[inaudible]
I work from home as a transcriptionist. Most of my work is market research: focus groups, two hours of eight strangers around a table telling a moderator what they think of a cereal box or a bank’s new logo. The firm sends me the audio, and I send back a transcript, and I get paid by the audio minute.
The rule is verbatim. You type everything. Ums, false starts, laughter, crosstalk. If you can’t make out what someone said, you don’t guess. You write [inaudible] and a timestamp, and the client knows you tried. A good transcriptionist has a low [inaudible] rate. Mine is the lowest on the team. I’m a little proud of that, which you should keep in mind.
The files from the Dunmore facility started coming in May. Same room every time: you learn to recognize rooms by their tone, the hum of the air system, the tick of a wall clock. Dunmore has a clock that’s slightly loud, and under the hum there’s a low, uneven murmur that I took for a television in another room.
It wasn’t a television. On the third Dunmore file, during a quiet stretch while the participants were filling out a form, I turned the gain up to check a mumble and realized the murmur was a voice. One voice. Speaking steadily, not to anyone at the table. Too low to make out.
I typed:
[00:41:17] [background speech, inaudible]
And went on.
It was in every Dunmore file after that. I’d catch it in the pauses. Same voice, man or woman I couldn’t say, very level, very patient, like someone reading aloud to a person who’s falling asleep. Once I’d noticed it, I couldn’t stop hearing it. And because I’m the person I am, I couldn’t stand leaving it as [inaudible]. It was speech. It was in the audio. Verbatim means everything.
So I cleaned it up. I’m good with the software. I notched out the air-system hum, and the clock, and pulled the voice forward, and on the fifth Dunmore file I got my first clear phrase.
[01:12:03] BACKGROUND SPEAKER: —is not on the guest list. Write that down.
I wrote it down.
I should say that I asked the project manager about it. She said Dunmore was a standard rented facility, and that the observation room behind the mirror was sometimes occupied by the client, and that people back there sometimes talked. She said I didn’t need to transcribe the observers.
But the observers would be in the observation room. In the cleaned-up audio, the voice wasn’t in another room. It was in this one. It was a little closer each file. On the seventh, it was clearly at the table.
[00:22:40] MODERATOR: Okay, so, let’s go around and say what three words come to mind when you look at this packaging.
[00:22:44] PARTICIPANT 3: Um. Fresh? Clean?
[00:22:47] BACKGROUND SPEAKER: Hungry.
[00:22:48] PARTICIPANT 3: Hungry. Sorry, I don’t know why I said that.
On the playback you can hear Participant 3 laugh. It’s not a good laugh.
I could have stopped. I could have marked it all [inaudible] and sent the transcripts in and cashed the checks. Nobody would have known. The voice wouldn’t have known.
That was the thought that kept me up. It was such an odd thought. The voice wouldn’t know. As if it could. As if it was waiting to see what I’d do.
[01:31:55] BACKGROUND SPEAKER: You can hear me. You’re very good at this.
[01:32:01] BACKGROUND SPEAKER: Nobody else could hear me. They wrote me down as nothing. Do you know how long I’ve been nothing?
I typed it. Verbatim.
[00:03:12] BACKGROUND SPEAKER: Don’t mark me inaudible. I’m right here.
[00:58:29] BACKGROUND SPEAKER: You type with your left hand faster than your right. Did you know that?
[01:40:02] BACKGROUND SPEAKER: There’s a glass of water by your keyboard. It’s warm. You poured it at two.
I had. I’d poured it at two.
I stopped cleaning up the audio after that. I didn’t have to. The voice was louder in the raw files now than the participants were. On the last file, the moderators’ and participants’ voices are faint, muffled, like they’re coming through a wall, and the voice is right at the microphone. Right at my microphone, I want to say, though that doesn’t mean anything. I only have headphones.
The last Dunmore file came in on a Thursday. It was two hours long. The participants were in it for about forty seconds before the audio drops to room tone, and then the room tone drops away too, and there is nothing at all, not even hiss, and then:
[00:01:03] BACKGROUND SPEAKER: Thank you for listening so carefully. Please introduce yourself for the recording.
That was all, for nearly two hours. Silence and that line. Then, at the very end:
[01:58:44] BACKGROUND SPEAKER: It’s all right. I’ll do it.
And it said my name. And my address, and the name of my street, and which window my desk faces, and what I’d had for lunch. It said them all very slowly, very clearly, the way you’d speak to a transcriptionist so there would be no doubt at all.
I sent the transcript in. I didn’t know what else to do. The project manager wrote back to say the Dunmore project had been cancelled before that session took place, and that the file I’d been sent didn’t exist on the firm’s server, and asked if I was feeling all right.
The voice doesn’t need the files anymore. That’s what I’m trying to tell you. I hear it with my headphones off. It’s under the hum of the fridge, under the traffic, under the tick of the kitchen clock. Low, level, patient. When I hold very still I can make out nearly all of it.
It’s dictating. I don’t know who to. I think there are other people out there who listen carefully for a living, and it wants them too. I can almost make out the names.
I’m trying so hard to write them down as [inaudible]. You have no idea how hard I’m trying.