Mini-Tutorial On Creating A Voice Story
By Wayne Rice

First you must have a good story to tell. Spend some time thinking about it. Write a first draft and put it aside. I suggest you keep it fairly short, under a thousand words. Before you pick it up again, try writing it down again without looking at the original - this will help you to identify the most important details (the ones you included in both), and the weaker details (the ones from the first that aren't in the second.) Develop the second draft from the two first drafts.

A word about dialog: the more characters you have, the more voices you will have to define. They all must be distinct from each other, so that the listener can easily tell who is speaking; still, it is a good idea to frequently identify the speaker, if you are using a narrator. Narration allows many things to happen more smoothly than a pure dialog story; it creates a perspective, it can move the story along without regard for time, logic, or other factors, it can provide motive and explanations, details and identities. When you have a really strong story, proceed to the next phase.

Get familiar with the software. Read the tutorials for each piece of software you will use. Practice with each one. Read and run all the tutorials found on the avnex site, including these: mp3-player.audio4fun.com/tutorials-voice-changer.htm and you can find more on the site. Use VCSD 6.0 (or whatever the newest version is) to create the voices. Use MMG to edit the output files.

Separate out all of the dialog for each voiced character, and for the narrator. Use the narrator's file as the master, and when the time comes, cut and paste the other dialog into it, saving often. Every so often, save the current file under a new name, like "file_b", so that you can step back through them if you need to recover an earlier edit or change. Be sure to write down the settings for each character's voice so you can go back and generate more dialog if needed. Also be sure to write down the name of each voice file that VSCD creates, as it uses a somewhat cryptic naming convention; note the file name and then the content. I like to do the narrator's file first, and when I come to the place for another voice, I read it in the narrator's voice, but very softly - that way the length of time for the dialog to be cut in is accurate, and the position will be easy to recognize in the graphic format of the editor when it comes time to splice in the other voices, because that space will show up as a very flat space in the audio graph.

When you are using VCSD, be sure to make sure you only have the filters and controls on that you want on, because they do not turn themselves off, you have to do that. I prefer to use the classic interface, because it is very easy to see where things are and it reminds you with tabs what your options are for filters, such as equalizer and effects, frequency morpher, etc. Check them frequently to make sure you have only the controls you want active. If not, you can produce undesired background effects and overlapping filters that will work against each other.

A word on filters - sometimes, more often than not, less is more. Instead of cranking the filters to the max, try a small adjustment first; drop the pitch to 95%, then 90% and see if either of those is good. If you use the frequency morpher, just drag the line a little at first. And don't forget to turn off the effect when you done.

When you put voice to your character's dialog, don't forget your most versatile tool - your own voice. If you are male, and you want to make several female voices, try altering your natural voice in different ways for each voice; speak higher, softer, slower, and then apply effects to your own altered voice. You will find that each change you make in your own voice will cause a somewhat different result, even with the same filter/effect in use.

Finally, a hint for recording if you make an error, mispronounce something, or lose your place, just pause, but keep recording. Do the line over again. You can even try a line more than one way while you are recording. You are going to be cutting and pasting anyway, so allow yourself the luxury of slightly longer raw files so that you can get it all in one sitting. It's a good idea to pause between dialog that is separated by other words or speakers; the pauses show up as flat areas in the audio graphic and help you to keep your place during the edits.

I like to splice all the voice parts into one master file by adding each one to the narrator file. Then I mark the time where each voice begins and write that down. When I am copying and pasting, I also write down the time spot that I will be pasting into; that way you can jump around the file easily, find what you need and get back to where it needs to go without a lot of frustration. Remember, it will take as long to produce the file as it will take to create the story, so plan your time well and be patient; that way you will have the best chance of making a really outstanding audio story.

Wayne Rice
United State

Voice story: Cinderella Jolie