phonetics courses

Analog-to-Digital Conversion

Digitizing Tape-Recorded Data with PC/Macquirer, Audacity, etc.

For digitization, best to use Simon (Audio Lab, 2101B) for PC, or Faust for Mac (Computational Lab, 2101D). These computers are in isolated rooms so that you can listen over speakers without bothering others. Quimby is also set up as an audio station.

To use Simon, see the instructions in the Audio Lab. Simon has PCQuirer/Pitchworks, Multispeech, Praat, Audacity, and other audio editors that you can use to capture speech from the analog cassette player.

Direct input of digital recordings from DAT

Simon's and Quimby's sound card allows digital capture as well as analog-to-digital capture. Follow the instructions in the Audio Lab for setting this up. You will need to use Audacity to acquire the digital speech, but if you save it in .wav files it can be analyzed in any program.

Recording with a microphone (that's A/D too!) - back to the Recording page

Digitizing Tape-Recorded Data with CSL

• Kay Elemetrics CSL (Computer Speech Laboratory):

This is the speech analysis program which comes with the external hardware box. The advantage of using the CSL box is that CSL digitizes analog-recorded data with much better quality than any program using the computer's sound card. The disadvantage is that since our old CSL is a DOS program, there is a severe limit on file length.

Speech data digitized with CSL can be saved in CSL's format (for use with CSL or MultiSpeech), or in .wav format (for use with PCQuirer or just about any other program).

Currently the only computer set up with CSL is Jasmine, in the Physiology Lab. However, there is currently no tape recorder there. Use the equipment on the rolling cart in the Perception Lab; ask Henry for help with connections.

You should be able to start CSL by typing "CSL" at the DOS prompt. Then follow these instructions:

• Adjust input level: We do not want to overload input level from the tape deck to computer. Correct adjustment of input level is critical for good sound quality of digital data.
1. Put a tape in the tape deck. You will see the control named "monitor" on the front right top corner. There are two options: "source" and "tape". Select "tape".

2. The analog signal from the tape deck will be output to Channel 1 of CSL. First, adjust "output" level by using the "output" knob on the front side of the tape deck. Second, adjust "input" level by moving the top left knob on the CSL box, which is labeled as "Input Ch1". Adjusting both input and output levels is important to maximize Signal-to-Noise ratio (i.e. the ratio of signal amplitude to quantized noise amplitude). In both cases, try to maximize level but not to exceed limit. Signal overload will be indicated by red color of the Channel 1 input indicator of the CSL box.

• Digitize and save speech input in zip disk
1. Activate Window A of CSL.

2. As soon as you play the tape, press F1 key on the keyboard. Press any key on the keyboard to stop digitization, and stop the tape deck.

3. CSL has only a red cursor. Stay in the A window. Place the cursor and click the mouse to mark the beginning of the portion of the waveform that you want to save. Move the cursor and mark the end of the portion in the same manner.

4. Type "s". All digitized speech files will be automatically suffixed with "NSP". You will see the SAVE SAMPLED DATA box. Type a name of the speech file. Keep the prefix "d:\", which indicates the zip drive (d-drive). Click "OK".

6. Repeat the same process for the digitization of more data.

[Back to Top] [Back to Facilities Home Page]