im working on a masking experiment in which i present a probe, blank slide (to create ISI or SOA) and a mask. im using event mode and set prerelease values. my monitor refresh is 85Hz and the timing test from eprime website gives me OK results.

when i pilot the task (about 16 trials), i get 0 duration errors for the probe, ISI slide and mask but HUGE onset delays for these three items (from 12-70ms).

since im piloting this work, i still havent set the probe/mask durations but most likely i will use 12 ms. however, ive used 12/24/36 ms durations and varied prerelease values accordingly. so i've tried 5-10 ms prerelease with item duration of 12/24 ms and as high as 25 ms prerelease for durations of 36 ms. none of these manipulations help me get rid off the OnsetDelay errors although i have no probelm with actual item duration.

please help!

thank you!!

If after finding that the amount of PreRelease you need is longer than what you want the duration of the object to be, then this means that additional consideration is needed such as lower resolution, not using the entire frame 100% and using "pre load without using canvas" sample or SpriteManager.  If those methods result in the same delays, then machine may not be able to display screens that fast.  Using faster hardware is a consideration then.  If you are attempting to perform single refresh displays, entering a value of zero for the duration would be advised under event mode.

Note that to find the value that effectively represents the "on screen duration", you need to subtract the OnsetTime of the next object minus the current one.  This value is not logged automatically so would need to be calculated during analysis or via an InLine.  There are a number of reasons this is not logged by default one of which is that you can't guarantee this value is what you say it is w/o consideration.  For example if this was auto logged and you had a lengthy InLine after or if the object after only wrote to part of the screen not clearing it.

-Brandon