# timing issue

**URL:** <https://psychtoolbox.discourse.group/t/timing-issue/1485>\
**Category:** Uncategorized\
**Created:** [September 9, 2002, 8:30pm UTC](https://psychtoolbox.discourse.group/t/timing-issue/1485 "2002-09-09T20:30:12Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![aghuman2002](https://avatars.discourse-cdn.com/v4/letter/a/a9adbd/32.png) [@aghuman2002](https://psychtoolbox.discourse.group/u/aghuman2002)\
**Post date:** [September 9, 2002, 8:30pm UTC](https://psychtoolbox.discourse.group/t/timing-issue/1485/1 "2002-09-09T20:30:12Z")

</div>

I was wondering if anyone had seen this timing issue that I am   
having. Basically I am trying to see how long a picture is presented   
on the screen so my basic code looks like:  
  
for i = 1:n  
 SCREEN('CopyWindow', ...);  
 timeonset = GetSecs;   
 screen(window,'waitblanking',waitlength);   
 SCREEN('CopyWindow', ...);  
 timenextpic = GetSecs;  
 data(i) = timenextpic-timeonset;  
end  
  
The problem is that data looks like:  
  
data =   
  
 Columns 1 through 9   
  
 0.0333 0.0287 0.0287 0.0332 0.0332 0.0331   
0.0332 0.0332 0.0287  
  
 Columns 10 through 18   
  
 0.0288 0.0332 0.0334 0.0287 0.0332 0.0287   
0.0334 0.0334 0.0335  
  
 Columns 19 through 21   
  
 0.0287 0.0335 0.0335  
  
While the descrepency isn't all that bad. I often seem to get that 5   
msec descrepency (note: waitlength = 2). This is not a big deal, but   
it is very consistent that most of the lengths are around 33.3 msec   
(as they should be), but a significant portion are around 28.7 msec,   
which is strange. Any clue what is going on here? Is the refresh   
rate of the monitor flipping from 60 to 70 Hz for no apparent   
reason? When I use framerate it says that the frame rate is 60 Hz   
very very consistently.  
  
Thank you,  
  
Avniel

---

<div class="post-metadata">

**Author:** ![aghuman2002](https://avatars.discourse-cdn.com/v4/letter/a/a9adbd/32.png) [@aghuman2002](https://psychtoolbox.discourse.group/u/aghuman2002)\
**Post date:** [September 10, 2002, 11:20pm UTC](https://psychtoolbox.discourse.group/t/timing-issue/1485/2 "2002-09-10T23:20:26Z")

</div>

Hi David  
  
Thank you for the help. I should have mentioned in my miniml example   
that I do indeed do this. So the minimal example is actually:  
  
waitlength = 2;  
for i = 1:n  
 SCREEN(window,'WaitBlanking');  
 SCREEN('CopyWindow', ...);  
 timeonset = GetSecs;  
 SCREEN(window,'WaitBlanking',waitlength);  
 SCREEN('CopyWindow', ...);  
 timenextpic = GetSecs;  
 data(i) = timenextpic-timeonset;  
end  
  
Thanks again  
Avniel  

> --- In psychtoolbox@y..., David Jones \<djones@e...\> wrote:  
> \> Hi Avniel,  
> \>   
> \> I think you might get more consistent timing  
> \> if you first wait for the vertical blanking interval  
> \> and then start your timer.  
> \> When you begin your timer asynchronously with your video display,  
> \> but then call 'waitblanking', it will add a jitter  
> \> of up to one frame time (1/60 = 16.7 msec).  
> \> I am not surprised you are seeing variation in timing up to 5 msec.  
> \>   
> \> Add "SCREEN(window,'WaitBlanking');" at the top of your loop  
> \> and see if it reduces the variability in your timing.  
> \>   
> \> -- David Jones  
> \>   
> \>   
> \>   
> \> On Monday, September 9, 2002, at 04:30 PM, aghuman2002 wrote:  
> \>   
> \> \> I was wondering if anyone had seen this timing issue that I am  
> \> \> having. Basically I am trying to see how long a picture is   
> presented  
> \> \> on the screen so my basic code looks like:  
> \> \>  
> \> \> for i = 1:n  
> \> \> SCREEN('CopyWindow', ...);  
> \> \> timeonset = GetSecs;  
> \> \> screen(window,'waitblanking',waitlength);  
> \> \> SCREEN('CopyWindow', ...);  
> \> \> timenextpic = GetSecs;  
> \> \> data(i) = timenextpic-timeonset;  
> \> \> end  
> \> \>  
> \> \> The problem is that data looks like:  
> \> \>  
> \> \> data =  
> \> \>  
> \> \> Columns 1 through 9  
> \> \>  
> \> \> 0.0333 0.0287 0.0287 0.0332 0.0332 0.0331  
> \> \> 0.0332 0.0332 0.0287  
> \> \>  
> \> \> Columns 10 through 18  
> \> \>  
> \> \> 0.0288 0.0332 0.0334 0.0287 0.0332 0.0287  
> \> \> 0.0334 0.0334 0.0335  
> \> \>  
> \> \> Columns 19 through 21  
> \> \>  
> \> \> 0.0287 0.0335 0.0335  
> \> \>  
> \> \> While the descrepency isn't all that bad. I often seem to get   
> that 5  
> \> \> msec descrepency (note: waitlength = 2). This is not a big deal,   
> but  
> \> \> it is very consistent that most of the lengths are around 33.3   
> msec  
> \> \> (as they should be), but a significant portion are around 28.7   
> msec,  
> \> \> which is strange. Any clue what is going on here? Is the refresh  
> \> \> rate of the monitor flipping from 60 to 70 Hz for no apparent  
> \> \> reason? When I use framerate it says that the frame rate is 60 Hz  
> \> \> very very consistently.  
> \> \>  
> \> \> Thank you,  
> \> \>  
> \> \> Avniel  
> \>   
> \> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -  
> - -   
> \> - -  
> \> Prof. David Jones, Dept of Electrical and Computer Engineering  
> \> McMaster University, 1280 Main Street West, Hamilton, Ontario,   
> CANADA  
> \> email: djones@e... phone: (905) 525-9140 ext.24689  
> \> web: www.ece.mcmaster.ca/~djones fax: (905) 521-2922