The Narrator The Narrator

Created by David Winter

(“The Narrator”)

https://www.thenarrator.co.uk

The Narrator’s Rest

Audiobook narration FAQ

The Discord FAQ, with a little more room to read.

Warning: This is a work in progress

Choose a question to read its answer.

Getting started

01Is an audiobook narrator also responsible for producing the finished audio?

On ACX, normally yes. The narrator working as the producer is responsible for delivering the finished audiobook: recording, editing, proofing, pickups, mastering and correctly formatted files. Those tasks can be outsourced, but their cost and coordination remain part of the production.

Publisher and studio jobs may be narration-only, with an engineering team handling post-production. The contract should identify the deliverables before a rate or deadline is agreed.

That distinction affects almost everything that follows.

02Is a good voice enough?

No. Audiobook narration requires acting, comprehension and sustained attention.

A narrator needs to communicate what the text means, follow changes of thought and make dialogue believable. Non-fiction requires those skills too: a complicated explanation needs to sound understood, rather than merely pronounced correctly.

An attractive voice may draw someone’s attention. It will not compensate for monotonous delivery, misplaced emphasis or a failure to understand the book.

03What should a beginner do first?

Complete a short practice production before accepting a book.

A useful exercise is to prepare and record enough material for around 15 minutes of finished audio, then edit it, proof it against the manuscript, record the corrections and export the finished result. Time every stage.

That reveals the actual work involved. It also produces something useful for a coach or engineer to assess. Buying more equipment before identifying the weaknesses is an expensive way to guess.

04What training should come first?

Acting and text interpretation should be priorities. Audiobook coaching can then address the particular demands of long-form storytelling: narrative perspective, character distinction, pacing and sustaining a performance.

Technical training belongs alongside that work. A narrator producing through ACX needs to learn both performance and production.

A good teacher identifies specific problems and provides exercises or techniques for addressing them. Repeated encouragement without useful criticism is poor value, however pleasant the sessions feel.

05Which recording software and course are recommended?

REAPER is the recommended starting point for narrators establishing an audiobook production workflow. Its non-destructive editing, punch-and-roll recording, automation and chapter-export facilities make it particularly useful for long-form work.

David Winter’s free Recording Audiobooks with REAPER provides a practical route through setup, recording, processing, proofing, pickups and delivery. It gives narrators a working method to learn before they start customising everything.

REAPER has a 60-day evaluation. Its discounted licence currently costs US$60 before applicable taxes, subject to eligibility; the software remains the same whichever licence applies. REAPER licensing

Studio & software

06What should the first equipment budget prioritise?

A quiet, acoustically controlled recording space comes first. Then a suitable microphone and interface, a stable stand, pop protection and headphones that make faults easy to hear.

A moderately priced microphone in a good space can produce professional results. A costly microphone in a reflective room still records the reflections.

Before replacing equipment, a narrator should obtain an assessment of an unprocessed recording. That makes it possible to distinguish problems with the room, microphone position, recording level and equipment from problems introduced by processing.

07Will acoustic foam or a microphone shield soundproof a room?

No. Acoustic treatment reduces reflections inside a space. Sound isolation reduces noise entering or leaving it.

A blanket booth may improve an echoey recording while doing little about a passing motorbike. Thin foam and a small shield around the microphone are not substitutes for treating the recording space.

The first practical steps are to identify noise sources, choose the quietest workable location and control the reflective surfaces around the narrator and microphone.

08Must the microphone use XLR?

An XLR microphone and separate interface are a sensible default for a permanent studio because they offer flexibility and straightforward equipment replacement.

However, USB is not a verdict on audio quality. Some studio microphones, including the RØDE NT1 5th Generation, offer both connections. RØDE specifications

Existing equipment should be judged through a recording before it is replaced. A specific studio’s equipment requirements still matter when applying to work with that studio.

09Which plugins are worth buying?

For narrators handling their own audio cleanup, iZotope RX Elements is borderline essential and should be included in the initial software budget.

Its most immediately useful tools are Voice De-noise, for reducing background noise, and De-click, for repairing clicks. Cleaning every small problem manually becomes expensive in working time across an entire audiobook.

RX Standard is a useful upgrade for regular production, particularly for its dedicated Mouth De-click and additional repair facilities. Mouth De-click is not included in Elements. RX edition comparison

REAPER’s included processing can cover the basic EQ, compression and limiting requirements. A large collection of expensive effects is not a prerequisite.

