0:00

Every job has a set of

0:02

minimum requirements, things that you kind

0:04

of have to get right to

0:06

do the job. For an audio book

0:08

narrator, one of the biggest ones

0:10

is just saying the names of

0:13

people and places correctly. When you're

0:15

reading, you know, The Lord of the

0:17

Rings for the first time, you don't

0:20

have to say those names. They're

0:22

crazy. But my job is to know how

0:24

to say them. That's Sean Pratt. He's

0:26

been an audiobook narrator for 25

0:28

years and has 1,200 audiobook titles

0:30

under his belt. So here's how

0:32

he nails down pronunciations. If it's

0:35

a fiction book, he goes straight

0:37

to the source, the author. He

0:39

sends them a list of the

0:41

tricky names. And I say, go

0:43

to the voice memo on your

0:45

phone, say this name slowly, twice,

0:48

and then I'll write it up

0:50

phonetically on my end. For names

0:52

from non-English languages, he uses a

0:54

website. There's one called Forvo.com-F-O-R-V-O. This

0:56

stands for foreign voice. And so

0:58

you can hear a native speaker

1:00

say the word, or like, how

1:03

do you say this name that's

1:05

Turkish? Because it's got a funny-looking

1:07

C letter there. How do you

1:09

say that? Oh, it's a C-H

1:11

sound. Or it's a Z sound,

1:14

or it's a Z sound, or

1:16

it's a Z-Z sound, or whatever.

1:18

So we often hire researchers to

1:20

help him track down pronunciations, but

1:22

even with all those safeguards, mistakes

1:24

can creep in. So I did

1:27

this book on stoicism recently, and

1:29

there was a philosopher, she was

1:31

a cynic, and her name is

1:34

Hipparchia. Unfortunately, I misread my researcher's

1:36

notes, and I said, Hipparchicaica. And

1:38

I said it 75 times. Oh,

1:41

no. So now I get to

1:43

go back and you don't go

1:45

to Hipparkia once. You have to

1:48

read the sentence for the context.

1:50

Boy. So I got to reread

1:53

75 chunks of that little, you

1:55

know, little sentences throughout the piece

1:57

to go, Hipparkia. I'm

2:03

Dan Heath and this is

2:06

What It's Like to Be.

2:08

In every episode we walk

2:10

in the shoes of someone

2:13

from a different profession, a

2:15

couples therapist, a stand-up comedian,

2:17

a cattle rancher. We want

2:20

to know, what do they

2:22

do all day at work?

2:24

Today we'll ask Sean Pratt,

2:27

what it's like to be

2:29

an audiobook narrator. We'll talk

2:31

about how he develops character

2:34

voices. why he advises aspiring

2:36

audiobook narrators to record in

2:38

their closets, and what AI

2:41

voices get wrong about the

2:43

way people talk. Stay with

2:45

us. Before Sean Pratt sits

2:48

down to record a novel,

2:50

he reads the whole thing, cover to

2:52

cover. He takes notes, and he says

2:55

what he's looking for in that initial

2:57

read-through. aside from potential pronunciation landmines, are

2:59

clues about the characters and what they

3:02

might sound like. The more fantastical the

3:04

genre, the more fantastical the character choices

3:06

I'll make. If you know like a

3:08

Western or a sci-fi or a fantasy,

3:11

I'll have big broad choices with my

3:13

voices. So recently he narrated one of

3:15

Agatha Christie's parro novels and he wanted

3:18

it to sound like an old black

3:20

and white film. So the American sort

3:22

of sound like that kind of American

3:25

in the 1920s. Hey there, pal, how

3:27

are you going? And then, you know,

3:29

you have a cue problem, how are

3:32

you doing? It's very nice to see

3:34

you. Yes, I say, we should go

3:36

over here and look at this thing.

3:38

And then you have, you know, like,

3:41

oh, I'm sort of a chippy-chappy guys.

3:43

How are you doing? How are you

3:45

doing? And you know. And you can

3:48

go a little further with it. because

3:50

it doesn't feel right for the era.

3:52

and also what the audience would expect.

3:55

So that's where the artistic part comes

3:57

in. He mostly works alone, making the

3:59

character choices on his own based on

4:02

the text, but he does occasionally collaborate

4:04

with authors to make different choices. About

4:06

two years ago, I did a fantasy

4:08

piece, and the author was a public

4:11

school principal in North Dakota, Stephen Googlet,

4:13

to his name. So he had a

4:15

meeting. He sent over, because he wants

4:18

input, he's paying me. And so we're

4:20

going over the different character choices. And

4:22

we have elves and fairies and goblins

4:25

and all sorts of stuff. And he

4:27

wanted the fairies, the little wee people

4:29

with the pointed ears, as it were.

4:32

No, the elves, excuse me, the elves.

4:34

He had mentioned he wanted them to

4:36

sound Israeli. I think in his mind,

4:38

because they're like action-oriented elves, these guys.

