Echo Labs Style Guide

Hello, and congratulations! If you’re reading this, it means you’ve passed Echo's
rigorous transcription exams with a score that puts you in the 99% percentile of
everyone who applied.

As part of our team working to provide the highest accuracy and with genuine care
for accessibility, we value your feedback on everything from platform features to
functionality. Our work surpasses traditional transcription to provide quality
captioning and ensure equitable access, which is why our contractors make 2-3x
more than other transcription roles.

You have been offered a freelance contract with the potential to work up to 40 hours
a week as jobs become available. With more universities requesting
captioning/transcription services each week, jobs will continue to be added to the
dashboard daily after being initially transcribed through AI. The Development team
is continually working to increase the accuracy of capturing with AI so you have a
complete transcription to work with.

The goal of the style guide is to proactively answer any questions you might have as
a new transcriber about formatting so, together, we can provide a consistent
experience to the schools we work with. We want them to feel that their 100th video
was captioned with the same care and attention as their very first - something that
requires tremendous coordination between all of us, at scale. If you have any
questions or clarifications on the items below, don’t hesitate to reach out to the
internal team directly on Discord or drop a question in the community channels.

Without further ado, the Echo Labs Style Guide:

TABLE OF CONTENTS

Captioning vs. Transcription

Legal Requirements
 1
Quality Captioning

Terms and Definitions

Standards and Expectations

Steps for Job Completion

Speaker Identification

Caption Stacking & Segmentation

Timing & Alignment

Spoken Content Language Mechanics

Common Punctuation Table

Math & Equations

Atmospherics

Common Captions

Challenging Content & Unclaiming

Helpful Hacks

FAQ

Common Captioning Errors

Glossary
 2

CAPTIONING VS. TRANSCRIPTION

Captioning not only transcribes spoken words but also includes additional
information such as speaker identification, sound effects, and other auditory cues.
Captions are synchronized with the audio or video content, appearing on-screen in
real-time, and are primarily used to provide access to individuals who are deaf or
hard of hearing.

On the other hand, transcription involves converting spoken words from audio or
video content into written text. It aims to provide a textual representation of the
spoken content, making it accessible to individuals who may have difficulty hearing
or understanding the audio.

While both transcription and captioning serve to make audiovisual content

accessible, captioning goes beyond simple transcription by providing a more
comprehensive and synchronized representation of the audio content.

JOB RESPONSIBILITIES

● Editing the captions from complete transcriptions

● Adding speaker tags
● Adding atmospherics
● Ensuring accurate caption segmenting and caption stacking
● Synching captions with audio and video
● Ensuring the 64-character limit is adhered to

HELPFUL RESOURCES

● Using a mouse and headset are HIGHLY recommended for synching captions
with audio
● The Google Image reverse feature can assist in identifying speakers or other
content displayed in the video
● Find and Replace extensions on your browser can improve speed for editing
 3

LEGAL REQUIREMENTS
Accuracy: Captions must relay the speaker’s exact words with correct spelling,
punctuation, and grammar with 99% accuracy and no paraphrasing. Captions must
honor the original tone and intent of the speaker. Captions must match background
noises and other sounds to the fullest extent possible.

Time Synchronization: Captions must align with their corresponding spoken words
and sounds to the greatest extent possible. Captions must not proceed too quickly
for the viewer to read.

Program Completeness: Captions must be included from the beginning to the end
of the program to the fullest extent possible.

Placement: Captions must be positioned on the screen without blocking important

content.
 4

QUALITY CAPTIONING
Accurate: We aim for errorless captions.

Consistent: Style and presentation should be uniform for viewer understanding.

Clear: A complete textual representation of the audio, including speaker

identification and non-speech information, provides clarity.

Readable: Captions should be displayed with enough time to be read completely,

synchronized with the audio, and not obscured by (or obscure) the visual content.

Equal: The meaning and intention are completely preserved to maintain equal
access.

TERMS AND DEFINITIONS

● Caption: the converted audio dialogue and sounds into text that appears on a
video and is synchronized with audio to provide equitable access and
meaning
● Caption Segment/Line: a transcribed line of dialogue that is less than
64-characters
● Caption Stacking: Splitting a caption segment/line into two sections, each is
32-characters or less in length.
● Color Block: The corresponding audio recording for each caption.
● Red Line: The red line indicates the 32-character line limit. Any text over this
line should be stacked or segmented appropriately. There should be NO text
beyond the red line at any time, for any reason.
 5

STANDARDS & EXPECTATIONS

At Echo Labs, our standards and expectations underscore our genuine commitment
to delivering quality accuracy with timeliness for students everywhere. Your role in
this mission is critical, and we count on your exceptional skills to maintain our high
standards and uphold our reputation for excellence.

