Physics 4830 Lab 2: Complex Waveforms

Introduction

A sound produces the sensation of a distinct pitch when it repeats itself at regular intervals. Here, we will examine a few digital samples to verify this concept is correct and when it is applicable. In addition, there is a clear relationship between the physical quantity frequency and the perceived psychological quantity pitch. Frequency alone does not determine pitch. For example, if we increase the intensity of a sinusoidal sound at a fixed frequency, the perceived pitch can change. Often, musical sound is significantly altered during the final recording process (mastering). For example, very intense sounds are reduced in amplitude (compression) and very quiet sounds are increased in amplitude (expansion). In addition, pops and hisses are removed, etc. Digital sound editing makes these alterations or enhancements very easy to do and we will explore the impact of these alterations on the quality of the sound. Finally, the quality of the sound is very sensitive the sampling rate and resolution. We can easily explore this by re-sampling a sound at a lower sampling rate and/or resolution.

Purpose

We will continue our study of complex waveforms using the Soundprobe digital editing software. In Lab 1, you got comfortable using Soundprobe. In Lab 2, you will look at a few waveforms and measure the period/frequency. You will also test your own audible range and digitally enhance a few "wave files" using filtering and compression. The purpose of this lab is convince yourself that complex periodic waveforms have the properties we have been learning about in lecture, such as, pitch perception, frequency and period. In addition, sensitivity to the sampling rate and bit resolution will be examined.

Equipment

Computer, Soundprobe software, Microsoft Word or WordPad software, headphones, function generator, BNC to 1/8" stereo cable, BNC to 1/8" mono cable, 1/8" stereo y-jack.

Questions (to be answered in your lab summary)

1) What do the square, triangle and sine waves look like when hooked directly to the sound card “line in” input. How good of a job does the function generator do? Sketch the waveforms produced by the function generator.

2) Describe the square, triangle and sine waves timbre (the color of the sound) in words. Any ideas why these waves sound the way they do?

3) What did you determine to be your own personal audible hearing range? What are possible/probable problems with this measurement?

4) When the A=220 Hz sine wave increased in intensity, what happened to the perceived pitch? What seems odd about this?

5) What specifically is wrong with the oboe2.wav file? What is it about its sound that seems distorted? Can you fix this distortion (to some extent) using Soundprobe? Did you introduce further distortion that can be perceived?

6) What was the period and frequency of the guitar4.wav file?

7) What was being played in guitar1.wav, guitar2.wav and guitar3.wav? Did you find periodicity, and if so, what was the frequency? We know that a pure tone produces a distinct pitch because it is periodic, but what about more than one tone played simultaneously? How do you suppose we perceive that there are multiple notes being played?

8) Describe the distortion you hear when you re-sample a sound at a lower sampling rate and resolution? What seems to make more of a difference (sampling rate, resolution) and why?

9) Why is it reasonable, not to expect reverb.wav to be periodic?

Procedure

You will need the "wave" files: guitar1.wav, guitar2.wav, guitar3.wav, guitar4.wav, reverb.wav, oboe1.wav, oboe2.wav and bach.wav. These files may be downloaded from the course web site:

http://www.colorado.edu/physics/phys4830/

under “Lecture Notes, Labs and Course Material”. Or, you can get them from the instructor via a CD-ROM. These files might also already be on your computer in C:\scratch.

If you find a particular operation, e.g. "low pass filtering" confusing, try reading the Soundprobe "Help" documentation (click on "Help" and search for the option) to figure it out. Also, don't hesitate to discuss topics with the instructors.

1) Record a square, triangle and sine wave directly from the signal generator. Zoom in on the waveforms and examine their shape to see how well the function generator produces the exact waveforms? To do this, you must click on the speaker in the tool bar, then select “Options” in the “Play Control” dialog box, then “Properties”, then “Recording”, then select “Line-In”. Ask the instructors, if you run into problems.

2) Determine your audible hearing range. Plug the headphones directly into the signal generator output (watch the volume level). Set the signal generator to 1 kHz sine wave. Adjust the volume so that you can hear it well. Use a decent volume, but not too loud. If you would like to try to standardize the level with nearby lab groups, that is fine. The audible range is somewhat sensitive to the intensity; see Figure 6.4 in your text. Now, scan the frequency range and determine your the upper and lower frequency limits of your ability hear sound.

3) Adjust the function generator to produce an A=220Hz sine wave at a quiet volume. Now, increase the volume. Do you perceive the pitch to change? (go flat or sharp?)

4) Download/read in: guitar1.wav, guitar2.wav, guitar3.wav, guitar4.wav.

Plug the headphones back into green jack on the PC sound card. Listen to and examine these waveforms. Look for periodicity and measure the period and determine the frequency. A good way to determine the period is to change the “X-Coordinates” to "Samples". Then you can read off the sample number in the upper right corner. By selecting a beginning time and ending time, you can determine how many samples there are in a period and use the fact that one sample = 1/44,100 seconds. Remember to measure the true period. There may be smaller oscillations that appear to almost repeat at regular intervals (these are large amplitude overtones or other distinct notes being played).

5) Download/read in reverb.wav and look for periodicity in this sound. Do you expect this sound to be periodic?

6) Download/read in oboe1.wav and determine its frequency. Can you figure out what note is being played? We will learn how to do this a little later.

7) Download/read in oboe2.wav and listen to it. What distortion do you hear? This was a sample taken with my Dell Inspiron 7500, which has a lousy sound card, hence, the poor quality sample.

8) Try using some of the features in Soundprobe to artificially enhance or improve the sound of oboe2.wav. Experiment a little by clicking on "Restoration". At the very least try “FFT Frequency Filter”, “3kHz Lowpass”. Increase the cut-off frequency up to around 4kHz o r higher. Apply the filter and listen to see if you think the sound has been enhanced. What distortion (if any) has been added?

9) Go back to guitar4.wav and try re-sampling the wave at lower sampling rates and resolution. Can you hear the distortion? What does it sound like?

10) Download/read in bach.wav and listen to it.

During the recording process (when the album is mastered), often the waveform is altered significantly, especially, in recording popular music. A very common process that is applied is expansion/compression. This reduces the dynamic range of the sound, making low intensity sounds higher intensity (expansion) and higher intensity sounds lower intensity (compression). Classical music is not processed (as much) in this way. In addition, during mastering, the overall amplitude of the waveform is brought up as high as it can be. This is a more recent phenomenon. If you listen to older CDs you will notice that the sampled level is much lower. Why are these things done? The mass market likes compressed sound, and they want loud CDs. Compression has been used in radio, since the early days because it makes a weak signal easier to hear and understand. Some people speculate that this is influencing today's preferences. It is might be more than that. If, for example you watch a movie with audio with a very wide dynamic range, it can be distracting. You want to turn up the quiet parts and turn down the loud parts. I personally don't think this can be attributed to the cultural influence of radio.

11) Now, apply compression to bach.wav. Click on "Enhance", then "Expander", then choose "Preset: Compressor". Now click on "Apply" four times. Now, double the amplitude (as we did in last lab).

Listen to the sound and report on its quality in the lab summary. Also, discuss in your summary the difference in the shape of the waveforms (raw waveform versus digitally altered).