4:41

And I think he thought they were

4:43

going to sound like that. And I

4:45

said, Stephen. If we do that, they're

4:48

going to sound like a bunch of

4:50

old Jewish guys from Brooklyn. You know,

4:52

we'll work at a deli. Because that's

4:55

my, hey, how you doing? Let's go.

4:57

OK, fine. We're going to go fight

4:59

the OGO. OK, fine. Let's go. So

5:02

I said, let's not do that. I

5:04

said, I would vote for maybe a

5:06

British accent. So it is a conversation.

5:08

And because there are also some accents

5:11

that, you know, people think that we're

5:13

machines that we can just pull these

5:15

accents out of thin air, but you

5:18

can't. Sometimes you have an accent come

5:20

along that's actually quite challenging. Like for

5:22

me, the South African accent's quite challenging

5:25

to do. So if I have a

5:27

character that's South African, the character doesn't

5:29

exist in the accent. It's part of

5:32

the character. So when you have an

5:34

accent that's a little challenging, you just

5:36

want one or two or three vocal

5:38

effects that are unique to that are

5:41

unique to that accent. build the rest

5:43

of it around. Give me an example

5:45

if you remember one where you had

5:48

to sort of consciously create the accent

5:50

for the character and maybe it was

5:52

cobbled from different sources, but walk us

5:55

through like your process in getting there.

5:57

So if I'm reading the text and

5:59

it says this character, the way the

6:02

author describes them, that they're a 70-year-old

6:04

man from Georgia who's been a chain

6:06

smoker and he's very angry. All right,

6:08

well that just sort of tells you

6:11

what this guy's going to sound like

6:13

like this. He's just going to have

6:15

a voice like that. Right, so you

6:18

use the adjectives to guide you. Other

6:20

times, if they don't have that kind

6:22

of a sound, then you sort of

6:25

cast from this catalog of people that

6:27

you know. A good performer is a

6:29

good mimic, and a good mimic is

6:32

a good observer. So in this case,

6:34

when I'm observing by listening to the

6:36

people around me. So... I listen to

6:39

people the way they speak or celebrity.

6:41

It can be anybody. It could be,

6:43

you know, my brother-in-law or a celebrity

6:45

or someone I know in my neighborhood

6:48

or whatever. If they have a unique

6:50

voice, I've cataloged it in my head.

6:52

So going back to the fantasy piece

6:55

I mentioned earlier, our villain, he had

6:57

a very unique way of speaking, and

6:59

we ended up, I cast his voice

7:02

as James Spader, the guy from Blacklist.

7:04

Give me a name, Henry, or I'm

7:06

going to drag you out, throw you

7:09

in the trunk. fly you to Papua

7:11

New Guinea and have your head stuck

7:13

on a pole. It just felt right.

7:15

And so then I watched several episodes

7:18

of The Blacklist to get a feel

7:20

for his cadence and style, and then

7:22

I apply it to the character. I'm

7:25

just in awe of the kind of

7:27

voice collection. Like I am terrible at

7:29

accents and voices and I've always admired

7:32

people who can kind of shuffle among

7:34

them. Have you ever encountered one that

7:36

was an unusual challenge? Like earlier you

7:39

mentioned South African accents. What's the hardest

7:41

voice you've ever had to nail? Oh

7:43

goodness gracious, there's so many. You know.

7:45

As you can hear my voice, when

7:48

I have African-American characters in there, because

7:50

for the majority of Africa, they have

7:52

a much lower voice. And of course,

7:55

there's a different melody, rhythm, and tempo.

7:57

And you walk a fine line when

7:59

you do different ethnicities. You want to

8:02

acknowledge the ethnicity the way they speak,

8:04

but you do not want it to

8:06

be a caricature. And that's different. And

8:09

that's where you look to the writer,

8:11

the style. But sometimes you're forced into

8:13

a certain sound, so when I narrated

8:15

Sound in the Fury by William Faulkner,

8:18

that was especially hard on a number

8:20

of levels. It's only four chapters long,

8:22

but each chapter is like at least

8:25

two to three hours long. And it's

8:27

four different characters, they're sort of their

8:29

perspective of what's going on. And two

8:32

of the chapters are written in stream

8:34

of consciousness writing. In fact, one is

8:36

a young man who's mentally disabled. And

8:39

he goes backwards and forwards in his

8:41

own time as he's thinking. So you

8:43

have a character who is not fully

8:45

present going back and forth in memory

8:48

to the present when he was 3,

8:50

13, 23, 33. And you have to

8:52

make that coherent somehow. And then he

8:55

wrote the patois of the African-American characters

8:57

in their patois. So it would be

8:59

like if you were writing someone from

9:02

Boston and you took the ours out

9:04

and put an A-H, so pack-you-can-havied yad

9:06

kind of thing. And what's tricky about

9:09

when the author does that is you

9:11

have to decipher their writing and then

9:13

make it sound conversational on top of

9:15

it. It's quite challenging. That's so interesting.

