Can you turn a vocal into midi?
No, you can't. MIDI data is all
performance data -- which note number (pitch), how long it was held for (duration), which program number (which the receiving device maps to a particular sound patch), how hard you hit the key (velocity), how har you press the key until you release it (aftertouch), and so on -- but there is absolutely zero inherent
audio information. No way at all for the MIDI information to incorporate digitized audio data.
and I would be able to dice it up in Digital Performer3?
Well, DP can record audio, and you can dice up the audio data... it's probably also possible to map MIDI messages to other functions in DP, so perhaps you could do something like hit a C on a keyboard repeatedly and instead of recording it as a Note On, record it as a differnt bit of MIDI data that might control the level of a volume envelope. That way you could stutter the sound in response to your manual playing via MIDI. I guess you might do that if you really wanted to control the sound's stutter manually rather than just divide the tempo by, say, 64 or some other rhythmic fraction of a beat and have the audio file chopped up in even increments by the software. I have never used DP but Sound Forge has a stutter effect like this.
What exactly is the use for a sampler as opposed to any regular recording device? In other words, what does a sampler do?
Well, you know how a wavetable synth has a set of sounds built in (or accessible on a memory card or something) that you can choose from and edit and play from the keyboard? A sampler is simply a synth that also lets you create your own sound sets by recording the raw material yourself, and it provides a slew of tools for manipulating the sounds you record -- truncating them, putting in loop points, blending them with others, manipulating their attack-sustain-decay-release envelope, etc. When you are satisfied, you have a sound set that can be triggered by any MIDI controller just like a regular wavetable synth.
The main difference between a digital audio recorder and a sampler is that the end result in a sampler is a MIDI-compliant sound or set of sounds that you can play via a MIDI controller; whereas the end result in a audio recorder is a set of audio tracks that you can play from beginning to end and mix down to a completed stereo song.
AS you can probably see from this description, the boundaries of the terminology are a bit fuzzy.
One approach to the particular thing you are asking about is to reord the speech into a sampler and edit it so that when the MIDI Note On message is sent, the volume drops immediately. Then, you play that sound via the keyboard and play staccato 32 notes. Each time you hit the key the sampler plays the sound from the beginning, and each time you lift your finger off the key it stops. So if the full sample says "Do the do" and you hit in 32nd notes, it would sound like "D-D-D-D-D-D-D-D-D-D-D-D-D-D-D-D-D-D-D-D-D-D-D-D-D-D-D-D-D-D-D-D..."
Another approach is to lay it out in a digital audio recorder like DP, The you can cut off the entire sound except the intial part that's the right duration for the rhytmic subdivision you want. Then you can copy and past as many of these truncated pieces in a grid. The end result is the same. It's probably just about as easy to do this with either a sampler and MIDI notes or with a digital audio snippet in a digtial audio editor.
Or, you could use a stutter effect like I described earlier and chop silences into the whole section of audio so that they occur rhythmically. This would result in something like "D-u-u-u-h- - - -t-h-h-h-e-e-a-a-a- - - -d-u-u-u-h..." This would be very easy to do in a digital audio editor, but much harder to do with a sampler.
