Vocal RT without SRBox

[Oscar Ferrante]

Hi everyone!
I'm an Italian student at Università "G. d'Annunzio" of Chieti-Pescara.
For my thesis, a dual-task experiment, and I need to register the vocal reaction time through a microphone.
There is the possibility to do this without the SRBox? If yes, how?
Thanks to everybody will answer!

[Deweber, Derick D. (HSC)]

Please unsubscribe me from the group.  Thanks, Derick.

Sent from my iPad

--
You received this message because you are subscribed to the Google Groups "E-Prime" group.
To unsubscribe from this group and stop receiving emails from it, send an email to e-prime+u...@googlegroups.com.
To post to this group, send email to e-p...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msg/e-prime/-/x9_qTpcihRsJ.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[Cognitology]

Hi,

Please correct me if I’m wrong, anyone, but although there *is* in E-Prime 2 a soundcapture device, with which you can record audio for your experiment, there is no way to use this as an alternative to the SRBox voicekey input.

However, you can still do what you want by using the soundcapture, to save the wave file for each trial (try to save the sound as a trial number or some such so that you know what condition it was in), and then get the reaction time from that. The hard way to do that would be to use any audio editor (I like cool edit, but audacity works), and see when a reaction is given. This is possible but will take a number of hours to work. If you know a little matlab, an alternative would be to do some signal processing and just detect for a threshold, which shouldn’t be too hard (find RT as time(dB > threshold)).

If you only have E-Prime 1, you could just record the whole experiment and have loud sounds as a synchrony marker.

 

I’ve noticed before that any voicekey type of experiment requires roughly the same amount of extra time of processing as (duration of experiment) x (number of subjects), though!

 

Cheers,

Michiel

--

[Oscar Ferrante]

Thank you very much for your answer, Michiel!

I use E-Prime 2 and the method that you mention it's the same that I thought.
I don't know Mat-Lab and neither my professor.

If someone else have other ideas I will glad to read it!

[David McFarlane]

As Michiel said already, for E-Prime to get a
voice RT you need an external voice key device,
which the SRBox provides. But if you do not want
to spend US$450 or more on an SRBox and are handy
with some light electronics, then you could build
your own voice key with readily available
supplies for very little money -- I did this
myself a couple decades ago using plans from a
published paper, and I could look that up again
for anyone interested. You would also have to
hook this up through a digital I/O port (e.g.,
parallel port), and make the appropriate adjustments in E-Prime.

FWIW, the folks building PsychoPy plan to add the
capability to use normal sound input as a voice
key, but I have no idea when they will get around to that.

-- David McFarlane

At 2/26/2013 08:08 AM Tuesday, Oscar Ferrante wrote:

>Thank you very much for your answer, Michiel!
>
>I use E-Prime 2 and the method that you mention it's the same that I thought.
>I don't know Mat-Lab and neither my professor.
>
>If someone else have other ideas I will glad to read it!
>Il giorno 26/feb/2013 13:55, "Cognitology"

><<mailto:msp...@cognitology.eu>msp...@cognitology.eu> ha scritto:

[David McFarlane]

Hmm, an observation at the thread at
https://groups.google.com/d/topic/e-prime/DcKdgNJKAlM/discussion may
dash the hope of using microphone sound recording to measure RT.

-- David McFarlane

[Alessia Coppola]

Hi Oscar,

I know it is now three years since you posted this message, but I wonder wheter and how you managed to solve your problem, because I'm in the same situation now.

I'm an Italian student too, preparing my thesis.

If you or anyone else have new informations to share about how to record vocal responses and RT without SRBox I would really appreciate that.

[David McFarlane]

How handy are you with an Arduino, or similar device
(https://www.arduino.cc/ )? For about US$25 in hardware, plus your
labor, you could make a nice little voice key. I have dabbled with
Arduino a bit myself, and this little gizmo has lots of potential. We
used it for getting tapping inputs from a piezoelectric device for use
with E-Prime. It should handle sound input, and digital output, very
handily, although I have not tried that myself. E-Prime could get
Arduino input either as a virtual Serial device via USB, or you could
wire one of its digital outputs directly to a parallel port or other
E-Prime compatible digital input.

Please let us know if you pursue this.

