Paper deep dive

Dramarrator: Object-Based Audio Editing for Audio Drama Production from Books

Karim Benharrak, Oriol Nieto, Bryan Wang, Zeyu Jin, Amy Pavel

Year: 2026Venue: arXiv preprintArea: cs.HCType: PreprintEmbeddings: 113

Intelligence

Status: succeeded | Model: Gemma-4-26B-A4B | Prompt: intel-v1 | Confidence: 90%

Last extracted: 8/12/2026, 1:26:03 AM

Summary

The paper introduces Dramarrator, an object-based audio editing tool that automates the adaptation of books into audio dramas. It represents narrative elements like characters and scenes as editable objects with linked audio assets (speech, SFX, music). Edits to these objects automatically propagate across all dependent assets, significantly reducing the manual labor and task load for creators compared to traditional linear timeline editing. User studies indicate that Dramarrator lowers entry barriers for novices and produces output quality comparable to professional tools.

Entities (8)

Dramarrator · software-system · 98%Object-Based Audio Editing · methodology · 95%Character · narrative-construct · 92%Scene · narrative-construct · 92%Listener Study · evaluation-method · 90%User Study · evaluation-method · 90%Adobe Audition · software-tool · 85%Pro Tools · software-tool · 85%

Relation Signals (10)

Dramarrator → implements → Object-Based Audio Editing

confidence 95% · Dramarrator, an audio drama authoring tool built around object-based audio editing

Dramarrator → models → Character

confidence 92% · Dramarrator models narrative constructs as editable objects: a Character object such as “Tom”

Dramarrator → models → Scene

confidence 92% · a Scene object such as “Market Square” has attributes like ambience design

Character → hasattribute → Voice Design

confidence 90% · a Character object such as “Tom” has attributes like voice design

Scene → hasattribute → Ambience Design

confidence 90% · a Scene object such as “Market Square” has attributes like ambience design

Dramarrator → reduces → Task Load

confidence 90% · Dramarrator significantly lowered task load when creating audio dramas

Scene → linksto → Ambience

confidence 88% · All sound effects, ambience layers ... that occur while a scene is active are linked to that scene object

Character → linksto → Speech

confidence 88% · A character’s dialogue lines and sound effects ... are linked to that character object

Dramarrator → comparesfavorablyto → Adobe Audition

confidence 80% · Dramarrator significantly lowered task load ... compared to existing tools ... Adobe Audition

Dramarrator → comparesfavorablyto → Pro Tools

confidence 80% · Dramarrator significantly lowered task load ... compared to existing tools ... Pro Tools

Cypher Suggestions (0)

No Cypher suggestions yet.

Abstract

Abstract:Audio dramas weave dialogue, sound effects, and music into immersive stories. Creators often adapt books into audio dramas, but this process remains labor-intensive, requiring them to interpret source material, author scripts, generate audio assets, and assemble them on a timeline. Because story elements like characters and scenes manifest across many interdependent assets, a single change can ripple into manual updates across the entire project. We present Dramarrator, an audio drama authoring tool built around object-based audio editing, where these story elements are represented as editable objects. Dramarrator extracts these objects from a book, generates linked audio assets (speech, sound effects, and music), and composes a multi-track audio drama. Edits to any object (e.g., a character's voice) automatically propagate to all dependent assets. In a user study with professionals (N=8), Dramarrator significantly lowered task load when creating audio dramas. A listener study (N=300) shows that creator-refined output from Dramarrator approaches the quality of productions made with existing professional tools, and an exploratory study (N=3) suggests object-based editing lowers entry barriers and generalizes beyond audio dramas.

PDF

Open source PDF →Open local PDF →

Trouble viewing inline? Open PDF directly →

Full Text

111,780 characters extracted from source content.

