Swift 228 houses five small sound booths for running perception and production experiments.
All booths are equipped with dual-boot Mac Minis (OSX and Windows 7).
The Basque, Hausa, English and Dakota booths are equipped with head-mounted microphones for recording.
The audio feed from these booths can be routed to a recording station located outside of the booths,
or to a neighboring booth, allowing pairs of subjects to communicate with each other.
- Five Intel Core 2 Duo 2.26 Mac Minis running Mac OSX and Windows 7
- Sony MDRV700 Headphones
- Aphex 454 Headphone Amplifier
- Shure WH20 Dynamic Head Mounted Mic
- Two MOTU Audio Express USB/Firewire Audio Interfaces
- Two Intel Core 2 Duo 2.4 Mac Minis
The computers in each sound booth are set up as dual boot machines running either Windows 7 or Mac OSX operating systems.
To log in, enter your university NetID and password.
A licensed copy of Superlab is activated on each computer in all of the sound booths for running experiments. Leased licenses of Superlab can be loaned out to students on a temporary basis.
The department also maintains three portable licensed copies of Max/MSP on USB iLok keys which may also be loaned out to students who wish to develop more complex perception or production experiments.
- Start up the experiment running software of your choice (Superlab, Max/MSP etc...)
- Check to make sure that the audio source select switch (on the rear of the Aphex Headphone Amp) is pushed ot the right
- Put on headphones
- Play a test sound file
- Set your volume levels on the Aphex Headphone amp.
Audio from headmounted microphones in each booth can be recorded at the recording station.
The "English" and "Dakota" booth mics are fed to the "Inferno" Mac Mini. The "Hausa" and "Basque" booth mics are fed to the "Smokescreen" Mac Mini.
- Dakota -> Inferno, Left Channel
- English -> Inferno, Right Channel
- Hausa -> Smokescreen, Left Channel
- Basque -> Smokescreen, Right Channel
To record audio, log into Smokescreen or Inferno, then launch Audacity. Adjust the appropriate microphone volume control on the front panel of the MOTU Audio Express audio interface.
Press "Record" in Audacity to begin recording. Press "Stop" to stop the recording. Select "File"->"Export"->"Wav FORMAT" to save the recorded audio to as a .wav file.
The sound booths in Swift 228 are set up to run Diapix experiments (described here).
This applies to any experimental design that requires two subjects to communicate with each other through the mic and headphones.
The sound booths are set up in pairs (Dakota and English, Hausa and Basque). A subject in the "Dakota" booth can communicate with the subject in the "English" booth.
The audio from both subjects is fed to the recording station and saved as a stereo .wav file.
Swift 228 Recording Lab (map)
The Swift 228 recording room contains one single-person double walled booth for experiments that require high quality audio recordings. Participants can respond to both video and/or audio stimulus using a keyboard, mouse or USB button box in the booth. Audio is recorded direct-to-disk through an external high quality analog to digital audio interface.
- Two MOTU Ultralite Audio Interfaces
- Two iMacs running OSX 10.8, 3.3GHz Intel Core i3 with 4GB memory
- APC Power backup
- Pair of Yamaha Monitor Speakers
- Shure SM81 Condenser Handheld Microphone
- Behringer Microphone Amplifier
- Presonus headphone amp
- Wireless Keyboard/Mouse/Trackpad
- 17-inch flat screen monitor
- Make sure that both MOTU ULtralite audio interfaces are on. If they are off, hold down the volume button to power up.
- Turn on the Behringer Micorphone Amp
- Wake up or turn on both iMacs (Sideswipe and Bumblebee). Log in with your netID and password.
- On both iMacs, click "System Preferences" and then set the input and output Sound Device to "MOTU ULtralite"
- Close the System Preferences window
- On Sideswipe, Launch Sound Studio, then select "Preferences..." from the menubar
- In the Audio section, select MOTU ULtralite for both "Input" and "Output"
- On the Input side, set "Left" to "2" and "Right" to "3"
- Close the preferences window
- If a group of experimenters are running the same experiment, you will need to make use of the Shared folder on both iMacs
- To access the shared folder, click anywhere on the Desktop, then select "Go" -> "Computer" from the menubar
- Then select "Macintosh HD" -> "Users" -> "Shared"
- Follow steps 1-3 of the Initial Setup
- Turn on the Flat Screen monitor in the booth.
