WORKSHOP WEEK 6 MMST 12016

SOUND DESIGN 1

INTRODUCTION

In this and the next two workshops, you are stepped through the design and creation of a sound to suit a hypothetical scenario.

In working through this task, you are introduced to the following technical concepts and software packages:

• Synthesis (extension) Orangator
• Digital Audio Editing (extension) SoundForge
• MIDI Anvil Studio

SCENARIO

You are a member of a community radio station. As an upcoming fundraising event, the station is holding a festival of 1950s B-grade science fiction films.

The station will be running an advertisement to promote this event, and you have been asked, at sort notice, to prepare a sound track to accompany the advertisement. You will have to do this using the resources you have on your computer, and will not have time to record any original audio.

You need to produce sounds that evoke this type of film, and then put them together into a soundtrack of about one minute in duration. Since the soundtrack may need to be altered to fit around a voiceover, you want to maintain flexibility. For that reason you do not want to design a single wave file, but will be creating a composition in Multiquence that will allow the file to be adjusted as required.

DESIGN

1. RESOURCES AND CONSTRAINTS.

2. DESIGN CONSIDERATIONS - FORM
You have been asked to create audio that will function as the background for a voice-over advertising the film. You do not know exactly what the voice-over will be, nor how long it will be. For that reason the final piece you produce must be open to manipulation and fine tuning.

3. DESIGN CONSIDERATIONS – PROCESS
In order to create a suitable piece, you must go through the following steps
a. Decide on appropriate sounds that will evoke old fashioned science fiction
b. Create these sounds on your computer
c. put the sounds together into a final, customizable audio product.

A: decide on appropriate sounds:

You need to decide on appropriate audio building blocks. Think for a moment about the sort of sounds one might associate with black and white B-grade science fiction movies. Then scroll down to see the sounds I have chosen to use as a basis for the audio piece.

When I think of old B-Grade Science Fiction films, I think of huge computers full of flashing lights that beep and bleep. I also think of monsters and aliens. You may have come up with different ideas. For the purposes of this example, we will be sticking with computers and monsters..

BEFORE YOU START:

Create a folder called "sci-fi_project" in your working directory for this project.

Create the following subfolders:

• wav
• mp3
• synth
• midi

1. CREATING A FIFTIES SCI-FI COMPUTER SOUND.

The sound I have in mind is made from a series of bleeps that cycle quite quickly and change tone randomly. I will create this sound using Orangator. I will not step through the process by which this sound was created, rather, I suggest you download the Orangator settings and investigate them. I will point out the features and explain how the sound "works"

Download the Orangator settings. The settings file is called "computernoise1.orp". Save this file in the "synth" subdirectory.

To use this file, open Orangator, and select "open preset" from the "presets" menu, and select "computernoise1.orp" from your "synth" folder.

Step 1: Length

To create a useable sample, I have chosen a length of approximately four seconds. To do this, I chose "custom" in the Length field, and moved the slider until I reached a length of approximately four seconds.

Step 2: oscillator.

Since I am creating a sound made up of beeps and bleeps, the "pure" tones of a sine wave seem appropriate. In the image below, note that I have made oscillator 1 active and selected "sin".

Step 3. LFO

I want the tone to change rapidly. To make the tone change, I need to apply an envelope or low frequency oscillator to the sine wave tone. Note that I have chosen LFO1 to modulate the frequency of the oscillator. I do not want the tone to ramp up and down, but to jump from one frequency to another. I have therefore chosen a randomised square wave to modulate the signal. Note that LFO2 has been selected on the view window, so the waveform being displayed is that of the LFO modulating the frequency. The tone is being raised and lowered to random levels twice a second (as the frequency has been set to 2.0 Hz)

At this point, you should play with these setting - try raising and lowering the frequency. Try applying a sin wave instead of a random signal for LFO2. Go back to the "Oscillators" tab and try different waveforms for Oscillator 1.Using a square wave provides a different yet equally effective sound.

We now have the basis of the first sound. The next step is to capture this sound as a wave file so it can be manipulated.

Step 4: capturing the sound.

Before you can capture this sound using SoundForge or GoldWave, you must check to see that your computer's volume settings are set correctly. The settings you need will be specific to your system. In general, you will need to ensure that "Wave" is selected in your computer's volume control, and that only "stereo out" or "what u hear" is selected in your recording volume control.

Open your sound editing package (SoundForge/SoundForge XP or Goldwave).

Open a new file

Select "record" - the location of this command will vary depending on which package you are using.

In SoundForge, you will see something like this:

Switch to Orangator, and play the sound. Now switch back to your audio editor and monitor the sound levels by watching the green bars, which represent the volume of the left and right channels. You will need to adjust your volume and/or your recording volume using your system's volume controls. The bars should reach a maximum near the top of the range, but should not enter the red area. In SoundForge, when the volume rises above what can be recorded without distortion, red "clip" marks appear on these bars.

