Synthesis Theory

Final exam: December 15, 1998

Course Outline   (fall, 1998)

  • 98/09/03: Introduction
  • 98/09/08: Digital Recording and Mixer (analog)
  • 98/09/10: Sound Designer
  • 98/09: SoundHack
  • 98/10: Csound / Cecilia
  • 98/11: SuperCollider
  • 98/12: Max / MSP
  • Csound.ppc
  • Audio File Format FAQ
  • AudioIFF 1.2.1 Spec
  • Apple AIFF-C Spec (pdf)
    Assignment #8 (Due 98/12/17, all assigments are due 2pm)

    Using MSP, build a usable Reverb patch. Use three to five comb filters in parallel and one to three allpass filters in series. The default interface should not have more than three sliders controlling input gain, output gain, and amount of reverb (wet to dry). Fine tuning of parameters should normally be hidden. Prepare at least four presets (shower stall, medium concert hall, church, cave, etc.)

    Assignment #7 (Due 98/12/10, all assigments are due 2pm)

    Build a graphic equalizer similar to the one in Assignment #6 using Max/MSP. Each slider should be controllable via MIDI mixer (controller).

    Assignment #6 (Due 98/12/03, all assigments are due 2pm)


    Assignment #5 (Due 98/10/29, all assigments are due 2pm)

    Using Csound, reproduce as accurately as possible, the following vibraphone excerpt. Use additive or FM synthesis to simulate the vibraphone sound. The motor should start up slowly. Note also that the B-flat is dampened by a mallet.

    Assignment #4 (Due 98/10/29, all assigments are due 2pm)

    Produce the following with CSOUND:

    1. Repeat Assignment #3 using sawtooth generate by using 10 sine oscillators and a simple envelope using linen
    2. Repeat above using one oscillator using sawtooth wavetable lookup
    3. Repeat above using Trumpet-like sound using the information on the handout.
      1. Use only the steady-state amplitude
      2. Use time-varying dynamic envelope
      3. Use time-varying dynamic envelope and frequency change
    4. Morph (interpolate) between Trumpet to Clarinet sound using the scale in Assignment #3
    Assignment #3 (Due 98/10/15, all assigments are due 2pm)

    Produce the following with CSOUND:


    Use the instrument of the form:

    a1 oscil p4, p5, 1  ; where p4 is the amplitude and 
    ; p5 is the frequency in the score file

    Submit the hardcopy of the orch and score files and place a copy in (314):Faculty:ich:synth assignments#2

    Assignment #2 (Due 98/09/24, all assigments are due 2pm)

    Submit the result of this assignment on a CD (there will be a tutorial on how to make audio CD after class next Tuesday (9/22) given by Meg in C307) and a hardcopy of your comments.

    1. Record as best as possible the phrase "Short-time synthesis" with your own voice. "Short-time" should be spoken and "synthesis" should be sung. (Track 1)
    2. Record the same phrase in the same manner but in a very noisy environment, preferably outside (traffic noise). (Track 2)
    3. Using the Phase Vocoder in SoundHack and the sound of Track 1 do the following. For each, provide two versions: a most natural and intelligible results and another which is to you most "interesting." Document the window type and the number of bins used for both versions and submit a hardcopy (start without using the Resynthesis Gating).
      • Pitch-shift up 2 octaves (Tracks 3-4)
      • Pitch-shift down 2 octaves (Tracks 5-6)
      • Time-expand 3 times (Tracks 7-8)
      • Time-shrink 3 times (Tracks 9-10)
    4. Repeat the last problem (#3) using Track 2 as the source. In addition to documenting the window type and the number of bins, describe the difference in the results compared to results using Track 1. (Tracks 11-18)
    5. Use Track 3 and Pitch-shift down 2 octaves. Try to make it as close as possible to Track 1. Which settings gave you the best results? What is the difference between this result and Track1? What did you lose? (Track 19)
    6. Repeat the last problem (#5) Use Track 9 and Time-expand 3 times. (Track 20)
    7. Take the first 50ms of the "Sh" in "Short..." and time-expand it 100 times. Document your settings use Resynthesis Gating for best results.(Track 21).
      Assignment #1 (Due 98/09/15, all assigments are due 2pm)

      Record with Sound Designer 2-3 sec. of your voice (e.g. a word). Create two soundfiles based on the above.
      The firs file (mono) should countain: the original, normalized version, reversed (unless otherwise noted, use the normalized version as the source for the following), reduce gain by 50%, change EQ using Parmetric EQ, change EQ using Graphic EQ, Compress/Limit, Expand (the compressed version), Noise Gate (the expanded version), Lengthen by factor of two without changing the pitch, Shorten by factor of two, and shift up by 120 cents. Each section should be preceded by a short verbal explanation of what follows.
      The second file (stereo) should be a 30-60 second composition using only the original sound of your voice.
      Submit a Zip disk or a CD.

      Ichiro Fujinaga 1998.12.14