9:18

So how much can you get done

9:20

in a session? So you go into

9:22

your home studio, you're ready to get

9:25

started. Do you start the first page

9:27

of a book or is there a

9:29

logic to starting in a different order?

9:32

No, you want to go front to

9:34

back. I mean, obviously in fiction there's

9:36

the narrative. There's still a narrative. You

9:39

as an author know, you know, you

9:41

set up your premise in nonfiction and

9:43

the intro preface. forward. And then chapter

9:45

by chapter, you take us on this

9:48

journey, you're revealing your knowledge and sharing

9:50

your truth with us. So no, you

9:52

want to go sequentially. And when I'm

9:55

working in the booth, for a professional

9:57

narrator, the ratio is about two to

9:59

one. So it takes about two hours

10:02

of work to generate one finished hour

10:04

of audio that now will be proofed

10:06

by the company or if I do

10:09

private work, my own proofer. That's pretty

10:11

impressive, that level of productivity. Well, if

10:13

you want an anecdote, so my very

10:15

first book is 1996. I'd moved down

10:18

to Washington DC to sort of start

10:20

my life over. I'd met somebody and

10:22

moved down there to be with her

10:25

and I was going to be using

10:27

a friend's studio and I was going

10:29

to have a monitor, a woman who

10:32

sat outside and was basically proofing me

10:34

in real time. So she was on

10:36

a headset, so if I made a

10:39

mistake, she'd click in like Sean, that

10:41

was street light and you said street

10:43

lamp, you know. Okay, thank you. And

10:45

you rewind the tape and then you

10:48

hit record and you listen to yourself

10:50

and then you punch in and record

10:52

over the mistake and you keep going.

10:55

It's called Punch and Roll. So I

10:57

practiced, you know, how to use the

10:59

machine and I read the book and

11:02

made my notes. The book was Cabbages

11:04

and Kings by O. Henry, a series

11:06

of short stories. And in my very

11:09

first session, there was a three hours

11:11

of work. I did exactly 15 minutes

11:13

of material. Oh my God, it was

11:15

beyond, it was the most stressful event

11:18

of my entire life. And so when

11:20

we finished, I thanked Bernadette, the monitor,

11:22

and I said, I'll see you tomorrow.

11:25

And I drove back to Alexandria, Virginia,

11:27

where I was living at the time.

11:29

And I walked into the row house

11:32

that we were living in, throw my

11:34

bag on the floor, and just collapse

11:36

on the rug in the living room.

11:39

And I'm staring up at the ceiling

11:41

and my girlfriend walked over and she

11:43

said, are you okay? And I was

11:45

like, this is so much hard. than

11:48

I thought it was going to be.

11:50

Wait, so what made that first day

11:52

so difficult? I was running everything. See,

11:55

that's the trick. It's like patting your

11:57

stomach and rubbing your head or whatever.

11:59

It's not enough. I'm not just performing

12:02

it. So like when you narrated your

12:04

audio books, right? You went into a

12:06

studio and someone did all the technician.

12:09

Okay. I'm doing all of that on

12:11

top of the performance. Wait, you're editing

12:13

it too and recording? Yes. I was

12:15

shocked to hear that Sean is essentially

12:18

a one-man band. Like not only is

12:20

he performing the work, which is hard

12:22

enough, but he's editing the audio on

12:25

the fly as he records it, and

12:27

he's relying on a technique that he

12:29

mentioned a minute ago called punch and

12:32

roll. It goes back to the days

12:34

of tape. So the idea would be

12:36

when you're recording on tape, you would

12:39

record, and then you make a mistake,

12:41

and then you make a mistake, You

12:43

hit play and you listen. And then

12:45

somewhere before that mistake, where there's a

12:48

pause somewhere, you hit record and the

12:50

spinning magnetic head would punch into the

12:52

tape as it were and roll over

12:55

the mistake with a new take. So

12:57

you're punching and rolling. You are editing

12:59

in real time. And this technique isn't

13:02

just useful for basic editing. It's what

13:04

allows him to do a whole scene

13:06

of dialogue between a bunch of different

13:09

characters. You're not doing a live radio

13:11

play. So you're not having to do

13:13

it. I can't pay the rent, but

13:15

you must pay the rent. I'll pay

13:18

the rent. Curses, you know, in my

13:20

hero, whatever. It's, when you have one

13:22

or two people, that's easy. When you

13:25

have four or five, it can get

13:27

tricky, and that's where a punch and

13:29

roll can save your bacon. So for

13:32

instance, I had a scene with, you

13:34

know, like four people in this room.

13:36

And they, you know, there was one

13:39

British lady, there was Urquil, and there

13:41

was an American, and then a British

13:43

man. So you build it literally one

13:45

exchange at a time. So in other

13:48

words, I'll start off with Urquil. saying

13:50

the first sentence and then stop. Go

13:52

back and listen to it as the

13:55

next character and then record her response

13:57

to him. And then go back, you

13:59

know, so I'm building a sentence by

14:02

sentence and I get a cleaner delineation

14:04

between those different character voices. Oh, that's

14:06

interesting. So it's like one comment, pause,

14:09

replay the comment, respond as though in

14:11

real time, pause again. Yeah, I mean,

14:13

that's how I do it. Punch and

14:15

roll helps him work just a little

14:18

quicker. And when you're recording, say, a

14:20

15-hour book, every ounce of efficiency you

14:22

can squeeze out of your workflow, it

14:25

matters to the bottom line. In the

14:27

world of voiceover, I have lots of

14:29

friends who do cartoons and video games

14:32

and commercial video and political spots, and

14:34

that's where the big money is. And

14:36

that audio books are sort of viewed

14:39

as the ugly red-headed stepchild of the

14:41

voiceover industry. Well, why is that? That

14:43

doesn't totally add up to me. Because

14:45

the amount of work for the amount

14:48

of compensation we get, comparatively speaking, is

14:50

a lot less. Well, it's sort of

14:52

a chicken and egg thing. Like, why

14:55

hasn't it become as valuable as some

14:57

of these other... Because when you do

14:59

a video game, the budget is already

15:02

there. And they know it's only going

15:04

to take the talent an afternoon or

15:06

two or three hours to do the

15:09

entire cartoon. And it's a much bigger

15:11

project in its scope and with the

15:13

budget. With audio books, think of an

15:15

audio book as almost like you're investing

15:18

in a stock. There's only so much

15:20

money you can invest in it. And

15:22

then it's going to take a certain

15:25

amount of time to recoup your money.

15:27

That's why you have audio companies that

15:29

have such large catalogs. As Sean mentioned

15:32

he recently narrated Faulkner's The Sound and

15:34

the Fury. That was because it went

15:36

into the public domain in 2025 so

15:39

the audiobook publisher wouldn't need to pay

15:41

for the rights to record it. So

15:43

you know that book is I think

15:45

it was like eight hours long. Okay

15:48

well they paid my fee and I'm

15:50

a union member so they paid my

15:52

union scale and all the stuff. with

15:55

that. So let's pretend I don't I

15:57

actually don't remember what the end cost

15:59

was. Well let's say it's like $5,000

16:02

all together. Well every download they only

16:04

get 40% because Amazon you know at

16:06

Audible doesn't obviously make enough money and

16:09

Jeff Basos needs a new yacht so

16:11

they take 60% of every download. And

16:13

so it takes quite a while to

16:15

recoup your money before you flip into

16:18

black. And do you get residuals once

16:20

they do pay out that initial investment?

16:22

No. My work as an audio book

16:25

narrator is strictly work for hire. That's

16:27

why I get a higher fee up

16:29

front. I wouldn't much prefer that, frankly.

16:32

Got it. Why do you say that?

16:34

Because I would have to wait a

16:36

lot longer to make that money over

16:39

a longer period of time. And sometimes

16:41

you do a book that just doesn't

16:43

sell very well. True. So you get

16:45

paid regardless, and you know, maybe one

16:48

out of 50 go on to be

16:50

a big home run, but at least

16:52

you have the certainty of the payment.

16:55

Yeah, it's like Hollywood. You know, they

16:57

have a couple of what they call

16:59

tentpole movies that support all those smaller

17:02

projects, and they're opening for a wild

17:04

card, a movie they didn't think was

17:06

going to take off that explodes and

17:09

becomes the next big thing. by Bessel

17:11

Vandercock. It's about how the body experiences

17:13

trauma and how it actually physicalizes trauma.

17:15

It was a wonderful nonfiction piece. That

17:18

book has sold, I kid you not,

17:20

hundreds of thousands of copies in the

17:22

audiobook world. And not to mention the

17:25

physical book, it's been translated, it was

17:27

a huge seller. And more recently I

17:29

did a book by Jonathan Hite. called

17:32

the Anxious Generation. Oh, no way. You

17:34

read that one. Yeah. Where that book

17:36

has been a phenomenon. Yeah, it's a

17:39

huge phenomenon. And I got it through

17:41

circumstance. Jonathan had a vocal issue that

17:43

happened. He lost his voice for several

17:45

weeks. And it was very tender when

17:48

he finally did get his voice back.

17:50

like he didn't have the stamina to

17:52

do the book because it was like,

17:55

I don't know, 13, 14 hours long.

17:57

Wow. So paint us a picture of

17:59

where you're working. Are you sitting? Are

18:02

you standing? Are you standing? Do you

18:04

have your finger on the mouse to

18:06

click record and stop? Presently I work

18:09

in a little, it's a booth that

18:11

I purchased off of Amazon, made of

18:13

PVC pipe and blankets. It's designed for

18:15

vocal recording. I put in a little

18:18

desk. And I have my equipment clipped

18:20

on here, the flat screen, so I

18:22

can see the software. I have a

18:25

bookstand from my iPad, because I narrate

18:27

the text from the iPad, so I

18:29

don't make any noise. I'm just flicking

18:32

my finger to make the page move

18:34

up. The mics here. You funnel in

18:36

some fresh air from your floor vent.

18:39

And you try to find the quietest

18:41

room in the house to record in.

18:43

And then it's really about what time

18:45

of day you work best. I tend

18:48

to record really well early in the

18:50

morning and late in the morning and

18:52

late in the afternoon and late in

18:55

the afternoon. Because I also

18:57

teach around that so I'm I teach

18:59

I teach people how to become audio

19:01

book narrators That's the other thing that

19:04

I do and I coach authors like

19:06

you who want coaching Because they're about

19:08

to go in and experience what you've

19:11

done like at least two or not

19:13

three of your books haven't you just

19:15

two yeah, and I'm about eight weeks

19:17

too late in seeking your coaching sadly

19:20

Yeah, so I have authors come to

19:22

me, and I work for them there.

19:24

But yes, so my setup is, it's

19:27

in a little four-by-four box. And I

19:29

tell people all the time, they were

19:31

interested in getting into audio books. I

19:33

say two things. I say, narrating books

19:36

is more about temperament than it is

19:38

about talent, because you're really just running

19:40

on your own. You have to make

19:43

all those decisions. You have to engineer

19:45

it to send to the client. If

19:47

you can't make those kinds of decisions,

19:49

if you need someone to hold your

19:52

hand every step of the way, then

19:54

you're in trouble. And the second thing

19:56

is, do you have the temperament to

19:58

sit in that little box, hour after

20:01

hour? and work, and a lot of

20:03

people can't. Sean says that when someone

20:05

mentions to him that they think they'd

20:08

like to be an audiobook narrator, he

20:10

suggests they try it out. You know,

20:12

if you actually sat in your closet

20:14

for two or three hours a day

20:17

and read a book out loud, you

20:19

would understand exactly what it is I

20:21

do and what you think you might

20:24

want to do, and it'll show you

20:26

faster than any other. Gosh, that is

20:28

such a good point. This is an

20:30

unusually trialable career, isn't it? I've had

20:33

people and believed messages saying, oh my

20:35

God, this is great. I loved it

20:37

every second of it. And then other

20:40

people going, I will never come near

20:42

a boat again, you know. Hey folks,

20:44

Dan here. So my new book, reset,

20:46

came out recently. We actually ran an

20:49

excerpt from the audio book last week,

20:51

if you missed it. And the book

20:53

is getting some kind attention. Amazon called

20:55

it one of the best nonfiction books

20:58

of the month. Apple Audio called it

21:00

a must listen for January. Oprah Daily

21:02

called it one of the best self-help

21:05

books of 2025 and Adam Grant called

21:07

it one of ten new books to

21:09

feed your mind this year. And my

21:11

mom and wife really like it and

21:14

encourage you to check it out too.

21:16

So if you are looking to get

21:18

unstuck this year... to move past the

21:21

way things always work and make the

21:23

changes that matter in your life and

21:25

work, go grab it. Reset how to

21:27

change what's not working wherever books are

21:30

sold. And by the way, on a

21:32

different note, stick around until the end

21:34

of this episode. We're trying something a

21:37

little bit new. And for now, back

21:39

to the show. What if you learned

21:41

on the business side of this work

21:43

over time? If you are a better...

21:46

business person today than you were 10

21:48

or 20 years ago, why is that?

21:50

When I started, the tempo of the

21:53

work was much more relaxed and now

21:55

it's much quicker on the turnaround. So

21:57

you have to have all of your

21:59

ducks in a row. Like I have

22:02

a researcher, I have a proffer, I

22:04

have people, a team that I can

22:06

work with, with a publisher. I just

22:08

got offered a big book about World

22:11

War II, The American Army and the

22:13

Philippines, and it's going to be like

22:15

a 15-hour project, but they wanted it

22:18

under a very tight deadline, and I

22:20

still haven't gotten the official text yet.

22:22

So I've already lined up a researcherer

22:24

a researcher, research all the pronunciations. How

22:27

do you say all those names in

22:29

Tagalog? Yep. Right? Or this military terminology.

22:31

We talked about pronouncing names. There's a

22:34

lot of other things. Like, how do

22:36

you say this math equation correctly? This

22:38

chemical formula. And that's why you hire

22:40

the researcher for, unless you have time

22:43

to do it yourself. So the publisher

22:45

comes to you, they say, do you

22:47

want to do this book? And what's

22:50

going through your head? Do they give

22:52

you a word count? Once you smooth

22:54

out the cadence the way people speak,

22:56

eventually everybody has a tempo. In other

22:59

words, how many words per finished hour

23:01

do you narrate? And my rate is

23:03

about 91 to 9200 words per finished

23:06

hour. That's so interesting that you know

23:08

that. So you could go back to

23:10

your own stuff. Go back to your

23:12

chapter, do a word count of the

23:15

chapter, and do the running time, and

23:17

do the math to figure out how

23:19

many words in an hour. So if

23:21

you did a chapter that was 3,000

23:24

words, and you did it exactly in

23:26

20 minutes, that means you're 9,000 words

23:28

an hour. And it changes. I mean,

23:31

you know, more complicated text is slower

23:33

than easier text, right? Fiction narrates slightly

23:35

lower than nonfiction because of all the

23:37

schmacking that goes on. But, you know,

23:40

when you do a book about the

23:42

history of brain surgery, which I did

23:44

a few months ago, that was longer

23:47

because you're doing these huge multisolabic words

23:49

and it slows the rate down. So

23:51

let's say in my rate is 9100.