ACCURACY & QUALITY

1. Quality of transcription is critically important. Due to the Americans with

Disabilities Act, all captions must be at least 99% accurate, and our current
standard of quality at EchoLabs is 100% accuracy, in even the most difficult
situations. Schools choose to work with us - because our team understands
how quality captioning helps make education accessible to everyone.
2. The most important standard is to keep captions behind the red line. Do not
go over the red line with anything (punctuation mark, extra space, etc.).
3. We aim to be as transparent as possible and utilize a three-chances review
system to ensure at least 99% accuracy. Each job is graded on a scale of 1-5.

4.
 6
5. A score of 2-3, will receive a strike against your three-chances. The grader will
add robust feedback for improvement and you still have two more chances to
improve. Errors that are scored automatically as 2-3 include:
a. text or characters going over the red line (spaces do not count)
b. no or incorrect speaker tags
c. no atmospherics
d. inappropriate captioning (not captioning exactly what was said)
e. overuse of (unintelligible)
6. When a transcriber reaches three “strikes,” they will lose their freelance
contract with Echo Labs.

TURNAROUND TIME

1. Every video comes with an associated ‘time allotment’ which is the total
amount of time you have to complete a video. You will have a timer running
when accepting the job. We do this to ensure that all videos are delivered to
students within 24 hours of submission (See Time Chart).
2. If work is not finished within the time allotment (when the timer expires) any
work completed will be erased and partial payment is not possible, this is
why we’ve extended the time allotment to 600% of the video length itself.
3. You may request an extension for extenuating circumstances on the
appropriate Discord channel.

STEPS FOR JOB COMPLETION

For accepting and completing a job with Echo Labs, follow these steps. We
recommend using a mouse and headset to simplify your editing.

1. Navigate to the Echo Labs dashboard portal: https://portal.el.ai/

2. Log in and you will see jobs as blue buttons on the dashboard with the video
length, allotted time, and payout.
3. Once you select a job and click the button, the editing dashboard will open.
4. Complete the Job Tracker Form with the requested information.
 7

SPEAKER IDENTIFICATION
Whether or not a speaker can be identified using visual cues, we must take the time
to include accurate speaker tags for students who rely on end transcripts as a guide.
Here’s the best way to prepare speaker tags for all users.

1. Speaker tags should always be enclosed in brackets with the speaker title
within it being written in full uppercase, e.g. [RACHEL ADAMS]
2. Please attempt to find the speaker’s full name by searching on the university
webpage, professional websites, or completing an internet search.
3. Maintain the same speaker tag throughout the video. Do not truncate the
speaker tag after the first use.
4. Please don’t use [SPEAKER ONE, TWO, THREE, ETC.] labels. This will result in
an automatic strike. Most of the time, you can see the names in the video,
glossary, or – if not – can infer from the video their position (e.g. STUDENT,
INSTRUCTOR, STUDENT ONE, TWO, etc.)
5. Use full names whenever possible, and for each speaker switch. Do not use
first names alone when first and last are available.
6. Speaker tags are required each time the speaker switches. If one speaker
continues talking across multiple caption segments, you only need to include
a speaker tag in the first caption segment of the running speech.
7. When numbering speakers, write out one through ten and then switch to
numerals for speakers 11+.
8. Follow this hierarchy for labeling speaker tags: 1) Full Name, 2) Role either
implicit or intuited.
9. When there are several speakers in one video, use the appropriate tag
according to their identity, not their speaking order in the video. For example,
if speaker one is named [DR. SMITH] and speaker two is not named specifically
they will be tagged with number one such as [STUDENT ONE]. Subsequent
speakers will be tagged either by their name if given, or the subsequent
number.
 8
10. Include given titles, if known, and use appropriate abbreviations, e.g. DR. MRS.,
MS., MR., PROF., etc.