10Can a narrator copy someone else’s FX chain?

A chain can provide the order of operations and a useful starting point. Its settings still need adjusting for the actual recording.

A gate threshold that works for one narrator may remove quiet words from another. An EQ setting that reduces muddiness in one voice may make another sound thin.

A practical starting chain addresses noise and clicks, corrects tonal problems, controls level variation and limits peaks. Each stage should earn its place through an audible improvement. Processing should be compared at similar playback loudness.

Auditions & finding work

11When is a narrator ready to audition?

When the audition demonstrates a performance and recording standard that can be sustained throughout the book.

That means more than producing one good paragraph. The narrator should be able to record a longer passage, correct mistakes, maintain character choices and deliver the promised audio quality.

A short assessment from an audiobook coach and an audio engineer can prevent months of submitting auditions with the same unnoticed problems.

12Should audiobook auditions be raw or mastered?

For a normal ACX audition, submit edited and mastered audio that represents the intended finished audiobook. The rights holder is choosing a producer as well as a performer. ACX explicitly advises that audition quality should indicate the final production quality. ACX audition guidance

This is an important difference from voice-acting work where an engineering team requests raw recordings.

For publisher or studio auditions, follow the submission instructions. If raw audio is requested, leave processing off. “Raw” does not normally mean leaving in stumbles and abandoned takes: editing mistakes and processing the sound are separate operations.

13How long should an audition be, and should it include a slate?

Follow a reasonable audition brief. A few minutes should demonstrate the performance; an entire chapter should not be recorded simply because someone uploaded one as the audition script.

Equity recommends audiobook audition samples of no more than five minutes. A substantially longer request deserves clarification before that time is committed. Equity audiobook guidance

Unless a slate is requested, begin with the audition text. Introductions, availability and relevant experience belong in the accompanying message.

14Does an audiobook narrator need a conventional demo reel?

Audiobook narrators need separate narration samples that demonstrate sustained storytelling.

A practical starting set is three strong samples of roughly two minutes each, showing relevant differences such as fiction narration, dialogue and non-fiction. The selection should match the work being pursued.

A rapid montage of voices, music and effects is a poor substitute. Casting needs to hear how a narrator handles connected prose. Samples should be clearly labelled and easy to play, with no implication that a practice excerpt is a published credit. Narrators’ Roadmap guidance

15Where should narrators look for work?

ACX provides a direct route to audiobook auditions. Other routes include publisher and production-company rosters, Ahab and relationships with authors and rights holders.

A useful routine is to check suitable opportunities, submit carefully prepared auditions and research companies producing books that fit the narrator’s strengths.

Uploading samples alone is not a job-search strategy. Neither is sending the same introduction to every company. Applications should follow the recipient’s submission procedure and link directly to relevant work.

16What should change when auditions produce no bookings?

The first step is an assessment of actual auditions.

Useful questions include whether the opening engages immediately, whether the performance suits the genre, whether dialogue sounds believable and whether the recording has audible technical faults. The projects being targeted also need reviewing.

More auditions will not fix a recurring problem that has never been identified. Equally, a handful of unsuccessful submissions cannot establish that someone is unsuitable for narration.

Rates & agreements

17What does “per finished hour” mean?

PFH is payment for the duration of the finished audiobook, not each hour spent working.

A ten-hour audiobook at £200 PFH pays £2,000. If preparation, recording, production and administration take 60 hours, the gross return is approximately £33.33 per working hour before expenses and tax.

A quote must also identify what it includes. Narration-only and fully produced audio are different deliverables and should not be priced as though they involve the same work.

18What is a useful benchmark for rates?

For UK work, Equity currently recommends a minimum of £150 PFH for remote recording and delivery of mistake-free audio, with £65 PFH for additional editing. The recording rate excludes work such as breath editing and mastering. These are recommended minimums, rather than a universal contract or a ceiling on fees. Equity’s rate guidance

That gives a much more useful reference than “charge whatever feels fair”.

An ACX quote covering the complete production must account for all of that production work, including any outsourced services.

19Is Royalty Share a good way to start?

Royalty Share is speculative work. It should be accepted because the book presents a worthwhile opportunity, rather than because a beginner assumes unpaid production is a compulsory entry fee.

The narrator should examine the author’s existing readership, sales history, marketing activity and track record. A long book with little evidence of demand can consume weeks without producing meaningful income.

