[solved] Record audio response with variable time span

solbero

Hi.

I was hoping someone here on the forum might have the knowledge and time to help me with a problem of mine.

I have an experiment where participants will be listening to some audio stimuli and afterwards describe them verbally. I want to record their response, but the recording phase will not have a set time span. Essentially, when the participant feel they have said what they want to say, they will click a button and progress to the next audio stimulus and recording phase.

Is there any way to accomplish such a thing in OpenSesame? I looked at the inline_script described in http://forum.cogsci.nl/index.php?p=/discussion/107/solved-recording-speech-response/p1, but it only allows for a fixed recording time. The audio plug-in essentially does what I want to accomplish, but it relies on Pymedia. Pymedia is not included in the Ubuntu 13.10 repositories, and I tried compiling Pymedia from source, but ran into all sorts of problems. Mainly that it required a lot of old audio libraries and gcc-3.4.

In the builder I image it would look something like this.

• Sequence.

  • Sampler (plays an audio file).
  • Start audio recording.
  • Feedback (wait for keypress)
  • Stop audio recording and write to file.

• Next sequence.

Any help is much appreciated

Njord

sebastiaan

Hi Njord,

A little bit down in the same discussion that you're referring to, you will find a little code snippet that you can use to stop recording when a key is pressed. It sounds like this is exactly what you're looking for:

• http://forum.cogsci.nl/index.php?p=/discussion/comment/611#Comment_611

Cheers!
Sebastiaan

solbero

Thank you very much for your help, Sebastiaan.

I got the script working and it is recording .wav files without a hitch, but I ran into another problem.

In the folder in which my OpenSesame experiment is saved, I want to create a sub-folder named after the "Participant ID". All the recorded sound files from one participant will then be saved in this folder.

In the script, I have these statements accomplishing this task:

SUBDIR = "./%d/" % self.get('subject_nr')
os.makedirs(SUBDIR)
WAVE_OUTPUT_FILENAME = "%d_%d.wav" % (self.get('subject_nr'),
self.get('count_inline_script'))
wf = wave.open( "./" + SUBDIR + WAVE_OUTPUT_FILENAME, 'wb')

This works well, except that the sub-directory is created in my $HOME folder on Ubuntu, and not in the folder where the experiment resides.

I know that this could easily be solved by specifying absolute paths, but I am going to move the experiment to a lab computer when it is finished. Therefore I want to use variable paths, so that I don't have to go back and change all the paths.

Do you know, or somebody else, how this can be accomplished?

• Njord

lvanderlinden

Hi Njord,

You could determine the path of the directory in which your OpenSesame is placed by using the following piece of Python inline code:

currentPath = self.get('experiment_path')

Does that help?

Cheers,

Lotje

solbero

lvanderlinden, that was just the piece of code I needed.

solbero

I though I would share my finished inline script so that others that might have a similar problem can use and modify it to their needs.

I am using this script to collect open-ended responses to a stimulus. When the participant is done giving their response and press the space bar, the experiment continues. But I also wanted to insert a time out, so that if one participant doesn't press the key, falls asleep, is uncooperative, the experiment will not halt at this part.

It has to be inserted into an inline_script to work.

""" Record a pre-defined amout of audio  and save to a WAVE file. """

import pyaudio
import wave
from openexp.keyboard import keyboard

# Name of sub-directory where WAVE files are placed
current_experiment_path = self.get('experiment_path')
subdir_recording = '/recording/'

# Variables for Pyaudio
chunk = 1024
format = pyaudio.paInt16
channels = 1
rate = 44100

# Set variable for the labelling of the recorded WAVE file.
wave_output_filename = '%s.wav' % self.get('subject_nr')

#Audio recording stuff
p = pyaudio.PyAudio()

stream = p.open(format = format,
                channels = channels,
                rate = rate,
                input = True,
                frames_per_buffer = chunk)

# Import keyboard and set accepted key
recorder_keyboard = keyboard(exp, keylist=['space'])

# Determine the timeout in ms
timeout = 60000

# Create an empty list for audio recording
frames = []

# Determine the timestamp of the start of the response interval
start_time = self.time()
current_time = self.time()

# Record audio until space is pressed
while (current_time - start_time) < timeout:

    # Record data audio data
    data = stream.read(chunk)
    # Add the data to a buffer (a list of chunks)
    frames.append(data)

    # Get new timestamp
    current_time = self.time()

    # Check if space key has been pressed
    key = recorder_keyboard.get_key(timeout=0)[0]

    # If space is pressed, advance to next part of the experiment
    if key == 'space':
        # Record data audio data
        data = stream.read(chunk)
        # Add the data to a buffer (a list of chunks)
        frames.append(data)
        break

else:
    # Record data audio data
    data = stream.read(chunk)
    # Add the data to a buffer (a list of chunks)
    frames.append(data)

# Close the audio recording stream
stream.close()
p.terminate()

# write data to WAVE file
wf = wave.open(current_experiment_path + subdir_recording + wave_output_filename, 'wb')
wf.setnchannels(channels)
wf.setsampwidth(p.get_sample_size(format))
wf.setframerate(rate)
wf.writeframes(''.join(frames))
wf.close()