Midnight Radio
Velvet · narrator
● Modern narrator
Close and low; suits mystery, urban and supernatural webtoons.
NarratorsMale9 languages
Audio comics · Motion comics
Audio Comic Dubbing: Voice Every Panel and Character
Audio comic dubbing gets hard when one chapter has a dozen panels, five or six characters, and narration and thought bubbles mixed in. With AI Drama Voice you sort the lines by panel first, then paste a clean voice script: the AI separates narrator and characters, applies the emotion you mark in brackets to each line, and exports an MP3 plus a line-by-line SRT you can lay back under every panel.
- Lines sorted by panel
- Narration / dialogue / thoughts
- MP3 + per-line SRT
- Free cast analysis
Updated Nov 5, 2026Fact-checked Nov 5, 2026
Urban fantasy · The last ten seconds on the phone
Chinese · Script excerpt · 4 characters
NarratorCalm
江迟拿勺子的手，停在了半空。
Jiang Chi's hand, holding the spoon, froze in mid-air.
Old ZhouCalm
这位是？
And who's this?
Song ZhiyaoCalm
第九处，宋知遥。来查一桩案子。
Ninth Bureau, Song Zhiyao. I'm here about a case.
Old ZhouTender
第九处？咱们这片，能出什么大案子。小迟，你没惹事吧？
The Ninth Bureau? What big case could happen round here? Xiao Chi, you're not in trouble, are you?
0:00 / 0:28
Try it now
Try a scene: paste one chapter's lines
The studio below is loaded with the comic sample. Swap in your own voice script (one line per row, speaker named) and watch the AI split narration from characters and tag emotions. Cast analysis costs no credits.
Unlimited free analysis · Dubbing is charged per character, failed runs refundedOpen in the studio
Listening examples
Two common audio comic sounds
These two clips are listening examples of multi-voice comic dubbing (placeholder audio to show how narration and characters are layered, not output made on this site). The first is modern urban fantasy with a close, low narrator and three characters; the second is a classic tale where a storyteller links the lines with “Pigsy said:”.
Script excerpt · Urban fantasy · The last ten seconds on the phone
NarratorCalm
江迟拿勺子的手，停在了半空。
Jiang Chi's hand, holding the spoon, froze in mid-air.
Old ZhouCalm
这位是？
And who's this?
Song ZhiyaoCalm
第九处，宋知遥。来查一桩案子。
Ninth Bureau, Song Zhiyao. I'm here about a case.
Old ZhouTender
第九处？咱们这片，能出什么大案子。小迟，你没惹事吧？
The Ninth Bureau? What big case could happen round here? Xiao Chi, you're not in trouble, are you?
Jiang ChiCalm
周叔，你钥匙上那个铃铛，以前没见过。
Uncle Zhou, that little bell on your keys. I've never seen it before.
Old ZhouTender
上个月我闺女送的，说是保平安。怎么，吵着你了？
My daughter gave it to me last month, for luck. Why, is it bothering you?
Jiang ChiCalm
没有。周叔，手机借我用一下，我的没电了，金姨刚才找我有点事。
No. Uncle Zhou, can I borrow your phone? Mine's dead, and Auntie Jin was looking for me.
Script excerpt · Classic tale · The White Bone Demon
TripitakaCalm
你有甚么不相应处？
What is it that does not sit right with you?
Storyteller
八戒道：
Pigsy said:
PigsySarcastic
师父，他要和你分行李哩。跟着你做了这几年和尚，不成空着手回去？你把那包袱内的甚么旧褊衫，破帽子、分两件与他罢。
Master, he wants his share of the luggage! Years as your monk, and he goes home empty-handed? Give him a couple of old robes and a battered hat from the bundle.
Storyteller
行者闻言，气得暴跳道：
Hearing this, Monkey leapt up in a fury:
Sun WukongFurious
我把你这个尖嘴的夯货！老孙一向秉教沙门，更无一毫嫉妒之意，贪恋之心，怎么要分甚么行李？
You pointy-snouted oaf! I have served the faith without a shred of envy or greed. Why would I want any luggage?
