|
|
|
Forum Newbie
Group: Forum Members
Last Login: 10/5/2007 9:58:37 AM
Posts: 2,
Visits: 6
|
|
Hello,
I am running a very simple experiment using E-Prime 2.0, which consists of presenting 42s video clips, separated by 14s fixations between each clip. Unfortunately, after running the experiment, I get a Clock.SystemTimeDrift error message (about 4.5s of drift), and my .edat files show about 700-800ms of OnsetDelay for each video clip (but not for fixations). I have tried "preloading" the videos by presenting them briefly (2s each) at the beginning of the experiment--this seems to reduce the OnsetDelay to about 600-700ms each, but it is still very significant.
The Refresh Clock Test seems to run fine, and does not indicate a timing issue with my system. Any suggestions regarding things I can do to speed up the video presentation would be appreciated (I am using FFDShow codecs and I have DirectX 9.0c installed).
Thanks very much,
Ben
|
|
|
|
|
Forum MVP
Group: Administrators
Last Login: Yesterday @ 9:26:59 AM
Posts: 576,
Visits: 1,254
|
|
| Thank you for your inquiry. The current generation of EP2 is taking additional time to prepare these movie files in an effort to ensure the best compatability amongst variety of machine classes and configurations of lab machines. Future releases will address this issue by providing a more efficient method of preparing the movie files upon load. Because of how movies are rendered, the time required to prepare/buffer movies in EP2 will take a significant amount of time in comparison to other types of static stimuli such as images and sounds. When any OnsetDelay of an object is detected, the two main ways to reduce them is to ensure that the Duration of the object is divisible by the refresh rate and to allow PreRelease of the object prior to the object loading the stimuli. Therefore, I would recommend providing a Duration and PreRelease value of at least 800ms in the object that runs just prior to your movie objects. If the object prior does not have that much duration to release, then you can consider using the same type of setup in the Sample: Pre-loading images without the use of Canvas sample where you load the movie at the top of the procedure. Note that the first object must still absorb the time needed to load the stimuli. -Brandon
|
|
|
|
|
Forum Newbie
Group: Forum Members
Last Login: 10/5/2007 9:58:37 AM
Posts: 2,
Visits: 6
|
|
Excellent! I will try your suggestions. Thank you for your help, Brandon.
~Ben
|
|
|
|
