Given a viewing distance, this function calculates the on-screen diameters of a hypothetical object of defined size approaching at a user-supplied variable speed profile. This allows a looming animation with precise parameters to be created.
variable_speed_model(
x,
screen_distance = 20,
frame_rate = 60,
attacker_diameter = 50
)numeric. Vector of speeds (cm/s) of the hypothetical approaching
attacker at the same frequency (Hz) as the frame_rate. Length of the
vector will thus determine total duration of the resulting animation.
numeric. Distance (cm) from the playback screen to your specimen.
numeric. Frames per second (Hz) you want the resulting animation to be played back at.
numeric. Diameter of the hypothetical approaching attacker. This affects the size of the simulation in the final frame of the animation.
List object containing the input parameters and the resulting diameter for each frame in the animation.
Calculates the screen diameters for a modelled object of specified
size approaching at a variable speed. The variable speed must be supplied
as a vector of speeds (operator x) in cm/s at the same frequency
(i.e. same Hz) as the entered frame_rate, which is the frame rate
the resulting animation created from this model will be played back at. If
the speed profile is in a different frequency than common video frame rates
(e.g, 24, 30, 60 fps) it is recommended to interpolate or subsample it to
be so. I can't guarantee all playback software will correctly play videos
encoded at odd frame rates, though you are free to try.
The output list object can be used to create a looming animation in
looming_animation. Unlike constant_speed_model, no
starting_distance operator is required: it is assumed the final
value in the speeds vector occurs at the point of zero distance
between the attacker and observing target. The function uses the
speeds to back-calculate distances from this point to the starting
entry in speeds. These distances and the attacker_diameter
are used to calculate a screen diameter for the simulation in each frame.
Obviously, the length of speeds and the frequency determines the
total length of the resulting animation.
Required inputs include the intended frame rate at which the subsequent animation will be played, and distance from the screen at which the observing specimen will be located. These details are important in experiments where you want to precisely determine at what time, perceived distance, or perceived velocity of an attack an escape response occurs. Note: if the specimen is closer or further away than the specified screen distance, the animation will be perceived as a different distance and a different velocity.
If you need to create a looming animation simply to elicit a response, and
are not concerned with the precise details, see
diameter_model.
Inputs should be in cm, speeds in cm/s, and frame rate in
Hz or Frames per Second.
## Create a speeds vector
## Here we create a vector of speeds of an attacker steadily accelerating from
## nearly stationary (1 cm/s) to 500 cm/s by 1 cm/s per frame
x <- seq(1, 500, 1)
## Use the variable speed vector to create the model
loom_model <- variable_speed_model(x, screen_distance = 20, frame_rate = 60, attacker_diameter = 50)