---------------
David McFarlane
E-Prime training online:
http://psychology.msu.edu/Workshops_Courses/eprime.aspx
Twitter: @EPrimeMaster (https://twitter.com/EPrimeMaster)

[David McFarlane]

As I recall, a SoundIn or SlideSoundIn continues recording sound for the
full duration of its Buffer Size, even if the object terminates earlier
than the Buffer Size (unless the SoundIn gets reset by rerunning it,
etc.). You might also have to set Stop After to No to get that to work,
it's been awhile since I tested this.

To test this, try the following: Put a SoundIn in front of your
succession of Slides. Set the Duration of that SoundIn to 0, and set
the Buffer Size to something suitable, say, 20000 ms. Then run that,
make some sound for 20 s while your Slides play, and see what happens.
Then report back here with the results. If that works then you could
adapt & modify the principles to your experiment.


Beyond that, how will you synchronize your sound recordings to your
stimulus to get valid reaction times? I would not trust E-Prime, or any
other software, to do a good job with that. At the very least, you need
to find some independent way of verifying your reaction time
measurements. Some time ago we tried to use sound recording to measure
RT with PsychoPy -- we carefully tested the timing with a Black Box
Toolkit, and to our great surprise, we found that sound recordings
started a variable time *before* we issued the command to start
recording! We never did figure this out (I have an idea of how this
might happen). In this case we just cheated by playing a tone when the
stimulus started, and then we used that tone in the recording as the
starting time for our RT measurements.

---------------
David McFarlane
Twitter: @EPrimeMaster (https://twitter.com/EPrimeMaster)


On 2017-01-25 6:57 AM, Alessia Coppola wrote:
> Hi David,
> Unfortunatley I'm not so handy with Arduino.
> But my tutor volunteered to get the latencies of the vocal responses
> recorded in eprime using Matlab.
> The problem is that my eprime experiment uses four Slides.
> Using a SoundIn subobject only allows me to record what the subject says
> during the presentation of a single Slide.
> Is there a way to record vocal responses during all the slides presentation?
>
>
>
> Il giorno giovedì 19 gennaio 2017 22:32:53 UTC+1, McFarlane, David ha
> scritto:

[Alessia Coppola]

I put a SoundIn in front of the first slide and set its properties 

as you described. It worked! Thank you so much for you help.

I set the Capture Filename to [AudioFileName], an Attribute 

I created that assign a number to each experimental condition 

(72 in total). In this way the audio files are saved with a name 

like "1.wav" "2.wav" and so on and I know to which condition 

that number matches, even though the selection order is random. 

I also set the Exit List parameter of my Testing List to "After 5 cycles" 

in order to have each subject tested five times on each condition

(360 trials in total). But I notice that in this way when a condition

is presented again during the experiment the new audio file that

is created overwrites the first because they have the same name, 

since the number matching to that condition is the same.

For the same reason I cannot save the audio files of different 

subjects in the same folder, so I should copy a subject's recordings

in a different folder after each performance,in order to not have 

them overwritten by the next subject's audio files.

I could copy and paste the 72 conditions other 4 times in the Testing List, 

keep the Selection order set to Random and set the Exit list parameter

to "After 1 cycle" but it doesn't seem an elegant solution and it doesn't 

allow me to keep all the subjects' files in the same folder anyway.

How could I figure this out?

As regard as the synchronization problem we still don't know how to solve it.

I'll fill you in. Thank you for sharing your experience.

[David McFarlane]

A few thoughts ...

First, remember that you may put together multiple attribute references,
plus arbitrary strings, to make longer strings. E.g., suppose [Subject]
resolves to "3", and [AudioFileName] resolves to "2.wav". Then the
reference "Subj#[Subject]-[AudioFileName]" would resolve to "Subj#3-2.wav".

Next, note that EP automatically produces ".Sample" and ".Cycle"
attributes for any running List. So, supposing that your List is named
"TrialList", then you might use something like the following for your
Capture Filename:

[Subject]-[TrialList.Cycle]-[AudioFileName]

or, to be a little more complete,

[ExperimentName]-[Subject]-[Session]-[TrialList.Cycle]-[AudioFileName]


Finally, if you want to put the subjects' data into different
subfolders, you might use something like

[Subject]/[TrialList.Cycle]-[AudioFileName]

(note the forward slash after "[Subject]"). (In fact, you could do
something like this out in the Data File tab of the Experiment Object
properties, see http://www.pstnet.com/support/kb.asp?TopicID=4335 ).
BUT BEWARE -- EP will *not* automatically create a folder if it does not
exist, so you must make sure that the folder exists before you start the
session! Probably best to just give each file a unique name.