COMMON SPEAKER TAGS

General Tags Synonyms for Professor

● [NARRATOR] ● [PROFESSOR]
● [HOST] ● [INSTRUCTOR]
● [COMMENTATOR] ● [LECTURER]
● [INTERVIEWER] ● [FACULTY]
● [INTERVIEWEE] ● [SCHOLAR]
● [ACADEMIC]
● [EDUCATOR]
● [TEACHER]
● [ADVISOR]

Synonyms for Student Synonyms for Moderator

● [STUDENT] ● [MODERATOR]
● [LEARNER] ● [FACILITATOR]
● [PUPIL] ● [CHAIRPERSON]
● [SCHOLAR] ● [MEDIATOR]
● [UNDERGRAD] ● [COORDINATOR]
● [GRADUATE] ● [HOST]
● [APPRENTICE] ● [PRESENTER]
● [TRAINEE] ● [DISCUSSION LEADER]
● [PARTICIPANT] ● [CONVENOR]
● [EMCEE]

● [VOICE OVER]
Special Roles ● [NEWS ANCHOR]
● [GUEST]
● [EXPERT]
● [PANELIST]
● [CONSULTANT]
● [KEYNOTE SPEAKER]
● [TUTOR]
● [AUDIENCE MEMBER]
 9

CAPTION STACKING & SEGMENTATION

As per legal requirements, all caption segments should contain less than 64
characters total. For an example of basi segmenting, check out this helpful video.

● As part of recent updates, we’ve made it easier to know how long each
segment should be, the red line indicates when you have reached
32-characters. If the caption is over the red line, it will need to be stacked
accordingly.
● When the caption moves beyond the red line, it must be split into two lines.
This is why we’ve created ‘caption stacking’, which you activate by clicking
SHIFT + ENTER while typing. The segment will be broken into two lines to
make it more readable. The caption segment should NOT be split into more
than two lines.
● Deciding where to stack captions is very important. The goal is to stack
captions so that whole phrases, nouns, sentences, and dialogue flow are
interrupted as minimally as possible.
● Readability is extremely important to segmenting and stacking. Make sure to
watch and review the video captions to ensure the segments and caption
stacks are not divided over too many screens visually which can impede the
student’s readability.

When a sentence is broken into two lines of captions, it should be broken at a logical
point where speech normally pauses and follows standard grammar constructions.
The following guidelines provide examples of appropriate segmentation.

● Do not break a modifier from the word it modifies. Example:

Inappropriate

Mark pushed his black

truck.

Appropriate

Mark pushed
 10
his black truck.

● Do not break a prepositional phrase. Example:

Inappropriate

Mary scampered under

the table.

Appropriate

Mary scampered
under the table.

● Do not break a person's name or a title from the name with which it is
associated. Example:

Inappropriate

Bob and Susan

Smythe are at the movies.

Appropriate

Bob and Susan Smythe

are at the movies.

● Do not break a line after a conjunction. Example:

Inappropriate

In seconds she arrived, and

he ordered a drink.

Appropriate

In seconds she arrived,

and he ordered a drink.

● Do not break an auxiliary verb from the word it modifies. Example:

Inappropriate

Mom said I could

have gone to the movies.

Appropriate
 11
Mom said I could have gone
to the movies.

● Never end a sentence and begin a new sentence on the same line
unless they are short, related sentences containing one or two words.
Example:

Inappropriate

This is an example.
You're not supposed to do this.

Appropriate

This could be one caption.

Right?

● If music or atmospherics occur before or during speaking, segment

the caption so that the atmospheric is on the second line unless the
sound happens with an action.

Inappropriate

[INSTRUCTOR] (opera singing)

Today’s class will cover opera.

Appropriate

[INSTRUCTOR] Welcome everyone,

(music continuous)

Task's assigned,
and (smacks hands) done.1

1
Adapted from the “Captioning Key.” Described and Captioned Media Program. 2024.
 12

TIMING & ALIGNMENT

The start and end times of each caption segment need to align with the beginning
and ending of the sound. While you should aim for precision, it’s ok for the start time
to be up to ½ second early from the true beginning of the sound. Slow down or
speed up the sound to ensure start and end times are marked as accurately as
possible.
 13

SPOKEN CONTENT LANGUAGE MECHANICS

Language mechanics incorporate the proper use of spelling, capitalization,
punctuation, grammar, and other factors deemed necessary for high-quality
captioned media. Rules included in these guidelines are primarily those that are
unique to captioning and speech-to-text. For conventional words, dictionaries and
style guides must be followed. Proper names, technical terms, and specialized
language must be verified through specialty references or directly from an
authoritative source.

GENERAL MECHANICS GUIDELINES

1. Maintain the integrity of the spoken words. When editing occurs, each
caption should maintain the original narration's meaning, content, and
essential vocabulary.

2. Lightly edit spoken content to avoid filler words, such as ‘uh’ and ‘um’. You
may remove filler words, such as “you know,” “well…” or “um” and other
non-essential information.

