Most indie authors proofread by eye, and most of them miss the same things every time. The sentence that reads fine on screen can sound like a tangle when spoken aloud. Proofreading by ear, meaning you have the computer read your manuscript to you, catches the clunky lines that your eyes have already forgiven because they know what you meant to write.
Pendrilo Studio includes Read Aloud, which speaks your chapters back to you. You can use it for a single scene or a whole draft. This article covers how to set up a listening pass, how to fix character names that get mangled, and how to choose between the two voice types Read Aloud offers.
Why your eye skips what your ear catches
When you reread your own prose, your brain fills in gaps. A doubled "the" disappears. A sentence with three clauses that technically parses still feels fine because you already know the intended meaning. Your eye travels faster than your attention does.
Your ear is slower and less forgiving. It has to take each word in sequence, and it notices when a phrase runs out of breath, when two sentences start the same way, or when a comma sits where a speaker would never pause. Those are the problems readers feel but rarely name.
Listening also exposes repetition. You might not notice that you used "suddenly" four times in one chapter until you hear it land four times in ninety seconds. If you use Echo Report in Studio alongside a listening pass, the two together catch more than either alone, but the ear is still the quicker net for rhythm.
Set up a listening pass that you will actually finish
A full novel read aloud can take several hours, so break the job into pieces. Work chapter by chapter, and keep each session to something you can finish in one sitting, usually one or two chapters.
Before you start, decide what you are listening for. Listening for everything at once tends to produce vague notes like "this feels off." Pick two or three targets for the first pass:
- Sentences that run past the point where you need a breath
- Repeated words or sentence openings within a paragraph
- Dialogue tags that interrupt the line instead of carrying it
Then do a second pass for names, timeline words, and anything a reader would trip on. Splitting the work this way keeps your attention sharp. A third pass for typos, if you want one, is far easier once the rhythm is already right.
Keep a notes file open while you listen. Write down the paragraph or line number rather than stopping to fix each problem mid-stream. Stopping every time you hear something breaks the flow and makes you hear less. When the pass is done, go back and fix the marked sections together.
Mark sentences by how they sound
A sentence that feels clunky usually has one of a few shapes. It stacks several "and" clauses in a row. It buries the verb at the end. It uses a subject and verb that are pulled apart by a long phrase. When you hear one of these, mark it and try a short rewrite. Often the fix is splitting one long sentence into two, or moving the strongest word to the end.
Be careful not to sand every sentence into the same length. Varied rhythm is part of what makes prose readable. The goal is to remove sentences that sound stuck, not to make everything sound like a bulletin.
Fix character names with custom pronunciations
Invented names are where text-to-speech goes wrong fastest. A name like Ysolde, Thessaly, or Bramwick may be read with the wrong stress, split in an odd place, or swapped for a common word that sounds similar. If you hear the narrator stumble on a name, your readers may stumble too, and the stumble will sit in their head every time that character appears.
Read Aloud supports custom pronunciations, so you can tell it how a particular word should sound. Set one for every invented name, place, and unusual term in the book. Do this early, before you start a listening pass, so you are hearing your prose and not the mispronunciations.
Keep a short list as you write. A character codex is a good place to record the spelling of each name and a rough spoken form. When you listen, check each name against your list. If a name still sounds wrong after you set the pronunciation, try a respelling that the voice handles more predictably.
One caution: custom pronunciations fix sound, not spelling. If you have two characters whose names look alike, the voice may still read them differently in ways that help or hurt. Listen for that specifically in dialogue-heavy scenes.
Built-in voices or natural voices: the trade-off
Read Aloud gives you two kinds of voice. Built-in voices come from your computer and are available immediately. Natural voices are 13 voices that you download once and then run locally on your machine. The download is several hundred megabytes, and the exact size is shown when you start it.
The choice comes down to a trade-off between convenience and how much the voice helps you hear your prose.
Built-in voices
The advantage is speed. Nothing to download, nothing to wait for, and you can start a listening pass right away. They are a good choice when you want a quick check of a single chapter or when disk space is tight.
The drawback is that they can flatten the text. Pauses come in the wrong places, emphasis lands oddly, and the voice may hide the very rhythm problems you are trying to catch. A built-in voice can make a clumsy sentence sound merely robotic, which makes it harder to tell whether the sentence is the problem or the voice is.
Natural voices
Natural voices handle phrasing, pauses, and emphasis more like a human reader. That makes them better for judging rhythm. When a sentence sounds awkward in a natural voice, the awkwardness is usually in the writing, not in the synthesis. For a full manuscript pass, this is the one most authors will want.
The cost is the one-time download and a little more processing. If you work on a laptop with limited space, plan for the download before you need it.
Which to choose
Use a built-in voice for quick checks and for the first look at a new chapter. Use a natural voice for the full listening pass where you are judging prose quality. Many authors find it useful to switch: built-in for a fast skim, natural for the careful pass. Either way, the same custom pronunciations apply, so you set them once.
What to listen for beyond rhythm
Once the basic rhythm is working, listen for the smaller things that a reader's mind catches. Dialogue tags are a common one. "She said, slowly, with a sigh, while glancing away" will sound like a jumble, and you will hear the fix at once: cut the extras and let the line carry the emotion.
Listen for timeline words too. Hearing "that morning" twice in a paragraph, or "later" and "after" landing one after another, makes it obvious when a scene is pacing unevenly or when a time reference has gone wrong. Pair this with your own continuity notes if you keep them.
Finally, listen for the places where you hesitate. If you stumble over a sentence while reading it aloud yourself, the voice will stumble too, and a reader will likely slow down there. Those hesitations are often the most useful marks in your notes.
Keep the ear pass in your revision routine
A listening pass works best as one step in a longer revision, not as a final polish. Do your structural and content edits first, then run the ear pass on the revised chapters, then fix what you marked. If you are sending the manuscript to beta readers, a clean listening pass before they see it means their feedback can focus on story rather than on sentences that a voice would have caught.
If you want to try Read Aloud, Pendrilo Studio is free during the beta with no key required. You can start at the registration page and load a chapter you already know well. Pick a passage you have read a dozen times, listen to it once, and see what your ear finds that your eye missed.
