User`s guide

E-Prime User’s Guide
Chapter 3: Critical Timing
Page 99
The durations were simply specified as 20, 60 and 100ms without regard for any potential
influences (e.g., refresh rate). The gray areas of the table show the actual length of time the
stimuli were displayed. Notice that the actual durations are integer multiples of the respective
refresh periods (rounded up) so that they accommodate the specified duration (i.e., 20, 60 or
100ms). For example, for a refresh rate of 60.15Hz, the refresh period would be 16.63ms. A
20ms display would result in two refreshes (2x16.63ms) and hence be presented for 33.26ms.
This is an error of 66% greater presentation time (33.26ms versus 20ms) than was intended.
When the same E-Prime experiment was moved across three computers, each having a different
display card, the target duration of 20ms yielded an actual duration ranging from 23.52 to
33.26ms; a 41% variation.
The important concepts to take away from the table are the following:
• Brief displays (i.e., under 100ms) must consider the refresh rate, as its influence can
significantly alter the observed duration of a stimulus.
• The refresh rate may vary significantly from one computer to another.
• Display durations should be set as a multiple of the determined refresh rate for the
computer in question.
E-Prime provides methods to report the true durations of displays, and methods to assess the
actual refresh rate. If millisecond precision is required, the display parameters will need to be
adjusted to match the constraints of the display cards. The video hardware will effectively always
round up the display time to the next refresh cycle (i.e., you will always get what you specify plus
up to potentially one additional refresh). Only by considering the constraints of the hardware
when setting display times can users optimize the settings to deliver consistent display durations.
In addition, the hardware and display times must be measured in order to be able to report what
happened in an experiment accurately. To present visual stimuli accurately, the durations must
be set based on the refresh cycle of the run-time display hardware. A good rule of thumb is to
set the stimulus duration to 10ms below the expected total duration of all refresh cycles
desired for the stimulus
10
. The display process may be thought of as the following sequence:
1. Display the first stimulus.
2. Wait for the specified duration.
3. Monitor the hardware for indication that the next refresh cycle has begun (vertical blanking).
4. Display the next stimulus.
The equation to use for determining what stimulus duration to specify in E-Prime is as follows:
Stimulus Duration to Specify = (Refresh Duration ms/cycle * Number of cycles) - 10ms
For instance, to display something for 3 refresh cycles on a monitor using a 60.15Hz refresh rate
(a 16.63ms refresh duration), we calculate the duration as follows:
10
Since the visual duration is always rounded up to the next refresh, the display duration must be specified
as some portion of one refresh below the targeted refresh duration plus the expected measurement error for
when to look for the vertical blanking signal indicating the time of a refresh. Expect a measurement error of
+/- 1ms plus operating system delays if they occur. If the refresh rate is between 60-75Hz, the refresh
duration is between 16.67-13.33ms respectively. If your duration is 10ms less than the targeted refresh
time, the program will start to look for the refresh 3.33-6.67ms after the previous refresh has occurred which
is more than the expected 1ms timing delay. Note, if the refresh rate is above 75Hz, we recommend setting
the duration to half of a refresh duration below the targeted time (e.g., for a 100Hz monitor, set it to 10ms/2 =
5ms below the targeted time). For the majority of monitors, the 10ms less rule will work well and is easy to
remember and check.