23:53

And you come to me with your

23:56

text and say, well, my book is

23:58

91,000. And I said, well, it's going

24:00

to be at least 10 hours long,

24:03

at least. And then that helps you

24:05

with your budgeting and me with my

24:07

calendar. And if I can get at the

24:09

booth, say, four to five hours a day,

24:11

so I'm generating at least two finished hours,

24:13

OK, that's a five to six days. So

24:15

then you give yourself a few extra days

24:18

on the back end in case you have

24:20

a train wreck of some kind. And then

24:22

you put it in your production

24:24

calendar. There's a new technology

24:26

that may one day upend

24:28

the economics of audio book

24:31

narration. It's text to voice

24:33

artificial intelligence. So these are

24:35

people's voices that have been

24:37

modeled by AI to read

24:39

text. The voices have gotten a

24:42

lot better in recent years. So

24:44

we thought we'd write a question

24:46

for an AI voice to ask

24:48

Sean. I'm an artificial voice

24:50

built for reading text. I don't

24:52

care if it's read it. or Rilke,

24:55

I'll read it on command,

24:57

and I am indefatigable.

24:59

I do not require food or

25:01

sleep or emotional support,

25:04

and while a voice actor might

25:06

scoff at my very existence,

25:08

most people wouldn't have

25:10

a clue that I'm

25:13

as artificial as astroturf.

25:15

So my question to

25:17

Sean is, in this brave new

25:19

world, do I worry him at all.

25:21

Is it my voice he hears

25:23

echoing in the alleyways of his

25:25

brain in the dark of the

25:27

night? Okay, so thank my therapist

25:29

thanks you for all the extra

25:32

sessions I'm going to be going

25:34

through for the PTSD of that.

25:36

I mean this has to be

25:38

kind of in the air for

25:40

narrators. Oh yeah, absolutely. So let's

25:43

break it down because I deal

25:45

with this with my students all

25:47

the time. So text a voice is

25:49

already here. And it's being

25:52

used for things like newspaper

25:54

articles, magazine things, phone

25:56

trees, and so on, and

25:58

basic learning, e-learning. It's making

26:00

biggie and runs there. So it's

26:02

a tool. And it just also

26:04

depends on what you want to

26:07

do with the tool and what

26:09

is the expectation for the listener

26:11

when they encounter the audio. So

26:13

if it's just learning the fax

26:15

man like a textbook, AI is

26:17

the way to go. To a

26:19

certain degree, because the language that

26:22

the computer used was very simple.

26:24

What happens when they have to

26:26

deal with citations, with charts? with

26:28

all the other stuff that's on

26:30

the page, right? So there's that

26:32

piece of it. You get points

26:34

for being clever. I had no

26:37

idea what was coming when you

26:39

said that. I thought my mother

26:41

going to appear on this. I

26:43

just had a question. Anyway, so

26:45

there's a place for it, and

26:47

it's already here. The thing that

26:49

is interesting with the audiobook world

26:52

is, once again, what is the

26:54

ultimate yardstick? in an audio book

26:56

performance that you want is to

26:58

be entertained. So it's fine for

27:00

the two minutes or whatever it

27:02

was that he was speaking, but

27:04

is that voice going to be

27:06

varied enough for a 10-hour book?

27:09

What makes for an interesting and

27:11

engaging performance is that, well, like

27:13

I tell my students, that their

27:15

narration needs to be consistently inconsistent.

27:17

Because that's the way we actually

27:19

speak. And you'll notice, just in

27:21

this, my response, I've used melody

27:24

rhythm and tempo in a variety

27:26

of ways that force your brain

27:28

to pay attention to me. And

27:30

AI has yet to figure that

27:32

out. And that's about text analysis.

27:34

Someone said, well, AI can tell

27:36

a joke. No, no, they can't.

27:39

It can say words in a

27:41

rhythm that we interpret as a

27:43

joke. It doesn't know that it's

27:45

told a joke. What makes a

27:47

good... voice for an audio book

27:49

narrator. Like a lot of the

27:51

best narrators, they don't have showy

27:54

voices. They don't have like Morgan

27:56

Freeman voices. So what is it

27:58

that makes for a good one?

28:00

Well, there's a couple things happening.

28:02

At the core of it, Miles

28:04

Davis, the jazz trumpeter, once said,

28:06

it takes a long time to

28:08

learn how to sound like yourself.

28:11

And when you hear a really

28:13

good narrator, they sound exactly like

28:15

themselves. They're not putting on a

28:17

voice when they narrate, right? They

28:19

sound like themselves. And that's my

28:21

goal, both as a performer and

28:23

as a teacher, you know, to

28:26

pass that on to my students.