Storyteller
唐僧道：
Tripitaka said:
TripitakaCalm
你既不嫉妒贪恋，如何不去？
If you feel no envy or greed, then why not go?
Get the terms straight
Audio comic vs motion comic vs AI comic drama
The three terms get used loosely, but each asks something different of the voice track. Work out which one you are making before you organise the lines.
| Format | Visuals | How lines are organised | What the voice track must do |
|---|---|---|---|
| Audio comic | The comic's original still panels, at most simple pans and zooms | By chapter and panel; a panel may hold several lines or none | Narrator links, clearly different characters, thoughts layered apart from speech |
| Motion comic | Comic art with layered animation: blinks, hair movement, camera moves | Still panel by panel, but each panel stays on screen longer | Line timing has to match the on-screen motion |
| AI comic drama | A continuous episode built from storyboards, with cuts and transitions | By shot and scene, close to a short-drama script | Acting and pacing like a short drama, with dense twists |
This page covers the first two: you already have the art (your own, licensed, or an AI-drawn webtoon) and want to add voices so the comic can be listened to. The big difference from live-action drama is that viewers read a panel while they listen, so knowing which line belongs to which panel matters more than performance. A voice that lands half a panel late feels wrong straight away.
The usual audio comic production steps
A chapter usually goes: get the art, sort the lines by panel, voice them, edit (place audio and subtitles under each panel, add transitions, music and sound effects), publish. Audio comic dubbing software handles only the voicing and the subtitle timing; the art, music and effects are yours to add in the editor. The full pipeline is in the AI comic drama workflow.
Five steps
How to dub an audio comic in five steps
One person can do all of it. The trick is to keep the panel sheet (for you) apart from the voice script (for the studio): the first tracks panels, the second only says who speaks.
- 01
Step 1
Build a panel sheet
List every panel of the chapter in order: panel number, what is drawn, which lines it holds, whether each is narration, dialogue or a thought, and any sound words. Keep silent panels on the list too; you will need them when lining up the art.
- 02
Step 2
Write the voice script: one line per row, no panel numbers
Copy the spoken lines from the sheet in order as “Name: line”, one per row. Leave panel numbers, art notes and sound cues out; any row without a colon is read aloud by the narrator.
Tip · Give each character one name for the whole chapter, e.g. always “Mia”, never “the clerk” halfway through.
- 03
Step 3
Paste it and get the cast list
Paste the script into the dubbing studio and run the cast analysis. The AI lists the narrator and every character, suggests a voice by gender, age and persona, and applies the emotion you mark in brackets to each line. This step is free, so adjust as often as you like.
- 04
Step 4
Tune thoughts, emotions and voices
Write thoughts as “Mia (thought): …” so they keep Mia's voice. A line marked “(softly)” is read as a whisper tag. Pick clearly different voices for leads and side characters; you can audition them first in the voice library.
- 05
Step 5
Generate, export, place lines under panels
Generate the chapter and download the MP3 and SRT. The SRT has one numbered cue per line, so with your panel sheet beside you, drop each cue under its panel art in CapCut and the chapter is laid out.
Template
A 12-panel chapter: panel sheet and voice script
Here is a 12-panel example (an original short, Rain at the Corner Store, chapter one). The table is the panel sheet you keep for yourself; only the last column goes into the audio comic dubbing tool.