---------------
David McFarlane
Twitter: @EPrimeMaster (https://twitter.com/EPrimeMaster)

[David McFarlane]

I'm glad you found an SRBox, and that you started by testing it with the
SRBox example from PST. Did you at least get the buttons and the lights
to work? That would verify that you at least got the SRBox to
communicate with E-Prime. Then you just have to get the voice key to
work for you.

Are you sure that the "Mic Sensitivity" is just an empty hole? On mine,
the hole may look empty from some angles, because nothing sticks out of
it. But if I look straight in, I can see a small blue potentiometer
("trimpot") soldered onto the circuit board about 1 cm inside the box
(if you cannot see that, try opening the box and look inside). That
trimpot has a slot for a small screwdriver, and you need to reach in
with a small screwdriver in order to adjust it. This inaccessibility is
by design, you would not want to change the sensitivity by accident.

My SRBox also came with a microphone with a 3,5 mm monophone (TS, or
tip-sleeve) plug. Note that not all microphones are the same -- see
https://en.wikipedia.org/wiki/Microphone . You must use a dynamic
microphone with the SRBox; most microphones made for computers are
condenser mics, and will not work with the SRBox. Also, many of those
mics, even if they use a 3,5 mm plug, use a TRS (tip-ring-sleeve) plug,
which will fit mechanically into a TS jack but will probably not connect
correctly electrically.

So my best advice is to find a certified dynamic microphone that uses a
3,5mm TS plug, preferably one that came originally with the SRBox. And
that microphone must plug in to the Mic jack on the SRBox, not to the
computer.

-- David McFarlane


On 2017-02-10 5:00 AM, Alessia Coppola wrote:
> Hi David,
> Thank you once again for your useful advices.
> As regard as the vocal RT issue, we manage
> to borrow a PST SR box from another laboratory.
> They gave us a microphone with a 6,3 mm jack too.
> I read on the SR Box manual that the microphone
> connector is for a 3.5 mm jack.
> I tested the SR Box with the SRBOX Sample for
> E-prime downloaded from the pst website.
> The voice key test failed. No response was detected
> either with the microphone they borrow us nor with
> another microphone with a 3,5 mm jack.
> I tought the problem was the fact that I had not
> connected a second microphone to the computer
> to the detect the vocal responses, but it didn't
> worked anyway.
> Finally I noticed that the microphone sensitivity
> button is broken or I guess so. I'm not sure if
> there should be a button but there's an empty hole.
> Could this be the problem?
> Any other precious thoughts to share?
>
>
> Il giorno giovedì 26 gennaio 2017 18:23:30 UTC+1, McFarlane, David ha

[Alessia Coppola]

Hi David,

The microphone problem has been "solved". I found out that the microphone was

actually the one that came originally with the SRBox (audio-technica ATR20) and 

opening the SRBox I noticed that the blue potentiometer was bent back so I could

not see it from the outside. After fixing it and regulating the sensitivity I tested the 

SRBox again and it worked: the lights flashed and the button press and the vocal 

responses were detected.

Unfortunately I was testing it on a computer different from the one i need to use 

for my experiment. I'm using an eye-tracking system that has its own computer.

I plugged the SRBox to it and tested it again. Only the flashing lamps test worked

while the buttons press and the vocal response were not detected.

I changed the SRBox COM port, trying all the port from 1 to 4 but nothing happened.

I don't understand what the problem could be. The joy deriving from finally having 

an SRBox has already vanished.

Thank you for your time.

[David McFarlane]

If the flashing lights test works, that shows that you chose the correct
COM port (actually, you should see that the flashing lights test works
with one COM port, but not with any of the others). I cannot think of
why the buttons and voice key would not work if you have the correct COM
port and the flashing lights test works. At this point you might want
to take this up with PST Support.

Best,
-- David McFarlane