Problem with multiple displays (Win7, 64bit, PTB Version 3.0.12)

1
Hello everyone,

I am using Psychtoolbox for a speech in noise task, in which the participants type in what they think they heard, and then an experimenter judges how many words they got correct (this is to prevent orthographic errors to influence the score). As this indicates, I care very little about timing accuracy here, so the issues
that multiple display may have for timing accuracy (as indicated in the various manual files) is not really an issue for me.

I am trying to do this with multiple displays, having two full-screen windows, one for the participant, and one for the experimenter. However, when I am flipping one of the windows, the other one goes black.

I also use the high-level function "Ask" to get the response from the participant, and whether this shows or not in fact depends on the order of initialization of the two windows. If the participant window was the last one to do something, this is visible, if the experimenter window was the last one to show something, the "Ask" command does nothing visible (but still works for the flow).

Has anybody had similar issues?
Currently, the multiple display Wiki site (http://docs.psychtoolbox.org/MultiDisplaySetups) is not much of a help.


Kind regards,
Holger Mitterer
University of Malta

PS: I am attaching the script, note however, that it makes use of a 300+ MB sound file repository, which I better not upload....


2
Dear Andreas,

thanks for the help. We just had a visitor who also uses PTB, and he suggested using one
"large screen" (so that both screens are one window).
Takes a bit more of coding to get the positions reasonable, but it worked.

Greetings from (an already unseasonably warm) Malta
Holger

From: "Andreas Widmann widmann@... [PSYCHTOOLBOX]" <PSYCHTOOLBOX@yahoogroups.com>
To: PSYCHTOOLBOX@yahoogroups.com
Cc: holgermitterer@...
Sent: Tuesday, 12 April 2016, 9:39
Subject: Re: [PSYCHTOOLBOX] Problem with multiple displays (Win7, 64bit, PTB Version 3.0.12)

 
Dear Holger,

sorry, not really a solution but possibly an alternative approach: We have a similar (but Linux) setup in our labs. However, instead of printing data for the experimenter in a second PTB window we output (fprintf) relevant information on the MATLAB command window on the experimenter display. As far as I understood your script you only output text (to this window). What was the reason you prefer output to a PTB win?

Best regards to Malta :)
Andreas

> Am 11.04.2016 um 11:24 schrieb holgermitterer@... [PSYCHTOOLBOX] <PSYCHTOOLBOX@yahoogroups.com>:
>
> [Attachment(s) from holgermitterer@... included below]
>
> Hello everyone,
>
> I am using Psychtoolbox for a speech in noise task, in which the participants type in what they think they heard, and then an experimenter judges how many words they got correct (this is to prevent orthographic errors to influence the score). As this indicates, I care very little about timing accuracy here, so the issues
> that multiple display may have for timing accuracy (as indicated in the various manual files) is not really an issue for me.
>
> I am trying to do this with multiple displays, having two full-screen windows, one for the participant, and one for the experimenter. However, when I am flipping one of the windows, the other one goes black.
>
> I also use the high-level function "Ask" to get the response from the participant, and whether this shows or not in fact depends on the order of initialization of the two windows. If the participant window was the last one to do something, this is visible, if the experimenter window was the last one to show something, the "Ask" command does nothing visible (but still works for the flow).
>
> Has anybody had similar issues?
> Currently, the multiple display Wiki site (http://docs.psychtoolbox.org/MultiDisplaySetups) is not much of a help.
>
>
> Kind regards,
> Holger Mitterer
> University of Malta
>
> PS: I am attaching the script, note however, that it makes use of a 300+ MB sound file repository, which I better not upload....
>
>
>
>
>



3
Agreed, that would also work, if Holger would need precise/trustworthy stimulation timing. *Only* connecting the Matrox to one video output of the graphics card, then driving two monitors from it, with stereoMode 4 so the "left eye" stim goes to the experimenter display, the "right eye" stim to the subject.

However, as Holger doesn't need precise/trustworthy visual timing or tear-free visual animations he can just use his current hack.

-mario

XX---In PSYCHTOOLBOX@yahoogroups.com, <vladvil@...> wrote :

Dear Holger,

Check out the "DualHead2Go" by Matrox:
In situation like yours it would be a dead simple solution. The device is connected to your graphics card and represents itself to the computer as a single wide monitor (e.g.3840x1200). The output of the device goes to two physical monitors, each showing half of a single logical desktop (window).
For all practical purposes the computer is dealing with a single wide monitor. Therefore there are no issues synchronizing the two physical monitors.
I've been using the device for 2-monitor optical stereo setup and had no problems with it.
The "DualHead2Go" has several modifications for PC and Mac with DVI, VGA, Display Port and Mini Display port connections.

Vladimir

Vladimir Vildavski, Ph.D.
Department of Psychology
Stanford University
Stanford, CA 94305



Hello everyone,
I am using Psychtoolbox for a speech in noise task, in which the participants type in what they think they heard, and then an experimenter judges how many words they got correct (this is to prevent orthographic errors to influence the score). As this indicates, I care very little about timing accuracy here, so the issues
that multiple display may have for timing accuracy (as indicated in the various manual files) is not really an issue for me.
I am trying to do this with multiple displays, having two full-screen windows, one for the participant, and one for the experimenter. However, when I am flipping one of the windows, the other one goes black.
I also use the high-level function "Ask" to get the response from the participant, and whether this shows or not in fact depends on the order of initialization of the two windows. If the participant window was the last one to do something, this is visible, if the experimenter window was the last one to show something, the "Ask" command does nothing visible (but still works for the flow).
Has anybody had similar issues?
Currently, the multiple display Wiki site (http://docs.psychtoolbox.org/MultiDisplaySetups) is not much of a help.

Kind regards,
Holger Mitterer
University of Malta
PS: I am attaching the script, note however, that it makes use of a 300+ MB sound file repository, which I better not upload....