| Panel | Art | Type | Row in the voice script |
|---|---|---|---|
| 1 | Rainy night, a corner store with its lights on | Narration | Narrator: Two in the morning, and the corner store was the only light left on the street. |
| 2 | Behind the counter, clerk Mia yawns | Thought | Mia (thought): Four more hours till my shift ends… |
| 3 | The door chimes; a soaked boy walks in | Sound: ding-dong | (not in the script; add a door-chime effect in the edit) |
| 4 | The boy stands by the shelves | Dialogue | Boy: Excuse me, do you sell umbrellas? |
| 5 | Mia points at the door | Dialogue | Mia: In the bucket by the door. That's the last one. |
| 6 | The boy takes the umbrella and doesn't move | Narration | Narrator: The boy picked up the umbrella, but stayed exactly where he was. |
| 7 | Close-up: the floor under his feet, bone dry | Thought | Mia (thought): Wait… he's soaked through. Why isn't there a single drop on the floor? |
| 8 | The boy looks up, half smiling | Dialogue | Boy (softly): You've noticed, haven't you? |
| 9 | Mia steps back into a shelf | Sound: clang | Narrator: With a clang, the shelf behind her rocked. |
| 10 | The manager leans out of the stockroom | Dialogue | Manager: Mia? Who are you talking to at this hour? |
| 11 | Mia turns; the doorway is empty | Dialogue | Mia (surprised): There was someone right… |
| 12 | The umbrella is back in the bucket | Narration | Narrator: In the bucket, the last umbrella stood where it always had, still dripping. |
Only the last column is pasted: 11 rows, from “Narrator: Two in the morning…” to “Narrator: In the bucket…”. After generation, cue 1 belongs to panel 1, cue 2 to panel 2, and cue 3 to panel 4, because panel 3 is only a door chime. Work down the sheet panel by panel and nothing slips.
Sound words: three ways to handle them
- Leave them to sound effects: a door chime, thunder or an explosion is a sound, not a line. Keep it out of the script and add an effect in the edit (panel 3).
- Let the narrator carry them: in storyteller-style comics, fold the sound into a narration line, such as “With a clang, the shelf behind her rocked” (panel 9).
- Let the character make them: sounds a character makes (“Hah!”, “Tch”) go straight into that character's line and are performed in their voice.
The two most common mistakes
One: putting “Panel 3” on its own row, which the narrator will then read out. Two: writing “SFX: ding-dong”, which creates a character called “SFX” and gives it a voice. Panel numbers and sound cues stay in the panel sheet.
Should inner thoughts be voiced?
Yes. In audio comics the thought bubbles are often the key lines that move the plot (panel 7). Written as “Mia (thought): …”, the line keeps Mia's voice, and the bracketed note goes to the voice model as a delivery cue so it sits apart from spoken dialogue. If a series leans heavily on thoughts, a touch of reverb on those lines in the edit makes them feel more inside the head.
Cast
Voices audio comics use most
An audio comic usually needs one narrator for the whole run, plus leads, side characters and a villain. In the audio comic dubbing studio each of them keeps one fixed voice. Click to audition.
Storyteller
Classic · folk tale
● Storyteller narrator
Traditional storytelling cadence for period pieces and “so-and-so said:” links.
NarratorsMale9 languagesYoung Hero
Clear · energetic
● Young hero
Bright and clear; the usual lead in action and school comics.
YouthMale9 languages
Sweetheart
Bright · playful
● Cheerful heroine / best friend
Bright and girlish, for slice-of-life and romance.
Female leadsFemale9 languagesElegant Lady
Mature · poised
● Cool investigator / senior
Mature and composed; stands out clearly next to a girlish voice in the same panel.
Female leadsFemale9 languages
Ice Queen
Cold · villainous
● Villainess
Cold with an edge; built for sneers and sarcasm.
VillainsFemale9 languages
The Father
Warm · grounded
● Manager / father / elder
Warm and weighty; gives a side character a memorable voice.
NarratorsMale9 languagesCartoon Villain
Over-the-top · comic
● Comic relief
Exaggerated and funny, Pigsy-style banter.
Anime & toonMale9 languagesMonkey King
Lively · anime
● Mischievous hero
Quick and wild; made for Monkey King types.
Anime & toonMale9 languagesPreview coming soon
Why this tool
Why use AI Drama Voice as your audio comic dubbing tool
Narration, dialogue and thoughts stay apart
Rows starting “Narrator:” and rows with no speaker go to the narrator; “Name:” rows go to that character; thoughts keep the character's voice with a different delivery.
