological order horizontally across the screen (Figure 1). The ordered timecodes of the cuts in a film can be exported as an edit decision list (EDL) for use in post-production (see Table 1). The timeline visualises and the EDL describes the editing in a film as points on a line and although filmmakers do not describe their work in terms point processes, the marking of points on a line representing the running time of a film is nevertheless a key concept in how editors think about time in the cinema.
Figure 2: Editing in The Perfect Storm (2000) modelled as a Hawkes process with an exponential kernel (Ap = 0.0502, « = 0.763, and B = 0.0106). (A) Counting process. (B) K-S plot: the rescaled times are plotted against the cumulative distribution function of a standard uniform distribution (Ujo, 1;), with a 99.9% confidence band. (C) Conditional intensity function of the fitted model.
Figure 2: Editing in The Perfect Storm (2000) modelled as a Hawkes process with an exponential kernel (Ap = 0.0502, « = 0.763, and B = 0.0106). (A) Counting process. (B) K-S plot: the rescaled times are plotted against the cumulative distribution function of a standard uniform distribution (Ujo, 1;), with a 99.9% confidence band. (C) Conditional intensity function of the fitted model.
1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005
1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005
Figure 5: The (A) background rate (Ao), (B) reproduction mean (j), and (C) decay rate (B) of a Hawkes process with an exponential kernel fitted to shot length data of 134 Hollywood films released from 1935 to 2005 by genre.
Figure 5: The (A) background rate (Ao), (B) reproduction mean (j), and (C) decay rate (B) of a Hawkes process with an exponential kernel fitted to shot length data of 134 Hollywood films released from 1935 to 2005 by genre.
Related topics:
Digital HumanitiesFilm StudiesFilm AestheticsFilm and Media StudiesData ScienceClassical Hollywood CinemaFilm EditingNew Hollywood CinemaFilm and TelevisionHawkes processes
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved