# Timestamp accuracy of PsychPortAudio on poor latency cards

**URL:** <https://psychtoolbox.discourse.group/t/timestamp-accuracy-of-psychportaudio-on-poor-latency-cards/798>\
**Category:** Uncategorized\
**Created:** [October 6, 2008, 1:53pm UTC](https://psychtoolbox.discourse.group/t/timestamp-accuracy-of-psychportaudio-on-poor-latency-cards/798 "2008-10-06T13:53:51Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Allen\_Ingling](https://avatars.discourse-cdn.com/v4/letter/a/a183cd/32.png) [@Allen\_Ingling](https://psychtoolbox.discourse.group/u/Allen_Ingling)\
**Post date:** [October 6, 2008, 1:53pm UTC](https://psychtoolbox.discourse.group/t/timestamp-accuracy-of-psychportaudio-on-poor-latency-cards/798/1 "2008-10-06T13:53:51Z")

</div>

At the moment, until I can either implement the Asio4All patch, or  
address the problem at a hardware level, I'm stuck using what  
InitializePsychSound says is a soundcard with high latency (running  
under Win XP).  
  
As I understand it from the PsychPortAudio('Start'...) help, this  
means the 'when' parameter will be subject to latency and variability.  
 (Fixed latency I can live with, but I'm guessing the variability will  
be bad.)  
  
Setting the waitForStart parameter to 1 means that "the optional  
return argument 'startTime' will contain an estimate of when the first  
audio sample hit the speakers, i.e., the real start time". Can i  
assume that this time stamp will be subject to inaccuracies on my poor  
hardware too? I.e., this timestamp will not be an accurate estimate  
of when that audio hits the speakers?  
  
Thanks.

---

<div class="post-metadata">

**Author:** ![mariokleiner](https://avatars.discourse-cdn.com/v4/letter/m/13edae/32.png) [@mariokleiner](https://psychtoolbox.discourse.group/u/mariokleiner)\
**Post date:** [October 7, 2008, 1:26am UTC](https://psychtoolbox.discourse.group/t/timestamp-accuracy-of-psychportaudio-on-poor-latency-cards/798/2 "2008-10-07T01:26:58Z")

</div>

Yes. All timing will be inaccurate and variable, both sound  
onset timing as requested by 'when' as well as all returned  
timestamps whatsoever.  
  
The only way around this is our ASIO plugin with  
either a native ASIO soundcard or some cheap card  
+ Asio4All & A bit of luck.  
  
Or, of course, you use a real operating system,  
like Linux or MacOS/X ;-)  
  
-mario  
  

> --- In [psychtoolbox@yahoogroups.com](mailto:psychtoolbox@yahoogroups.com), "jm\_parkinson" \<jm\_parkinson@...\> wrote:  
> \>  
> \> At the moment, until I can either implement the Asio4All patch, or  
> \> address the problem at a hardware level, I'm stuck using what  
> \> InitializePsychSound says is a soundcard with high latency (running  
> \> under Win XP).  
> \>   
> \> As I understand it from the PsychPortAudio('Start'...) help, this  
> \> means the 'when' parameter will be subject to latency and variability.  
> \> (Fixed latency I can live with, but I'm guessing the variability will  
> \> be bad.)  
> \>   
> \> Setting the waitForStart parameter to 1 means that "the optional  
> \> return argument 'startTime' will contain an estimate of when the first  
> \> audio sample hit the speakers, i.e., the real start time". Can i  
> \> assume that this time stamp will be subject to inaccuracies on my poor  
> \> hardware too? I.e., this timestamp will not be an accurate estimate  
> \> of when that audio hits the speakers?  
> \>   
> \> Thanks.  
> \>