One fixed voice per character
Within a chapter, one name always gets one voice. Note the voice you chose for each character and pick the same one every chapter so the series never recasts.
Emotion on every line
Sneer, choked, furious, whisper and more apply line by line, and notes like “(softly)” or “(surprised)” are matched to the right tag.
One SRT cue per line
Cues follow the voice script in order, timed to where each line sits in the MP3, so in CapCut you just drop them under each panel in sequence.
Foreign-language versions too
Once the Chinese version is done, the same cast can voice English, Japanese, Thai and five more languages, handy for taking a webtoon audio version abroad.
Who it's for
Who audio comic dubbing software is for
Audio comic channels
You hold the rights to a webtoon and publish an audio version daily or weekly. One voice script per chapter, one run per chapter, and one person keeps the schedule.
Motion comic studios
Voice first, animate second: time each panel's motion and camera moves to the real line lengths, instead of animating and then hunting for voice actors.
Artists voicing their own work
Make an audio preview of your own comic to promote the series on short-video platforms, with every voice chosen by the person who created the characters.
Novel-to-webtoon teams
Webtoons adapted from novels carry a lot of narration. Splitting narration from dialogue makes the chapter sound like a small drama, not one person reading start to finish.
Tips
Six tips for audio comic dubbing software
Never squeeze one panel into one row
If two people talk in a panel, write two rows. One speaker per row, or both lines land on the first name.
Keep narration lean
If the art already shows the action, the narrator does not need to repeat it. Narration carries what the panel can't: time, inner state, suspense.
Write a one-line profile per character
At the top of the panel sheet: “Mia: night-shift clerk, early twenties, all bark and no nerve”. Choose voices from it, and you won't forget them ten chapters on.
Keep a cast sheet
Record “character → voice” for each series. Every chapter is a separate run, so pick the voices again from the sheet each time; see consistent character voices.
Test one chapter first
Run your most typical chapter end to end, settle the narrator style and the lead voices, then process the rest in the same format.
Line panels up by the SRT
Import the MP3 and SRT into CapCut together: each subtitle block marks where a line starts and ends, so you drag panels to the blocks instead of hunting by ear. Steps in syncing MP3 and SRT in CapCut.
FAQ
Frequently asked questions
Four broad kinds: general text-to-speech tools (fine for a single narrator), the read-aloud built into video editors (convenient, but casting is manual), script-dubbing tools that cast characters automatically (such as AI Drama Voice, which voices narrator and every character in one run and exports an SRT), and hiring human voice actors. The number of characters and how often you publish decide which fits.
An audio comic uses the comic's own still panels and organises lines by chapter and panel. An AI comic drama is a continuous episode built from storyboards, with cuts and transitions, closer to a short drama. The first is about panel timing and narrator links, the second about acting and pace; for the latter see AI comic dubbing.
With audio comic dubbing software, one person: build a panel sheet, write a one-row-per-line voice script, paste it and confirm the cast, then place each SRT cue under its panel. Human dubbing usually means one voice actor per role plus someone on post-production.
Voice the thoughts, in the character's own voice, written “Name (thought): line”. Sound words depend: door chimes and thunder are better as sound effects in the edit; storyteller-style comics can fold them into narration; sounds a character makes go into their own line.
Usually through platform creator programmes, sponsored work and paid content, under whatever rules each platform has at the time; we make no promises about income. The precondition is rights to the art and story: your own original work, or a work you are licensed to adapt and distribute. For commercial use of AI voices see can AI voices be used commercially.
Credits are deducted by character count, narration and dialogue alike, with translation counted for foreign versions. Cast analysis is free and failed lines are refunded automatically. Plans are on the pricing page.
Your next audio comic chapter, voiced solo Every panel, every voice.
The panel sheet tracks panels, the voice script says who speaks. Paste it, check the cast, then generate the chapter MP3 and per-line subtitles.
Billed by characters generated · Failed runs refunded automatically · Alipay, WeChat Pay and cards