Hello everyone!

This time I’d like to share a short story called “Between.” The listening time is just 3 minutes and 15 seconds, but producing this video took far more time.
The difficulty was that it combines three voices: a narrator (male), a protagonist (male), and a female character. At first, I decided to read the narrator myself and use two voices from https://elevenlabs.io/](https://elevenlabs.io/ for the characters. So I began selecting voices to match the characters I had chosen in advance.

But wait — first I need to explain how I chose the characters in the first place.
There’s an interesting moment here. You’re supposedly writing not about yourself, yet somehow "you" inevitably break through. But that’s not even the main point. No matter how you try, you inevitably filter what you’re writing through your own moral lens.
The story has two characters: a guy named Alex and a girl named Becky. As I imagined them, I started browsing portraits of contemporary actors. Here is what I settled on eventually:

The guy looked roughly like this:
…
or like this.

And the girl looked roughly like this:

Naturally, I fed their descriptions into an AI image generator, and here’s what it produced — for Alex

and for Becky

What’s interesting is that once you choose your characters, the internal connection with yourself seems to break off, and even the sound of your inner monologue begins to change, adjusting itself to these people.

But to express that, you need to have very good control over your own voice. And the issue here isn’t really the sound of the voice itself — which, between us, I’d prefer to be lower and deeper. By the way, if I had realized this when I was younger and seriously worked on my voice, squeezing out everything it’s capable of, I might have had much more success with women.

In any case, the way I pronounce words when recording a video is very different from how I speak in everyday life. For one thing, I don’t talk that much — my circle of communication is fairly limited. So I have a habit of “filtering” and cutting off words. Instead of saying “Good morning,” I’ll say “morning,” or sometimes just “m—” (you get the idea).

But you’re not recording narration for family and friends. Here, everything needs to be articulated clearly. And in the process, I discovered that some word combinations are difficult to pronounce quickly. Though it’s not really about speed. It’s about the vocal variations and modulations required to match the situation being described.

You have to keep track of several things at the same time: clarity of pronunciation and articulatin and whether your delivery fits the dramatic situation in the story. It sounds pretentious, but that’s exactly how it feels.

So, as I said earlier, I first selected voices to match the chosen characters. And of course, nothing ever matches perfectly — you choose from what’s available. In this case, the voice Vinsent won out for Alex, and Jessa for Becky.

And when the dialogue was recorded and I tried inserting my own voice as the narrator, it became obvious: it didn’t work. And not because of recording quality — not because of apartment noise, not because I had to clean the audio or normalize levels. The real difficulty was something else.

A narrator can’t just speak “in general.” His voice has to adjust every time to what the character has just said — in meaning, in tension, in the internal state of the scene. And that’s where I hit the limit of my abilities.

You can’t simply record the text and then cut out the necessary pieces and place them where needed. Each time, you have to listen to the character’s line and then read so that you land exactly in the moment — in tempo, pitch, tension, and so on.

In the end, I had to abandon my own voice and replace it with yet another voice from the site. That one also took some searching. It was crucial that the narrator’s voice not resemble Alex’s voice. For that role, I chose Spuds Oxley — in terms of timbre and overall feel, he fit best.

Even then, it wasn’t always possible to get the desired dramatic effect. Sometimes, though, his calm, comfortable delivery clashed with the tension of the scene — almost as if the narrator were standing slightly apart, internally amused, thinking: why all this drama? At moments like that, you start to understand how directors feel when actors do something that doesn’t match their intention.

But at that point, you run into the limits of what’s possible and tell yourself: one can’t embrace the infinite. You do the best you can with what you have.

So — that’s what came out in the end.