● Omit speech disfluencies: unnecessary filler words, false starts,

stutters, repetitions, etc.
● Omit quick interjections, such as an interviewer saying
“mm-hmm”, unless a direct response to a question.
● Correct egregious phonetic and pronunciation errors that inhibit
readability. However, never change the story.
● Don’t correct a speaker’s grammar or pronunciation that is easily
understood. E.g., “gonna” must stay as “gonna”.
● NEVER censor or edit expletives. If the word is censored with a
beep sound, use (beep) or (bleep) in-line where the sound occurs.
● Never omit special words, entire sentences, or expletives.
● Don’t be excessive. If in doubt, don’t omit the word(s).
 14

3. If the filler or sound word adds meaning, please caption it. The chart
below indicates common filler words that might communicate specific
meanings.

Common Filler Words With Meaning

Well Sort of Like Actually

Basically Seriously Literally Totally

Clearly You see You know I mean

You know what I At the end of the Believe me I guess

mean? day
Or something Okay So I suppose

And Kind of

4. The most important concept for “captioning accessibility” is to capture the

speaker’s meaning and the overall readability of the captions. This is why we
add atmospheric sounds and audible speaking/utterances that add significant
meaning. For example, a speaker using “umm” or “uh” as a filler word does not
necessarily add meaning. However, if the “umm” is longer than 3-5 seconds,
not captioning this sound might detract from the meaning for the student
since there would be a significant length of time without a caption.

5. When a speaker uses grammatically incorrect language or a dialect, it

should be reflected in the captions.

6. Hesitations should be replicated in the captions - long pauses should be

shown using ellipses (...), and shorter pauses should be shown using commas.

7. Caption contractions, formal and informal, as spoken.

8. You are expected to research proper nouns and terminology for

representation and proper spelling (See Glossary as needed for common
spellings).
 15
9. Use the spelling of words that are included in the video, e.g. PowerPoint
presentations or written on the board.

10. Googling with a bit of context from your video/audio is also helpful such as
speaker’s names, or academic discipline.

11. URLs, hashtags, social media tags should be captioned using common
convention: www.el.ai/#echolabscaptions/@echolabs

12. Never type out an already censored word.

13. Use an appropriate atmospheric for the sound heard when the word is
censored, e.g. (beep).

14. Include proper punctuation per common English grammar rules.

 16

SPELLING AND PUNCTUATION

Common Punctuation Guidelines

,
Used when listing, to separate clauses, after filler words,
before quotes, and when addressing someone. No space
before and one space after.

[]
Use for speaker tags.

.!?
Used at the end of whole sentences. No space before
and one space after.

‘
Used in contractions and to indicate a possessive. No
space before or after.

‘’ Use for short quotes, answers, and media titles. Also, for
emphasis when a speaker is emphasizing a specific
term, phrase, or quote. Wrap words at the beginning
and the end of the quote.

“”
Use for long and direct quotes. Wrap words at the
beginning and end of the quote.

–
Use an Em-dash/double hyphen for additions or asides,
such as when a speaker interrupts themselves or
changes the direction of the conversation. Use a space
before and a space after.

-
Use a single dash/hyphen for hyphenated words.

…
Use when there is a significant pause, (longer than 5
seconds) or interruption by another speaker. Use only
when necessary.

:
Use to give emphasis, present dialogue, introduce lists
or text, and clarify composition titles.

;
DO NOT USE.
 17

1. When in doubt, check the word in the dictionary: Merriam-Webster:

America's Most Trusted Dictionary, The Oxford English Dictionary, or the
Glossary of standard academic vocabulary.

2. Do not emphasize a word using all capital letters except to indicate

screaming or shouting.

3. Be consistent in the spelling of words throughout the media. This includes

vocabulary that can be spelled either as one or two words or in hyphenated
form.

4. Capitalize proper nouns for speaker identification. All other speaker

identification should be lowercase unless this identification is being used as a
proper noun.

5. Lowercase sound effects, including both description and onomatopoeia,

except when a proper noun is part of the description.

COMMAS

● When captioning a list separated by commas, use a serial, or Oxford,

comma. Example:

Inappropriate

I'm having eggs, bacon and toast.

Appropriate

I'm having eggs, bacon, and toast.