28:28

So that's the first step. And

28:30

then they sound of their generation.

28:32

So you can spot a 20-something

28:34

voice as opposed to a 60-something

28:36

voice. And then it's about the

28:38

acting portion. It's about connecting emotionally

28:41

with the text and responding to

28:43

it. So yeah, there's no special

28:45

voices. In fact... A lot of

28:47

people make that mistake. They come

28:49

to me like, people tell me

28:51

I have a very nice voice

28:53

and I've done radio for 50

28:55

years and I've done this kind

28:58

of voice for this. And I'm

29:00

like, I can't listen to you

29:02

do even a five-hour book. Once

29:04

upon a time there was a

29:06

night and he did this thing

29:08

and I'm like, oh God, no.

29:10

So Sean, we always end our

29:13

episodes with a quick lightning round

29:15

of questions. Here goes, what's the

29:17

most insulting thing you could say

29:19

about an an audio book narratorator

29:21

work? You sound just like AI.

29:23

You haven't heard that, I expect.

29:25

No. I'm very grateful. What's a

29:28

tool specific to your profession that

29:30

you really like using? Speaking of

29:32

AI, there's a product called positron.

29:34

It's a subscription service and what

29:36

it allows you to do is

29:38

you can upload a text depositron.

29:40

It will do a word frequency

29:43

search to help you generate a

29:45

word list to do your research

29:47

with, but also you can upload

29:49

the audio of the chapter you

29:51

just did and the text and

29:53

it will proof it for you.

29:55

So if you said street lighting,

29:57

it was street lamp. What phrase

30:00

or sentence strikes fear in the

30:02

heart of an audiobook narrator? When

30:04

you have to go over the

30:06

corrections, because sometimes it may not

30:08

recognize a foreign language or an

30:10

oddity that I add to add

30:12

in the text for explanation to

30:15

make it listener-friendly. But at that

30:17

speed, it says everybody a ton

30:19

of time and money. What phrase

30:21

or sentence strikes fear in the

30:23

heart of an audiobook narrator? We

30:25

can't get a vocal match on

30:27

your corrections. What does that mean?

30:30

Okay, so when you record, I'm

30:32

like I'm recording this right now,

30:34

let's say I record it and

30:36

there's a certain audio quality to

30:38

it. Okay, so weeks and weeks

30:40

later, I get my package of

30:42

pickups. These are the sentences I

30:45

got wrong in the piece. And

30:47

what you do is you record

30:49

them as one long track, so

30:51

you record each sentence, and then

30:53

the engineer will go back and

30:55

drop it in to... that spot

30:57

with a correction. It's like cutting

30:59

out the tape and then putting

31:02

in the new piece of section

31:04

of audio there, right? Well, there

31:06

are times, for whatever reason, your

31:08

voice may be really off, or

31:10

maybe you're traveling and have to

31:12

use somebody else's studio, or there's

31:14

ambient noise that wasn't there before,

31:17

so you can really hear the

31:19

change in the audio through it

31:21

and becomes distracting. So when an

31:23

engineer comes back and says, We

31:25

have a problem with your audio

31:27

corrections. Oh my God, my, my,

31:29

just chill ran down my spine

31:32

just thinking about it. What's a

31:34

sound specific to your profession that

31:36

you're likely to hear? Oh, ironically,

31:38

we would hear it, but the

31:40

listener never would, and that's things

31:42

like lawnmowers and jet airplanes and

31:44

cars or Canadian geese while you're

31:46

recording. There's no such thing as

31:49

a sound proof booth. The booth

31:51

cancels out the majority of sound

31:53

and inside the booth it's what

31:55

we call sonically dry to make

31:57

it a nice clean recording. But

31:59

if you've got leaf blowers and,

32:01

you know, you're in your house.

32:04

Yeah, I mean, if your neighbor

32:06

is using a leaf blower, like,

32:08

you can't have a career. No,

32:10

you're host. Oh, no, you do.

32:12

You just start working at night.

32:14

When I was married, my ex,

32:16

we were both narrators. And there

32:19

were times when, depending on time

32:21

of year or what was happening

32:23

in the neighborhood, our schedule just

32:25

got flipped. And we still had

32:27

to get the kids to school

32:29

and feed them, but we worked

32:31

all night and slept during the

32:34

day, because you just didn't have

32:36

a choice. You were on deadline,

32:38

and next door they're doing, you

32:40

know, right in front of your

32:42

house, they're doing road work from

32:44

nine to five. You're hosed. So

32:46

it's the noises, actually, the listener

32:48

never hears, that are the ones

32:51

that resonate most with the narrator.

32:53

When I was narrating my book,

32:55

we had to stop several times

32:57

because my stomach was growling so

32:59

loud they could hear it. I

33:01

never thought of like eating as

33:03

a preemptive problem-solving intervention. Yeah. What

33:06

you eat, you know, or depending

33:08

on which end it comes out,

33:10

you know, I belch a lot

33:12

when I narrate, but if you

33:14

get gassy, well, good luck there.