- Turn on both Yamaha monitor speakers.
- Setup your experiment on Bumblebee. Bumblebee's monitor is mirrored inside the booth. Your subject will be able to interact with the experiment inside the booth using the mouse, keyboard or button box.
- Launch "Sound Studio" on Sideswipe.
- Have the subject sit in the booth with the microphone placed as close as possible to the mouth and slightly below the lower lip.
- Have the subject vocalize at normal volume while you adjust the volume level using the "Input Gain" knob on the Behringer amp. You want to obtain the loudest possible signal without clipping (within the green and yellow zones but not in the red).
- If more volume is required, adjust the "Output Gain" knob on the Behringer amp/>
- The left channel input level corresponds to the subject's voice through the mic in the booth. The right channel corresponds to the TriggerWave signal used to mark/segment recordings
- Close the booth door and click the record icon in Sound Studio to begin recording.
- You should be able to hear the subject through the Yamaha monitor speakers.
- Keep track of the incoming volume level, making sure that the levels are not clipping.
- After the recording is complete, select File->Save in "Sound Studio" and save your audio file. Select the "wave" file format and a 16 bit sample size.
HINT software calibrated Jan 2011
- HP4 coupled with Sony MDRV700 headphones
- Left Channel = 8.7mV when noise or speech in HINT is set to 65dB
- Right Channel = 9.0mV when noise or speech in HINT is set to 65dB
When measuring inserts using Radio Shack SLM, readings should be 3-4dB lower. For example, if HINT is generating 65dB, the Radio Shack SLM should read 61-62dB
Cresap 105 (map)
Cresap 105 contains a large sound booth that can accomodate up to two subjects at any one time.
Audio is recorded direct to disk using a multichanel analog to digital audio interface. Each subject can be miked and recorded on individual channels.
A multi monitor setup allows for individual video feeds to each subject.
Cresap 105 is also set up for simultaneous audio and video recording. To maintain synchronization, both the audio and video are recorded to digital tape and then transfered to disk via firewire.
- Apple Mac Pro 2.66GHz Quad Core Intel Xeon with 6GB DDR3 SDRAM (Cylon)
- Apple iMac 2GHz Intel Core 2 Duo with 2GB DDR2 SDRAM (Rosie)
- MOTU Ultralight
- MOTU 8pre
- AKG C420 Headmounted Mics
- Shure SM81 Handheld Mic
- Panasonic PV-GS500 Handheld Camcorder
- DXA-6VU Adapter
- The Mac Pro in Cresap 105 (Cylon) is connected to two flat screen monitors outside the booth which are in turn, mirrored on two flat screen monitors inside the booth.
- Monitor 1 outside the booth is mirrored with the Station 1 monitor inside the booth. Monitor 2 outside the booth is mirrored with the monitor in Station 2 inside the booth.
- When recording audio only, use Station 1.
- Set up your experiment running program (Superlab, Max/MSP etc.) on Cylon. The subject will be able to interact with the experiment using the keyboard, mouse or button box and flat screen monitor at Station 1.
- Place the AKG C420 headmounted mic on the subject and position the mic so that it is close to the mouth and just below the lower lip
- Launch "Sound Studio" on Rosie
- If you are using TriggerWave, Select "New" and then "New File" from the menubar. By default, Sound Studio opens a blank stereo audio file.
- If you are not using TriggerWave, Select "New" and then "New Mono File" from the menubar.
Have the subject vocalize and make sure that the left channel level meter in Sound Studio is within the green/orange range. Avoid recording at levels that are too low (blue range) or too high (red).
The right channel level meter corresponds to the level of the audio markers if you are using TriggerWave.
- Use the MOTU 8Pre gain control knob to increase or decrease the mic volume. You may also need to adjust the position of the mic, bringing it either closer to further away from the subject
- Click record.
- When you are done recording, click "stop".
- Select File->Save from the menubar and save your audio file. Use the "wave" file format and a 16 bit sample size.
High pitched ringing (usually coming from either one of the battery backup units)
- This is usually caused by power outages. The battery units kick in when the power goes out and the built alarm goes off.