HYPHENS & DASHES

● When captioning spelling (including fingerspelling), separate capital

letters with hyphens. Example:
 18
A-N-T-I-O-N-E-T-T-E

ELLIPSES

1. Use an ellipsis when a caption has a significant pause, longer than 5 seconds.

2. Do not use an ellipsis to indicate that the sentence continues into the next
caption.

3. Use an ellipsis to lead into or out of audio relating to an onscreen graphic.

QUOTATION MARKS

1. Use quotation marks for on-screen readings from a song, poem, book, play,
journal, or letter. However, use quotation marks and italics for offscreen readings
or voice-overs.

● Beginning quotation marks should be used for each caption group of

quoted material except for the last caption. The last caption should
have only the ending quotation mark. Example:

Reading from a journal…

(Caption group 1) "Mother knelt down

and began thoughtfully fitting

(Caption group 2) “the ragged edges

of paper together.

(Caption group 3) The process was watched

with spellbound interest."

SPACING

1. Spaces should not be inserted before the ending punctuation, after opening
and before closing parentheses and brackets, or before/between/after the periods
of an ellipsis.

● A space should be inserted after the beginning music icon (♪) and
before the ending music icon(s). Example:
 19
♪ There's a bad moon rising ♪

MUSIC

● When captioning music, use objective descriptions that indicate the

mood. Avoid subjective words, such as "delightful," "beautiful," or
"melodic."
● If music contains lyrics, caption the lyrics verbatim. The lyrics should be
introduced with the name of the vocalist/vocal group, the title (in
brackets) if known/significant, and if the presentation rate permits.
● Caption lyrics with music icons (♪). Use one music icon at the beginning
and end of each caption within a song, but use two music icons at the
end of the last line of a song.
● A description (in parenthesis) should be used for
instrumental/background music when it's essential to the
understanding of the program.
● Offscreen background music description should be italicized.
● If possible, the description should include the performer/composer and
the title.
● Beware of misplaced modifiers in descriptions.

Incorrect

(frantic piano playing)

Correct

(frantic piano music)

● Nonessential background music should be captioned by placing a

music icon (♪) and a brief description.
● Do not caption background music with a duration under 5 seconds.
● If music contains lyrics, caption the lyrics verbatim. The lyrics should be
introduced with the name of the artist and the title in brackets, if the
presentation rate permits.

Correct Examples
 20
[Ella Fitzgerald singing
“Old MacDonald Had a Farm”]

[The Beatles singing

“Come Together”]

[Fred Rogers singing

“Won’t You Be My Neighbor”]

● Caption lyrics with music icons (♪). Use one music icon at the beginning
and end of each caption within a song, but use two music icons at the
end of the last line of a song. A space should be inserted after the
beginning music icon (♪) and before the ending music icon(s).

Correct Examples

♪ I’m pickin’ up good vibrations ♪♪

♪ And I can’t wait to get

on the road again ♪♪

● Use descriptions that indicate the mood. Be as objective as possible.

Avoid subjective words, such as “delightful,” “beautiful,” or “melodic”.

Correct Examples

(monotone boy croons)

(hypnotic orchestra serenades)

(ethereal choir harmonizes)

(animated quartet chimes)

(soporific voice mesmerizes)

NUMBERS

● Unless otherwise specified below, spell out all numbers from one to
ten, but use numerals for all numbers beyond ten. Examples:
 21
Inappropriate

The fifty-four DVDs need to be shelved.

Appropriate

The 54 DVDs need to be shelved.

Inappropriate

He's at the thirty, the twenty, and scores!

Appropriate

He's at the 30, the 20, and scores!

● Spell out any number that begins a sentence as well as any related
numbers. Example:

Two hundred guests and eleven guides entered.

● Spell out casual, nonemphatic numbers. Example:

He gave me hundreds of reasons.

● Numerals with four digits can either have a comma or not. Be

consistent throughout the media production. For numerals having over
five digits, a comma is necessary. Example:

Inappropriate

50000

Appropriate

50,000

● Use numerals in a listing of numbers if one or more is above ten and

these occur in one caption or one sentence. Example:

Inappropriate

Steven has 21 books, 11 oranges, and three cats.

Appropriate
 22
Steven has 21 books, 11 oranges, and 3 cats.

● Use numerals when referring to technical and athletic terms. Example:

He scored 3 goals in today's game!

● When indicating sequence, capitalize the noun and use numerals.

Exceptions are the indication of line, note, page, paragraph, size, step,
or verse. Examples:

Building 2 page 31

Channel 5 size 12

Chapter III step 3

Room 438 paragraph 2

LOCATION NUMBERS

● Use numerals for location numbers such as phone numbers and zip
codes. Example:

Inappropriate

One, one, one, five, five five, one five, one five

Two zero zero four eight

Appropriate

(111) 555-1515

20048
 23
DATES

● Use the numeral plus the lowercase "th," "st," or "nd" when a day of the
month is mentioned by itself (no month is referred to). Example:

Original Narration

"ninth"

Captioned As

Bob went fishing

on the 9th.