PFH work is the clearer choice when dependable payment is the priority. Royalty Share Plus can cover some production costs while retaining potential royalties. Current arrangements should be checked against ACX’s payment terms.

20How should a possible scam be checked?

Verify that the person offering the job is connected to the book and controls the audiobook rights. Where necessary, contact the author or publisher through an independently located official channel.

Warning signs include vague project details, pressure to act immediately, implausible payment arrangements and requests to forward money to someone else.

An unsolicited offer is not automatically a scam. A listing on ACX is not automatically legitimate. Neither should replace checking the actual person and project.

21What must be settled before accepting a book?

At minimum:

  • The manuscript, approximate word count and any demanding or objectionable content.
  • Required accents, languages and other performance expectations.
  • Whether the job is narration-only or full production.
  • The payment arrangement and payment dates.
  • The checkpoint, final deadline and correction process.
  • Responsibility for manuscript changes and additional work.
  • Credits, permitted uses, AI restrictions and cancellation terms.

A vague brief becomes a practical problem when the parties discover that they expected different things. The contract is the place to resolve those differences.

Preparing a book

22How should a manuscript be prepared?

Use the final manuscript and establish the information needed to record it confidently.

For fiction, that usually includes a character list, speaker identification, pronunciation notes and details affecting performance. Colour-coded dialogue can make a large cast much easier to follow. An editable DOCX is particularly useful for automated formatting and preparation tools.

Automated markup must be checked. A line attributed to the wrong character can survive an otherwise accurate recording and become a pickup later.

The purpose of preparation is to prevent mistakes and support performance. Elaborate annotation that does neither is wasted effort.

23How can character voices remain consistent?

Record references for recurring characters and make them quick to access.

A useful reference gives the character’s name followed by a representative line. Placing a marker immediately before the example allows the narrator to jump straight to the voice during recording.

Update the reference if the approved performance changes. Keep references for subsequent books in a series, including apparently minor characters who may return.

Written descriptions help, but “slightly rough, northern accent” is much less precise than hearing the original performance.

24What should the 15-minute checkpoint demonstrate?

The intended final sound and the important performance choices.

The checkpoint should be edited and processed, with representative narration, dialogue and any significant accents or character voices. A separate character-reference recording can cover choices absent from the excerpt.

ACX’s current guidance calls for a representative excerpt rather than an inflexible, exactly 15-minute recording. Approval should be obtained before continuing with the full production. ACX checkpoint guidance

25How should a production deadline be calculated?

Estimate finished duration, multiply it by a realistic production ratio and add time for approvals and corrections.

ACX estimates approximately five to seven working hours per finished hour for new narrators handling production. At six hours of work per finished hour, an eight-hour audiobook represents about 48 working hours—not eight. ACX production estimates

Personal records should replace that initial estimate as experience develops. A deadline should fit the narrator’s available working time, including existing commitments.

26Should a narrator learn punch and roll?

Yes. It is the recommended default for self-recorded audiobook work.

Punch and roll replaces a mistake during recording, reducing the amount of unwanted material that must be sorted through later. Pre-roll plays the preceding audio so the narrator can speak along with it and resume in the same tone and rhythm.

It should be learned early, before an inefficient recording method becomes habitual. A studio-directed session may use a different method because an engineer is managing the recording and edit.

Recording & editing

27What recording level should be used?

For a conventional 24-bit recording workflow, peaks around −12 dBFS during a representative passage provide a useful starting point. Louder passages still need enough headroom to avoid clipping.

The microphone gain should be set for a clean source recording. ACX delivery loudness is addressed during mastering.

Record and preserve a lossless source, normally WAV. An MP3 delivery requirement is not a reason to record the original performance as MP3.

28What is room tone for?

Room tone maintains a consistent background through pauses and edits.

Record a clean section using the same microphone, gain and room conditions as the narration. Remove movement and intermittent noises before using it. A carefully prepared loop can supply room tone where an edit leaves a gap.

Room tone is not a cure for noise underneath speech. Nor should noise be added merely to reach a numerical target: a maximum permitted noise floor is not an instruction to make a quiet recording noisier.

29Should breaths be removed?

Distracting breaths should be reduced or removed. The pause they occupy should be preserved unless the pacing also needs changing.

A breath edit should be checked within the sentence, not merely in isolation. Cutting every breath and closing every gap can produce relentless, unnatural speech.

Gates can automate part of the work, but quiet consonants and word endings need particular attention. A gate that removes part of a word requires correction, however tidy the waveform looks.

