HOME  â€¢  SYLLABUS/SCHEDULE  •  ASSIGNMENTS  â€¢   COURSE RESOURCES   •   YOUR PAGES   •   YOUR FEEDBACK

Creating Sound Textures

Choosing the Right Tool for the Right Texture

So, by now you've messed around a little with Audacity, and you've done some very basic audio editing (cut, paste, copy; amplify, normalize, fade-in, fade-out; change pitch/speed/tempo). 

I like to take very small, sustained fragments of interesting sounds, expand them, give them nice hairpin shapes (fade in, sustain, fade out), and layer them together. 

First, start in audacity by selecting between 3 and 5 seconds of an interesting sound file. Select this segment, and select 'Export Selection  . . . " , to create a new file (save as an .aiff file). Or use this modified vocal sample:   chadChoirRemixSweetSpot.mp3  Whatever you choose, open the file in SoundHack.

Now, SoundHack is a program that doesn't really have an interface - - at least not one that's designed for a user-friendly experience. This is, however, part of the enduring beauty of SoundHack: you simply open files and apply these weird filters to the sounds. When you open a file, a little window opens on the left, and you see a player bar at the bottom of your screen. Select the file, then hit the 'space' bar and the sound file will play or stop. I don't really even use the player bar - its sound quality is not as good as, say, playing the sound through Audacity.

So, you've opened your sound in SoundHack (go SoundHack>File>Open and find your sound, or drag the icon of the sound file onto the SoundHack icon in the Dock (if it's there, in the dock - - otherwise open SoundHack in the Applications folder. And SoundHack is only available on the Mac - - sorry, PC users. . .

Once your file is open, under the 'Hack' menu, select 'Phase Vocoder'.

This is one of my favorite hacks, because it's a somewhat obvious what's going on in the hack (at least in part). Phase vocoder will expand/contract the Time Scale (multiplying the length of sample) or the Pitch Scale (multiplying the frequency) by the number in the 'Scaling' box.  You can also click on the 'Scaling Function' button, and visually draw a scaling function.

My general process in SoundHack is to try something, see if you like it, then apply a hack to your result. You can create any number of new sounds by continuing to apply different hacks to a soundfile which itself is a product of a hack. Here's what my desktop looked like after carrying out this process six times:

You can alternate a SoundHack hack with an Audacity hack - - see if you understand what's happening, and if you can predict - - at least on a really broad level - - what new characteristics of the sound file will be revealed.

If you download an unzip this archive (  SoundTextures.zip ), you'll hear 10 sound files, moving from simple pitch gesture (#1) to rather nicely crafted ambient texture (#6) to rather intrusive solo texture that grabs our attention because it is broken apart and fairly pitchy (#9), to a hairpin rhythmic texture (#10) created from a short segment of #9. Here's how I created each:

1. I created this first sample in Audacity, by generating a few seconds of a sine wave at A440. I then selected bands of the sound and applied the Change Pitch filter to raise the pitch two, four, and ten semitones (half steps), corresponding to the notes B, C#, and G.
2. Next, I opened #1 in SoundHack and applied the Phase Vocodor with a pitch scaling of 4.0000. This transposes the file up 2 octaves.
3. For the next example, I took file #1 and (in SoundHack) applied Phase Vocodor with a time scaling that I drew on screen. The screen image looked like an extreme city skyline.
4. Then, I took #3 and applied the Mutation hack, 512 bands, USIM type, Mutation Index 0.5000. I picked as the Time Scale Target the file #1.
5. Almost done - - I took #4 and used the Phase Vocodor hack, but this time with time scaling. I selected 'Scaling Function' and clicked on the 'noise' icon, resulting in a fairly dramatic time scaling.
6. I opened #5 in Audacity, and inserted 8 or 10 seconds of silence at the end of the file. Then, I applied the Apple Matrix Reverb at 'Cathedral' setting. Now we have an ambient texture!
7. To create the Solo and Rhythmic textures, I copied about one third of a second of a fairly noisy part of the sound file #6 (I'm still in Audacity). This sample is called a noise profile. This is the noise profile. I selected it, and copied it into the clipboard.
8. With #7 selected, open the 'Noise Removal' filter (Audacity) and click on the 'Get Sound Profile' button. I selected all of file #6 and applied 'Noise Removal' - - about 1/3 on the horizontal slide bar. The application of the filter removes the audio content that corresponds to the the Sound Profile (#7) from the main file (#6). Blips and bleeps, and some extended sustained areas, remain.
9. This file is a bit of a pastische. I manually copied and pasted short segments from file #8 into itself multiple times in multiple places. I grabbed a short segment and applied 'change pitch', then I grabbed another short segment and re-applied that filter (use command-R). I selected some silent areas and normalized them to see if there was any interesting sounds there. And finally, I grabbed other small segments and applied the reverse filter. So, the process is select, apply filter, select new segment, re-apply filter, etc.
10. And finally, I grabbed a 3 or 4 second segment and pasted it into a new file. I clicked on the purple fast forward (end of file) button, and pasted the segment again. I repeated this until I had a file about 30 seconds long. I compressed the file to about half its lenth using the 'Change Tempo' filter, and then fading in the beginning, and fading out the end. Presto, rhythmic texture!