Cross-gender voices are a tricky business. Even if you can really do a convincing cross-gender voice–and I know folks who can–the fact is that in a live reading, the audience knows that there’s only one person doing all the voices. There are two ways cross-gender voices can throw people out of listening. It’s really bad, and embarrassing, or it’s really good and shocking that a female voice is coming out of a man’s mouth. Either way, the listener drops the story for a moment.

This is like a turn of phrase that’s really stunning in a story. You stop reading for a moment and think, “Wow, that’s lovely.” That may be true, but the story has stopped, right there. Same thing with voicing. Any time you make the listener stop to think, you’ve injured your story.

The point of doing different voices is to make it clear who is speaking–it’s not to make it sound like there are fifty people sharing the stage with you. If you really want it to sound like there are completely different people, hire some actors.

Now, with that said, you also want to use your voice to enhance the character and to help paint a picture in your listener’s mind. Even when I’m doing same gender voices, I tend to “lighten” my voice a little to make it more feminine.

But, besides the “audio picture” I’m trying to paint, part of the reason I alter my voice for female voices is so that when I do male voices, I’m altering my voice to a similar degree.

Let me use a visual analogy. If you are watching a cartoon, you don’t think about the fact that there is no texture in hair or clothing. But, as soon as the animated character wanders onto a digitally rendered lawn, the fact that you can see every blade of grass is jarring. It makes the grass look unreal, and the character look unreal. They don’t and shouldn’t live in the same universe.

With voicing, if you want your cross-gender voices to sound real they must live in the same universe. So if you’re a guy and you’ve got to do a female voice, then don’t use your “natural” voice for a male character. Color your male voices to the same degree that you color your women’s.

And remember that you can be subtle.

So let’s look at some of the tools for doing this.  In my previous post I talked about the basic tools  of pitch, placement, pacing, accent, and attitude.  For working cross-gender the first three are going to be your most useful tools.

The key to performing a voice that is not your own is to play with stereotypes. Sad, I know, but people have preconceptions in their heads about what men and women sound like.

The perception is that men’s voices are deeper, placed more in the throat and more staccato.

Women’s voices are perceived as being higher, placed more in the front of the mouth and more fluid.

There are clear exceptions to these ideas, for instance, I have a friend who is a female tenor. When she calls, before she identifies herself, it’s not uncommon for me to think that a man is on the other end of the line.  Likewise, I have a friend who is a high tenor and his natural speaking voice is in my range. Again, I’ve had moments of wondering who the woman on the other end of the line was.

The point here is that even though there are exceptions, the societal expectation that a low voice means male and a high voice means female can work in your favor when voicing.  Shifting your voice up from your narrator voice by only a tiny fraction can make the character sound more feminine because you are using the built in preconceptions of your audience.

And remember, keep it subtle.

Mary Robinette Kowal was the 2008 recipient of the Campbell Award for Best New Writer. Her short fiction has appeared in Strange Horizons, Cosmos and Asimov’s. Mary, a professional puppeteer and voice actor, lives in Portland, OR with her husband Rob and nine manual typewriters. Tor is publishing her debut novel, Shades of Milk and Honey, in 2010.