● When the day precedes the month, use the numeral plus the lowercase
"th," "st," or "nd" if the ending is spoken. Example:

Original Narration

"seventeenth"

Captioned As

My birthday is the 17th of June.

● Use the numeral alone when the day follows the month. Example:

Original Narration

"nine" or "ninth"

Captioned As

I will meet you on May 9.

● When the month, day, and year are spoken, use the numeral alone for
the day, even if an ending ("th," "st," or "nd") is spoken. Example:

Original Narration

"six" or "sixth"

Captioned As

Paul will marry on July 6, 1996.

 24
● Write the year using numerals. Examples:

1907

● Years that are shortened should always include an apostrophe before

the year. Example:

Original Narration

June of 1999

Captioned As

June of ‘99

TIME

● Indicate the time of day with numerals only. Examples:

I awoke at 5:17.

If you wish to attend, you must arrive by 6:25 p.m.

We were expected to report no later than 1400 hours.

I awoke at 4 o'clock.

● Always use numerals when the abbreviation "a.m." or "p.m." is present.

Double zeros are not necessary to indicate minutes of the hour when a
whole number is used with a.m. or p.m. Examples:

She leaves at 3:20 p.m. for the airport.

Our hours are from 9 a.m. to 5 p.m.

We're leaving at 6 in the morning.

PERIODS OF TIME

● A decade should be Captioned As "the 1980s" (not "the 1980's") and "the
'50s" (not "the 50's").
 25
● Do not use hyphens if a decade or century is in noun form. Example:

This vase is from the 17th century.

● If a period of time is used as an adjective, use a hyphen. Example:

This 19th-century painting was done by Van Gogh.

FRACTIONS

● Either spell out or use numerals for fractions, keeping this rule
consistent throughout the media. If using numerals, insert a space
between a whole number and its fraction. Example:

Numeral Used

Do you plan to eat 1 ½ pizzas?

Fraction Spelled Out

Do you plan to eat one and one-half pizzas?

● Do not mix numerals and spelled-out words within the same sentence.
Example:

Inappropriate

Malika is 13 and a half years old.

Appropriate

Malika is 13 ½ years old.

● If a fraction is used with "million," "billion," "trillion," etc., spell out the
fraction. Example:

The population was over one-half million.

● Fractions expressed in figures should not be followed by endings, such

as "sts," "rds," "nds," or "ths." Example:

Inappropriate
 26
3/10ths

Appropriate

3/10

DOLLAR AMOUNTS

● Use the numeral plus "cents" or "¢" for amounts under one dollar.
Examples:

I need 15 cents.

I owe you 32¢.

● Use the dollar sign plus the numeral for dollar amounts under one
million. For whole-dollar amounts of one million and greater, spell out
"million," "billion," etc. Examples:

Barbara brought only $11.

Bob brought $6.12.

The budget of $13,000 will be sufficient.

Taxes will be reduced by a total of $13 million.

He owes $13,656,000.

● Use the word "dollar" only once for a range up to ten. Example:

I hope to find three or four dollars.

● Use the dollar sign and numerals when captioning a range of currency
over ten dollars. Example:

Alice expected a raise of $6,000 to $7,000.

MEASUREMENTS

● Spell out units of measurement, such as "inches," "feet," "yards," "miles,"

"ounces," "pounds," and "tablespoons." However, if spoken in a
 27
shortened form, symbols should be used. For example, if the Original
Narration is "I'm five eight," it should be Captioned As:

I'm 5'8".

● For whole numbers, use numerals. For example, caption "3 cups of
sugar" instead of "three cups of sugar."

DECIMALS

● Spell out decimals, such as one point four or nine point nine seven.
Unless the content is math specific. Then use integers such as 1.4 or
9.97.

MATH & EQUATIONS

● When captioning math specific content (arithmetic, algebra, equations,

fractions, etc.) always use integers instead of spelling out the number
words even for numbers 1-10.
● Follow basic number rules; use numerals for fractions (since a fraction
uses more than one digit). Example:

One plus one equals two.

2,000 plus three equals 2,003.

1/2 of this and 10 3/4 of that.

● Follow the conventions for ordinal place. For example: “Zeroth”, “first in
line”, “tenth place”, “21st century”, “100th time”.
● For anything that’s not a number / digit or percentage, write out the full
word instead of the symbol. Example:

X squared times 0.32%.