30Can automated proofing replace a complete listen?

No. Automated proofing helps find likely errors; it does not certify the finished audiobook.

The useful workflow is to review flagged discrepancies, decide which are real, and listen through the complete recording against the manuscript. That full pass also needs to catch wrong voices, accent slips, poor joins, background noises and awkward pauses.

Tools such as Pozotron can make comparison and pickup preparation more efficient. Their flags still need judgement, and unflagged passages still need listening.

31Should proofing use raw recordings or rendered audio?

For a narrator producing the finished audiobook, proof the rendered audio.

The listener receives the processed result. Gates, noise reduction, de-clicking and other effects can introduce or expose faults that are absent from the raw recording.

The original project remains the place to make corrections. A rendered waveform can also provide a useful visual check, but watching the waveform does not replace listening.

32How should pickups be recorded?

Match the original microphone position, recording level and environment, then listen to the surrounding passage.

Use pre-roll to recover the original pace and energy. Recording a technically clean replacement is only half the task; the performance must also fit.

After inserting the pickup, listen from before the edit until after it. If the join remains obvious, re-record a wider passage. Several minutes spent forcing one word into place may be less efficient than replacing the sentence.

Mastering & delivery

33What are the main ACX mastering targets?

A practical working target is around −20 dB RMS, within ACX’s permitted range. That does not remove the need to check each exported file.

  • Average level: between −23 and −18 dB RMS.
  • Peaks: no higher than −3 dB.
  • Noise floor: no higher than −60 dB RMS.
  • File format: MP3, 44.1 kHz, constant bit rate of at least 192 kbps.
  • Channels: consistent mono or stereo throughout.
  • File length: no more than 120 minutes.

For ordinary single-voice narration, mono is a sensible default. RMS and LUFS are different measurements; a LUFS target should not be substituted for the stated RMS requirement.

ACX currently recommends one to five seconds of room tone at each end, with spacing not exceeding five seconds. The complete specification also covers credits, chapter files and the retail sample.

34Does passing an ACX checker mean the audiobook is ready?

No. A checker can confirm particular measurements. It cannot confirm correct words, consistent characterisation, natural editing or an engaging performance.

Before delivery, check the actual exported files. Confirm that every chapter is present, the boundaries are correct, the intended tracks were included and the files use the required format.

A project playing correctly inside REAPER is not proof that its export is correct.

35Should a narrator add music, sound effects or extra performers?

No—not as part of a standard narration job or ordinary ACX narration production.

The narrator-producer’s responsibility for editing and mastering does not make those additions necessary. An audition should not be decorated with music, and a completed narration should not acquire effects or extra voices without a specific commission for that work.

Deciding whether a book should become a sound-designed or full-cast production is outside the scope of this FAQ.

36How should files and backups be organised?

Keep source audio, project files, proofing exports, pickups and final delivery files clearly separated.

Use consistent chapter numbering, such as 001_Opening_Credits and 002_Chapter_01, unless the client supplies a different convention. Final delivery should come from one clearly identified folder.

Back up both the project and its source audio. A project file contains editing instructions; it cannot recreate a missing recording. At least one backup should survive failure of the working drive.

Payment, protection & support

37When should an ACX producer confirm payment?

Only after the agreed payment has actually been received.

For Pay-for-Production and the fee portion of Royalty Share Plus, payment is arranged with the rights holder. The ACX payment confirmation is not a formality to complete in anticipation of payment. ACX payment guidance

The amount, method and timing should be agreed before production. Any deposit arrangement should also be recorded in the agreement.

38What should contracts say about AI?

They should explicitly address voice cloning, model training and the generation of new performances.

A general right to use the audiobook recording should not quietly become unrestricted permission to reproduce the narrator’s voice. NAVA’s free Synthetic Voice and AI Rider provides a useful starting point for those protections.

Uploading audio or manuscripts to production services is a separate issue. The narrator needs permission to use the material that way and should understand the service’s retention and training terms.

39Where can narrators find further help?

The Voice Acting Club is a treasure trove of resources and practical discussion across voice acting, recording and professional practice.

Narrators’ Roadmap provides extensive audiobook-specific guidance. David Winter’s course provides the practical REAPER workflow recommended earlier in this FAQ.

Questions posted to a community should include enough information to answer them. For an audio problem, a short unprocessed sample and a description of the setup are much more useful than “the microphone sounds bad”.