# CopyWindow & stimulus appearance on the screen

**URL:** <https://psychtoolbox.discourse.group/t/copywindow-stimulus-appearance-on-the-screen/991>\
**Category:** Uncategorized\
**Created:** [October 29, 2003, 5:23pm UTC](https://psychtoolbox.discourse.group/t/copywindow-stimulus-appearance-on-the-screen/991 "2003-10-29T17:23:56Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![eye1272](https://avatars.discourse-cdn.com/v4/letter/e/da6949/32.png) [@eye1272](https://psychtoolbox.discourse.group/u/eye1272)\
**Post date:** [October 29, 2003, 5:23pm UTC](https://psychtoolbox.discourse.group/t/copywindow-stimulus-appearance-on-the-screen/991/1 "2003-10-29T17:23:56Z")

</div>

Hi everybody,  
  
setup:  
Mac OS 9.2,   
MATLAB 5.2.1.1421,   
PsychToolbox 2.52,   
Eyelink II,   
Eyelink Toolbox  
monitor refresh rate 120 Hz  
screen resolution: 1024x 768 pixel  
I am working on saccade contingent experiments having people read   
single sentences.  
  
I have three types of delay: the eye tracker delay (at least 3 ms),   
the drawing delay, and the display delay.  
this is a short description of the code:  
  
screen(visual.main,`WaitBlanking');  
time\_on = getSecs \* 10000;  
- online saccade detection (reading out the link)  
- computations  
t2 = getSecs \* 10000 – time\_on;  
CopyWindow (I copy a 1024 x 20 pixel rect [0 374 1024 394] from an   
OFF screen to the main screen) using scrCopyQuickly  
t3 = getSecs \* 10000 – time\_on;  
  
As I read in another message, `CopyWindow' runs synchronously meaning   
that a call to Screen(`CopyWindow') does not return until it has   
finished copying.  
  
t3 – t2 is on average 0.68 ms  
I thought that t3 is the time the new rect is appearing on the screen.  
Obviously, this is not the case.  
Can anybody tell me how long it takes between t3 and the actual   
appearance on the screen (t4). How can I measure it? On what does it   
depend? Does it take more than 1 ms (roughly)?  
  
The underlying logic is the following:  
On the vertical axis, the to be copied rect is in the middle of the   
screen. With a refresh rate of 120 Hz (=8.3 ms for a refresh), the   
pointer (or whatever it is) is getting to the critical area after   
around 4 ms.  
Therefore, t3 (or t4?) has to be smaller than 4 ms.  
  
Until now, I relied on t3 (time at which CopyWindow returned) with   
the following logic: If t3 (recomputed as an absolute measure) is   
smaller than the time of saccade end (which I can reliably   
determine), then I conclude that the subject didn't see the display   
change since we're supposed to be visually blind during a saccade. I   
don't want the subject to see the display change.  
I tested myself and according to my analyses I wasn't supposed to see   
any of the display changes, but on quite a few occasions I definitely   
did. Therefore, I think I have to know how long it takes between the   
return of CopyWindow and the actual stimulus appearance on the screen.  
  
Any help is highly appreciated.  
  
thanks a lot  
Antje Nuthmann