The derivative of y squared.

 28
Five x times negative three x equals negative 15x squared.

● A speaker may write this on a whiteboard: while speaking the equation

aloud. You would Captioned as:

Percentage of the bankroll equals

the odds received,

multiplied by probability of winning,

minus probability of losing,
divided by odds received.

GRAPHING TERMS

● Write it out as the speaker says it, following basic number conventions
and using integers instead of writing out the words (-10,3).
● Quadrants are labeled with Roman numerals, such as “quadrant IV”
● Axes and coordinate references are hyphenated as follows: x-coordinate,
y-axis.

ACRONYMS

● Only type an acronym if spoken that way by the speaker.

● For acronyms like FOIL (First, Outer, Inner, Last, used for binomial
multiplication), be sure to keep it capitalized for all uses (e.g., “Now you
try FOILing this next problem”).
● To make an acronym plural,, add a lowercase “s” without an apostrophe
(e.g., SATs,), unless the acronym ends in an S, in which case, add an “es”.
● Do not abbreviate unless the speaker specifically says it. Example:

Appropriate

Gigabytes

Inappropriate

gigs or GB
 29
● If a speaker spells out a word, use the format “W-O-R-D”.

NON-LETTER SYMBOLS

● Do not use symbols or special characters such as é, £, €, or ². Only use

what is available on a standard American keyboard. This includes Greek
letters such as alpha, beta, gamma. Write each of these as words.
● Non-letter symbols, such as pi, should have spaces in between them
and both the preceding and next variable or term.
● Try to be as clear and consistent as possible using spaces as needed to
avoid confusion, such as pi being mistaken for p times i. Example:

Appropriate

2 pi r

2 times pi times r

Inappropriate

Twopir

● Spell out all units (joule, gram, ampere, volt, meter, pascal, kelvin, hertz,
coulomb and newton.)
● Spell out all functions such as “f of x” instead of f(x). When referring to
notations such as dy/dx (and all other related derivative references) in calculus,
engineering, etc, have the captions reflect what the speaker says (e.g., “dy dx”,
including a space in between). Example:

SUBSCRIPTS & POWERS/EXPONENTS

● Treat the subscript or superscript term the same way you might another term
like pi, tau, or sine, all together. Denote these as the speaker reads it. Example:

x2

x sub 2
 30
xj2

● Anything to a power should be written as spoken. Use the integer plus the
lowercase "th," "st," or "nd" if the ending is spoken. Example:

9 squared

y to the 10th

12 to the 2nd

x to the 50th power

PROGRAMMING LANGUAGE

● HTML tags: Write them between backticks, like `<b>`

● Keywords: Write them between backticks, like `function`
● Write out “underscore” when necessary.
● Coding and programming language: Write out “dot” when speaker is
describing coding vocabulary, like pd.concat, “pd dot concat.”
 31

ATMOSPHERICS
Captions need to indicate sounds heard on screen. We call these identifiers
atmospherics. These are critical in providing visual indicators of non-verbal sounds to
viewers who may be hard of hearing.

1. Sound effects describe any meaningful, audible content that is not

spoken.

2. The sound should add important context and meaning to the video such
as sounds that:

Interrupt the speaker

Result in a reaction (laughing, gasping, etc.)

Convey emotion

Include gesturing or body language

Intentional speaker sounds (tutting, mumbling, tongue click, etc.)

Include unintelligible reactions (class chatter, student interrupts)

Typing or clicking when the person is describing or discussing the topic

Background music

3. Avoid adding atmospherics that are excessive and do not add meaning:
(lip smacking, mouse clicking, pen clicking, etc.).

4. Use parentheses ( ) and lowercase unless a proper noun is used for all
atmospherics, e.g. (Erin sketching).
 32
5. Describe the sound or sounds heard on screen by following this
convention: noun + descriptor/verb in the present tense form, e.g. (water
boiling), (door slams).

6. Don’t use onomatopoeia e.g. (ribbit ribbit); instead, describe what’s

creating the sound, e.g. (frog croaking).

7. When a video also contains spoken words, only include background music
if there’s a significant time gap and it would benefit the viewer, e.g.
(background music).

8. Try and be as detailed as possible with your music atmospherics, e.g.

(whimsical piano music).

9. Avoid subjective adjectives. Instead, opt for objective descriptors, e.g.

Objective = Unbiased Observation, Subjective = Personal Evaluation
 33

CHALLENGING CONTENT & UNCLAIMING

You should always do your best to caption all spoken words. For extremely
challenging content, follow the guidelines below.

1. If an occasional word cannot be understood, use (unintelligible) in place of the

word. (unintelligible) should only be used if you absolutely cannot determine the
word.

2. Words in another language should not be translated. Add an appropriate

atmospheric (speaking French).

UNCLAIMING A JOB

Currently, the only reason to unclaim a job is if a job is extremely poor audio
recording. When you encounter this, unclaim the job, and notify the IT
troubleshooter. When in doubt, please email support@el.ai first for guidance.

If there are repeated words, incorrect wording, or incorrect lyrics to music, please
correct and complete the job.
 34

SYNCING AND EDITING

Most of your job with a job will be to edit captions to the accurate length and ensure
speaker tags, atmospherics, and grammar/punctuation are correct.

Syncing the audio accurately with the captions is vital for understanding. The color
blocks at the bottom of the screen align the sound waves with the caption. If the
caption extends beyond the red line, it needs to be stacked to the next line and the
color blocks will be adjusted so that each caption corresponds to the accurate color
block of sound.
 35

● There should be no gaps between color blocks unless there is no sound

whatsoever (as seen below between caption 580 and 581).

● You can access the video timeline by right-clicking the video screen, and
selecting “show all controls.”
 36

FAQ
Q - For long pauses, do you need to include an atmospheric tag such as (silence) or
(no audio)

A - No, this is not necessary. The screen can be caption-free when there is no one
speaking and then have the captions resume when the sentence resumes.

Q - How do I caption speaker interruptions?

A - Use an ellipsis when there is a significant pause, (longer than 5 seconds) or

interruption by another speaker. Use only when necessary.

Q - For math lectures, should you write out the equations using words or the way a
professor is writing them out?

A - Write the equations using the words spoken by the professor.

Q - Do we start the audio blocks when there is an uh or um, or exclude them from
the audio blocks and start with the typed text?

A- Start with the typed text, ignore the ums, and uh.

Q - Do we want the caption blocks to fill the whole bar or just where speaking is
occurring?

A - We only want to have captions appear where the speaker is talking. Try to match
the cadence of speech as much as possible.

Q - When the speaker stutters or repeats words, do we start the caption block where
they first start speaking or where they say it correctly?

A - Start where they say it correctly to optimize the readability for the students.

Q - When the speaker trails off and doesn't finish a sentence should we use ...?
 37
A - We should use … to indicate that the sentence wasn't completed. This is being
updated in the AI soon. If it is just a pause in the sentence and they continue the
sentence, there is no need for ... or -, the screen can just be caption-free until they
resume speaking.

Q - If a video begins with music before the speaker does that need to be captioned
or not?

A - Yes! And please be as specific as possible, i.e. whimsical piano music, rock music,
etc.
 38
COMMON CAPTIONING ERRORS

Below, are several examples of inappropriate captioning and common errors that
have been submitted for review.

● This is an example of inappropriate atmospherics and no speaker tags.

 39
● This is an example of an incorrectly spelled name. The transcriptionist did not
research the spelling of the professor.

● This is an example where the transcriptionist did not ensure the caption block
continued with the speaker before submitting.

● The caption group is over the red line, over 32 characters.

 40

● The caption group is not segmented appropriately

and there are spelling errors.
 41

QUICK GUIDE

ITALICS (As of 4/21, there is no ability to italicize but it will be coming soon!)

1. A voice-over reading of a poem, book, play, journal, letter, etc. (This is also
quoted material, so quotation marks are also needed.)

2. When a person is dreaming, thinking, or reminiscing and the dialogue is

internal.

3. When there is background audio that is essential to the plot, such as a PA

system or TV.

4. The first time a new word is being defined, but do not italicize the word
thereafter.

5. Offscreen dialogue, narrator (see Exception 2 below), sound effects, or music

(this includes background music).

6. Offscreen narrator when there are multiple speakers onscreen or offscreen.

7. Speaker identification when the captioned dialogue is in italics.

8. Foreign words and phrases, unless they are in an English dictionary.

● When a particular word is heavily emphasized in speech. Example:

You must go!

Exceptions to the use of italics include:

1. When an entire caption is already in italicized format, use Roman type to set
off a word that would normally be italicized.
 42
2. If there is only one person speaking throughout the program (including the
narrator), whether onscreen or offscreen, use Roman type with no italics.

3. Do not italicize when a person who is offscreen is translating for a speaker who
is onscreen.