33:16

So... Those are all real things

33:18

that happen. Or, you know, sounds

33:21

like... I've worked with authors and

33:23

I have a checklist and I

33:25

say, eat light on the day,

33:27

wear clothes that aren't what we

33:29

call scratchy. So if you have

33:31

a nice starched shirt on, it's

33:33

going to go shh, shh, shh,

33:36

every time you move your arms

33:38

around. So you want to wear

33:40

sweats like cotton and you want

33:42

to have plenty of water, but

33:44

you don't have too much because

33:46

then you're going to start gurgling.

33:48

And this is so, I love

33:50

these like insider tips. What is

33:53

an aspect of your work that

33:55

you consistently savor? In the world

33:57

of fiction, it's when you really

33:59

play a scene with the characters

34:01

and you go back... and listen,

34:03

and they sound like four different

34:05

people, that you really nailed the

34:08

dialogue, that the subtext is there,

34:10

the voices are unique, and it

34:12

flows at the right melody, rhythm,

34:14

and tempo. In nonfiction, I really

34:16

enjoy when I get a challenging

34:18

piece with dense, you know, really

34:20

well-written, but dense material, and I

34:23

make it accessible to the listener.

34:25

And you can feel it when

34:28

you had a good day. You're

34:30

like, oh, yeah, you get that

34:33

groove with the text, and it

34:35

just starts coming out. And the

34:37

next thing you know, it's not

34:40

the author speaking anymore. It's

34:42

me. And that's a really wonderful

34:44

feeling, that blending of their words

34:47

with my style of speaking, and

34:49

it just locks in. It keeps

34:51

me coming back, you know, year

34:54

after year. is an audiobook narrator.

34:56

He also coaches aspiring audiobook narrators.

34:59

You can find out more about

35:01

his voice over work and his

35:03

coaching at shanpratpresents.com. I had such

35:06

fun talking to Sean. I've narrated

35:08

two audiobooks and he's narrated 1,200

35:11

so I feel like we're

35:13

pretty much peers. A couple of

35:15

things stood out to me about

35:17

his work. It's a job that

35:20

requires such an odd combo of

35:22

skills. You've got the narration side,

35:25

which is all about performance and

35:27

artistry, and then you've got the

35:29

sound recording and editing side, which

35:32

is completely different. Now obviously those

35:34

skills can go together, just ask

35:37

Sean, but there's no reason they

35:39

should go together. I love jobs

35:41

like that. I think of

35:43

them as odd couple jobs. It

35:46

reminds me of the TV meteorologist

35:48

episode and... Remember the way she

35:51

had to be both an expert

35:53

on the science of weather forecasting

35:55

as well as a likable on-screen

35:58

performer? The other thing that's striking

36:00

about being an audiobook narrator is

36:03

that it's a very high autonomy

36:05

job. You can pick what you

36:07

work on and how you work

36:10

on it. I'm thinking of some

36:12

of the other high autonomy jobs

36:15

we've had on the show, the

36:17

mystery novelist, the Christmas Tree Farmer,

36:19

the stand-up comedian, even the long-haul

36:22

truckers to some extent. For the

36:24

young people listening... This is one

36:27

of those things that it would

36:29

be useful to figure out about

36:31

yourself early on in your career.

36:34

Do you need to be your

36:36

own boss? Or do you thrive

36:39

more when you're part of a

36:41

larger team or system? Mastering the

36:43

art of punch and roll, workshoping

36:46

the perfect accents for a host

36:48

of characters, tracking down obscure pronunciations,

36:51

scoping the workload of new projects,

36:53

and using your voice. as an

36:55

instrument to bring an author's world

36:58

to life. Folks, that's what it's

37:00

like, to be an audio book

37:03

narrator. So the new thing I

37:05

mentioned earlier is that we have

37:07

an extra for this episode. First

37:10

time, it's a five-minute clip of

37:12

Sean talking about his work narrating

37:15

the famous David Foster Wallace Literary

37:17

Behemoth Infinite Just. Now it didn't

37:19

fit into this episode, but if

37:22

you're a DFW fan, you might

37:24

enjoy it. There is a link

37:27

to this extra in the show

37:29

notes. A shout out to recent

37:31

Apple podcast reviewers, Megan Bridget, Najwat

37:34

R, Sadie Barcelona, Amabala, Sarah Ellen

37:36

Atwood, Allah Resherst du Tom Perdue.

37:39

Sorry I'm not bruised. Bonita-do? Moody

37:41

M.B. 77, K.E. 287, and California

37:43

Sandman. Thanks to Paul Fowley for

37:46

connecting us with Sean. This episode

37:48

was produced by Matt Purdie, who

37:51

by the way folks also created

37:53

the theme. music for

37:55

the show. for the

37:58

I'm Dan Heath.

38:00

See you next

38:03

time. time.