Once you have set your volume and recording volumes such that the levels are optimised (i.e. the loudest part of the repeating sound does not send the level indicators into the red) press record. When you have recorded a little more than one full loop, press stop on the record window. You may need to zoom in to see what the sound looks like:

In this magnified view we can see that the sound is made up of a series of relatively pure tones at different frequencies and amplitudes.

Step five: creating a loopable sample

Since you want this sound to be versatile, it is a good idea to select a portion of the sound that will work well as a loop. To do this in Soundforge, press the "loop play" button, and select part of your wave by clicking and dragging the mouse. When you have selected a slice that sounds good as a loop, trim the file so that only this section remains. In SoundForge, this is done by selecting "trim/crop" from the "edit" menu.

Then save the sound. I saved the sound as computernoise1.wav

Note

If you can't get a recording you need to set up your record volume settings, this is different depending on the windows version you are using. You need to check that your volume is set to record not playback and that either of the two options are selected Wav Out Mix Balance or Wave Stereo.

You then need to adjust the slider in the panel and the volume control in Orangator until you get the required levels on the bar graph in Sound Forge.

If this is beyond you then go on, you can down load the file.

Because of your settings if you can't record to Sound Forge, download the following file.
To do this right click and save target as to the directory you have created.

By adjusting the Orangator settings, I produced a slightly different computer noise, which I captured and saved as a longer sample, called computernoise2.wav. If you want to explore these changes, download the modified presets file:

I played this sound in Orangator for approximately fifteen seconds. Before I could select a portion of this that would successfully loop, I had to fix one problem that arose: At the end of each loop of the sound as it was played by Orangator was a small period of silence, which was undesirable. I fixed this by highlighting the silence as shown below:

and then hitting the "delete"key. This process was repeated to eliminate the second area of silence. A similar procedure was carried out on computernoise1.wav

CAUTION One way I may have approached this task would be to zoom in and carefully select the area to be deleted, ensuring that there was a good match between the regions directly before and after the silent area. Since SoundForge allows multiple undos, mistakes can easily be rolled back. If your audio software does not allow such undoing, it is a good idea to save an original version, so that you can repeat the process if required.

The next step is to highlight as large as possible an area, and play the selection as a loop. Experiment with setting your beginning and end point until you have a sound that works, with now audible popping or jarring discontinuity at the end of each loop. Then trim the sound, saving your selection.

Have a listen to the result. Play it as a loop.

Sound 2: monsters

The second sound I have chosen to develop for this project is that of an alien monster. In order to do this, I am going to use one of the MIDI sound effects built in to my computer's sound hardware. To access these sounds, I am going to use Anvil Studio. Anvil is a powerful package with many more features than we need to cover for this course. In order to give you an understanding of this package and help you gain a practical grasp of MIDI, it will be necessary to detour here to cover ANvil Studio in a little more depth than is necessary for this example or your assignment work.

Exploring Anvil Studio and MIDI

Step 1: selecting a MIDI sound effect.

First, open Anvil Studio. By default, anvil opens in composer mode, You can see that Anvil is designed for composing music using MIDI. We are not focusing on such production in this course. Rather, we are considering MIDI as a feature that provides us with a ready made bank of sounds. For the most part, these are musical instruments, but there are also many useful sound effects. Bear in mind that the software makes no distinction between musical instruments and sound effects.To Anvil, they are all "instruments". When you first open Anvil you will see that by default it contains one MIDI track, and it is set to "Acoustic Grand", the first instrument on the list. Look at the range of instruments by clicking on "acoustic grand."

While we are using Anvil Studio to access MIDI sounds for the purposes of sound design, it is most usually used as music composition tool. If you have access to one, you can play a MIDI keyboard and Anvil will record a score for playback. We will be using the on-screen musical keyboard for input. To use this keyboard, you select a note duration (A) and then click on a key to place the note on the musical staff.

You do not need to know how to play or read music to access these sounds, and musical notation is outside the scope of this course. For the benefit of those without basic musical knowledge I am going to walk through some of the compositional abilities of Anvil. It is important that you gain a conceptual understanding of what Anvil and similar software does.

You will notice that this note (a crotchet) is selected by default. In the simplest of terms, a crotchet represents a beat. The duration of a beat is determined by the tempo of the score. We can view (or adjust) this by selecting "metronome/tempo settings" from the view menu. By default, Anvil has a tempo of 80 beats per minute. Each crotchet therefore represents an eightieth of a minute, or 60/80 seconds. In this case, a crotchet represents 0.75 seconds.

The time values of the available notes at the default setting of 80 beats per minute (bpm) can be represented as follows.