- Simply power cycle the battery backups to turn off the alarm. First, make sure the computers are shut down. There should be a power button on both battery backup units. Hold the button down to turn off the unit and then push the button again to start it back up.
Audio output is not playing through the headphones
- Make sure both MOTU interfaces are turned on
- Make sure that the MOTU audio device is the selected input and output device in the System Preferences Panel on Cylon and Rosie.
- Open the Sound Studio Preferences panel (Sound Studio->Preferences)
- Make sure both Input and output devices are set to MOTU 8pre.
- Make sure both left and right channels are set to 1 and 2
- If using Max/MSP, make sure that the MOTU is set as the output audio device.
- Restart the computer.
- Please contact us if you are still unable to resolve any hardware or software related problems.
Connection failed error message
- Undergraduate students accessing either Rosie or Cylon may see this error messsage upon login.
- This is a known issue that we are aware of. For now, simply Click the "OK" button to dismiss the error message.
HINT software calibrated Jan 2011
- Supercal HINT. Left = 4.53. Right = 4.2.
- HP4 Coupled with Sony MDRV700 headphones
- Leftmost output of HP4 = 7.1mV when noise in HINT is set to 65dB
- Leftmost output of HP4 = 7.1mV when speech in HINT is set to 65dB
NNCS Experiment calibrated Jan 2011
- Using cal_test.maxpat with output set to 65
- Left output of Aphex coupled with ER-1 blue = 22.3mV
- Right output of Aphex coupled with ER-1 red = 23.3mV
- When measuring inserts using Radio Shack SLM, readings should be 3-4dB lower
Location: Room 009 (Basement)
Swift 009 has two lab spaces. 009A houses several desktop computers for developing and running Linger experiments.
Swift 009B is set up for running eye tracking experiments.
Swift 009A Equipment List:
- Apple iMac (OS 10.5.8) 2.4 GHz Intel Core 2 Duo
- Razer Lycosa backlit gaming keyboards
Basic instructions for running Linger:
Swift 009B Equipment List:
- Dell Optiplex 760 2.80GHz Intel Core 2 Duo CPU E7400
- running Microsoft Windows XP Professional
- EyeLink 1000 Tower System for Human use, including EyeLink CL High-speed Camera with Dual Illuminator Module, PCI Card, Head Rest with IR Mirror, EyeLink 1000 System Software for Host PC, EyeLink C API, Standard Host PC.
Eye Tracking Experiments:
1. Prep Stimuli text as follows:
- Experiment name on own line
- Followed by empty line
- Followed by:
item_number condition_name stimulus
? question answer(Y/N)
- Click here for an example text file
- Each line must be less than 70 characters
- All sub experiments must have the same number of conditions
- Only alphabetic characters in experiment names
2. Prep Eye-tracking Stimuli:
- - run eyetrack_reading.py
> python eyetrack_reading.py MyStims.txt
- - run pseudorandomizer.py (NO PRACTICE SENTENCES)
> python pseudorandomizer.py output.script
- Troubleshooting: change "filler" category to experimental category
3. Run Eyelink
- Run EyeTrack.0.7.10m
- Open experiment (output.script file, whatever you have named it, generated by eyetrack_reading.py)
- Click "Validate" experiment
- Click "Run" experiment
- Choose "Separate File" for order of stimuli (a file generated by pseudorandomizer.py)
- Name file with ONLY alphanumeric characters (no dashes/underscores/spaces)
- Determine participant's dominant eye
- Trick 1: make small window between hands at arm-length, focus on distant point through window, bring hands back toward face until touching nose (window should land on dominant eye)
- Trick 2: which side of their face do they see their nose more
- Place clean surface (tissue, paper towel) on chin rest
- Click "Calibrate"
- Click "Validate" calibration
- Click "Record output"
- Repeat calibration as often as necessary
The Child Language Development Lab (affectionately known as the ChiLDlab) lives in 102-103 Cresap Laboratory. The lab is decorated in warm, bright orange and green colors, and features large windows with a view onto Lake Michigan. It is designed to be an inviting space where parents can bring their children to participate in language acquisition studies.
To find out more about the lab's mission, its personnel, and current projects, visit the ChiLDlab website.
Our front door:
Our family-friendly reception area:
Our administrative desk:
Pockets for give-aways and organization:
Testing room with pint-sized furniture: