One of the unexpected extras that come with FL Studio 7 is the Speech Synthesizer. If you're doing a track that calls for a phrase from an alien or robotic voice, then you'll probably find that one of the 20 speech models in the menu does the trick. Just type your phrase in the box, and you can hear it spoken or sung by a man, a woman, a munchkin, a nerd, or a robot, among others.
When you click on the check mark to accept what you're hearing, FL Studio asks you to save a .speech file, which contains the phrase as separate words. The program then creates a Fruity Slicer Generator (see Fig. 1), in which each word is assigned to a separate key, as well as a piano-roll window from which the words can be played back with the original rhythm.
FIG. 1: The words produced by the Speech Synthesizer appear in the Fruity Slicer Generator.
There are many ways to extend the power of the Speech Synthesizer by rearranging and processing its output. Here are a few of my favorite techniques.
Use the Speech Synthesizer three or four times, entering the same phrase each time but choosing different Personalities. Select Monotone as the Style for each phrase and enter a different pitch. Choose the pitches to create a chord.
Assuming you didn't switch to a new step-sequencer pattern during the process, you'll now have a pattern that will play back a choir singing your phrase as a chord. The sense that it's an ensemble will be enhanced by the differences among the Personalities, but for the same reason, you may find that the choir doesn't have good rhythmic cohesion. You can solve this by quantizing or hand editing the notes in the piano roll so that they have the same start times. Rerendering one of the voices with a different Rate setting may give you a better rhythmic match.
Each note in the Slicer's piano-roll data can be tuned up or down in the event editor strip. Moving the note pitch by more than a couple of half steps will introduce timbral changes, but if you choose initial pitches in such a way that each of them needs to be transposed up or down by only a few semitones, it's quite practical to produce a synthetic choir with moving harmonies (see Web Clip 1).
Massaging a Word
FL Studio's new audio editor, Edison, provides some nice tools for processing samples, including time-stretching, a convolution reverb, and a blur tool. Edison can't load .speech files, but there's an easy work-around. After bringing up the Channel Settings window for the Slicer that's playing your phrase, click on the Plugin tab and then on the second button where you see a file folder. From the drop-down menu, select Save Original Sample. This command can export the .speech file as a WAV file, which you can then load into Edison with markers for the words.
If you want to process individual words in a way that would change their length, create a second Edison. Select the word(s) you want to process, click and hold the Drag button (the one at the right end of the Edison toolbar), and drag into the second Edison.
Drag the resulting file into a new Sampler Generator to trigger it in patterns. Before you do the drag-copy, however, open the Edit Properties window for Edison (the shortcut key is F2) and switch off Tempo Sync. In the early-release version of FL 7, the Sampler Generator has problems playing tempo-synced files.
Scratch Me Up
From the Channels menu, create a Wave Traveller and drag your synthesized word(s) from Edison into the Wave Traveller. Here you can create a different scratching sound for each MIDI key. A Wave Traveller can contain many scratches but only one sound file. If you need to scratch two or more different sounds from the keyboard, then create a Layer Generator, assign several Wave Travellers to the Layer Generator, and choose Split Children from its menu.
FL Studio is capable of many creative effects, from ring modulation to envelope-controlled morphing EQ. Using the Speech Synthesizer is a great way to start experimenting with them.
Jim Aikin writes regularly for EM and other music-technology magazines. Visit him online atwww.musicwords.net.