I want to input a scale of crotchets (single beat notes) using the inbuilt keyboard. Make sure you have the crotchet selected, and then click once on each of the keys indicated below, in the order I have indicated. (Note the yellow dot representing the note middle C. You need to begin two white keys to the left of this):

The notes you have entered (which are A, B, C, D, E, F and G) are shown on the staff as follows:

When you have done this, press play:

You should hear the sound of a synthesised piano playing he notes A, B, C, D, E, F and G at 80 beats per minute. The exact nature of the sound will depend on the quality and model of sound hardware on you system. On my system, the sound was:

Explore some of the other sounds. Add a second track by selecting "Create -> Instrument Track" from the Track menu. Once you have create a new track, it will be marked as active by a green square on the left of the track listings. All keyboard input will now go to the new track, and the notes on the staff will represent the notes on the new track. In other words, the green square represents the active track:

When you want to go back to working with the first track, you would click where indicated by the red cross:

For now, work with the new track (track two). The track listing shows that track two is set to Acoustic Grand Piano by default. Change this by clicking on "Acoustic Grand" in track two (i.e. as indicated by the red cross):

A range of instruments pops up. Select one. For now, pick something that you recognise as an installment, such as Church Organ or Viola.

Now, direct your attention to the onscreen keyboard again. Make sure that the crotchet is selected. The first track (the piano) consists of seven crotchets, that is seven notes of 0.75 seconds. In order that the sounds do not overlap, press "add rest" seven times, then press the notes A,B,C,D,E,F and G as before. Adding seven rests means that the second track will be silent for seven periods of 0.75 seconds, while the piano notes of track one are playing.

Now, press play. I chose Accordion as the instrument for the second track.

Experiment by changing the instruments in your two tracks.

I created a MIDI file with several tracks. After running through several instruments, I chose the Bird Tweet, as this is the sound that will be used in the sound design example. The bird tweet begins with the same seven notes, but then notes from further up and down the keyboard are aded, including some longer notes.

Download the file "abcdefbirds.mid"and save it in your "midi" folder.

Right Click and select Save Target as to download the file in Anvil Studio, and press play.

On my system, it sounds like this.

A caution about MIDI.

MIDI files are very small in size, as they simply point to a standard set of sounds in your computer system. The General MIDI format allows for quite powerful control over the volume and frequency envelopes, and extensions to MIDI allow many other effects to be added in order to enhance the naturalism of the sound. However, no matter how sophisticate your MIDI file may be, it still plays back the virtual installments available on your system. A system with an expensive sound card will have better quality instruments (MIDI patches) than will a system with onboard sound. This workshop has been prepared on a Fujitsu notebook computer with on board Yamaha audio software. The sounds you hear when you listen to the audio samples on this page may differ to your system - in fact they almost certainly will unless you are using a comparable Yamaha chip set.

Depending on your system, you may have a choice of synthesizer hardware to choose from. For instance, there may be cases where your computer's motherboard has built-in audio hardware, and in addition you have a dedicated soundcard in stalled. Windows XP (and possibly earlier versions) contain system software that emulates the functionality of audio hardware. In Anvil, selecting "Synthesizers" from the "View" menu allows you to choose from among the available options. In my case:

I put together a small MIDI file that plays some notes using the following MIDI patches: Acoustic Grand, Reed Organ, Slap Bass 1, String Ensemble 2, and Baritone Sax. I then recorded the results using both the Yamaha AC-XG synthesiser on board my notebook, and then using the Microsoft Windows software synthesiser. Compare the results:

(Yamaha AC-XG)

(Windows XP softsynth)

You might want to download the MIDI file, open it in Anvil, and see how your sound card compares.
Right Click and select Save Target as to download the file.

That ends the diversion into Anvil Studio. Let's now return to the example that is being worked through in this workshop. Remember, we have created two versions of a computer noise, and need to create the sound of a science fiction monster. We are going to build this basis of this sound using Anvil Studio and the Bird Tweet noise

There are not many instruments that look like they could form the basis of a an alien monster. The most "animal like" is "bird tweet", so select that, and press OK. Your track view should now look like this:

Select the longest note (the minim) and use the keyboard to place a few notes on the staff. Stay down at the low end of the keyboard. My results looked like this:

I played this and recorded it in SoundForge, saving the file as monsterlow.wav.

In order to add more texture and make this sound less like a bird, I created a new track in Anvil, and using the on screen keyboard, placed some higher shorter notes and rests. This took some trial and error. Since the MIDI tweet is a relatively high frequency sound sample, the very high notes on the keyboard did not produce good results. After a few attempts, I was happy with a score that looked like this:

The result, saved as monsterhilo.wav was:

I now have four sounds which will form the basis of the building blocks of the soundscape. You might like to refresh your memory:

computernoise1

computernoise2

monsterlow

monsterhilo

The next workshop will involve taking these sounds and working with them in SoundForge to produce the range of sounds which will be used in the final soundscape.