Expand or collapse full text

by-nc-nd Dramarrator: Object-Based Audio Editing for Audio Drama Production from Books Karim Benharrak 0009-0002-3279-5664 karimbenharrak@berkeley.edu University of California, BerkeleyBerkeleyCAUSA , Oriol Nieto 0000-0001-6459-7609 onieto@adobe.com Adobe ResearchSan FranciscoCAUSA , Bryan Wang 0000-0001-9016-038X bryanw@adobe.com Adobe ResearchSeattleWAUSA , Zeyu Jin 0000-0003-0161-5915 zejin@adobe.com Adobe ResearchSan FranciscoCAUSA and Amy Pavel 0000-0002-3908-4366 amypavel@eecs.berkeley.edu University of California, BerkeleyBerkeleyCAUSA (2026) Abstract. Audio dramas weave dialogue, sound effects, and music into immersive stories. Creators often adapt books into audio dramas, but this process remains labor-intensive, requiring them to interpret source material, author scripts, generate audio assets, and assemble them on a timeline. Because story elements like characters and scenes manifest across many interdependent assets, a single change can ripple into manual updates across the entire project. We present Dramarrator, an audio drama authoring tool built around object-based audio editing, where these story elements are represented as editable objects. Dramarrator extracts these objects from a book, generates linked audio assets (speech, sound effects, and music), and composes a multi-track audio drama. Edits to any object (e.g., a character’s voice) automatically propagate to all dependent assets. In a user study with professionals (N=8), Dramarrator significantly lowered task load when creating audio dramas. A listener study (N=300) shows that creator-refined output from Dramarrator approaches the quality of productions made with existing professional tools, and an exploratory study (N=3) suggests object-based editing lowers entry barriers and generalizes beyond audio dramas. Audio Editing, Creativity Support Tools, Human-AI Co-Creation †journalyear: 2026†copyright: c†conference: The 39th Annual ACM Symposium on User Interface Software and Technology; November 02–05, 2026; Detroit, MI, USA†booktitle: The 39th Annual ACM Symposium on User Interface Software and Technology (UIST ’26), November 02–05, 2026, Detroit, MI, USA†doi: 10.1145/3830398.3830546†isbn: 979-8-4007-2856-3/2026/11†ccs: Human-centered computing Interactive systems and tools Figure 1. Dramarrator’s interface for object-based audio drama authoring. The Script Editor (left) represents a multi-track audio timeline via a transcript with dialogue, sound effects, and music, nested within scenes. The Object Pane (right) visualizes all Character and Scene objects: a Character Vignette exposes a character’s voice design and their linked speech lines and SFX, and a Scene Vignette exposes a scene’s ambience design and its linked ambience and SFX assets. When creators edit an object (e.g., changing a character’s voice), Dramarrator automatically propagates edits across all affected object-dependent assets. 1. Introduction Audio dramas adapt source material (e.g., books) into immersive experiences told through dialogue, sound effects (SFX), and music (Wayland et al., 2020; Rattigan, 2002; Crook, 2002). Adapting books into audio dramas requires creators to decompose expository prose into dialogue and continually decide how to convey actions, environments, and atmospheres through speech, sound effects, and music (Toscan, 2023; Wayland et al., 2020). For example, a spoken line “I’m opening this crate” may be replaced with a sound effect that conveys the same action. Recent generative AI advances dramatically lowered the cost of producing individual audio assets from text prompts (ElevenLabs, 2026a; Borsos et al., 2023; ElevenLabs, 2026b; Suno, 2026; Adobe, 2026c). However, assembling assets into a coherent narrative and iteratively refining it remains challenging. Current audio editing tools such as Adobe Audition (Adobe, 2026a), Pro Tools (Avid, 2026), and Logic Pro (Apple, 2026) represent audio dramas as numerous isolated audio clips on a linear timeline, with limited explicit connection to the story constructs (e.g., characters, scenes, story arcs) that creators reason about (Wayland et al., 2020; De Fossard, 2005). For example, a character “Tom” may span many speech clips across multiple scenes, yet existing tools do not model “Tom” as a coherent entity. Thus, single high-level decisions, such as changing a scene’s weather or rewriting a character’s dialogue tone, require manually locating, regenerating, and replacing every affected clip across all timeline tracks, which discourages revision of early decisions and raises the barrier to multi-episode production. Prior work has shown that editing through higher-level abstractions simplifies workflows in creativity support tools (Li et al., 2023; Beaudouin-Lafon, 2000) by grouping lower-level elements into editable objects across programming (Rentsch, 1982; Stefik and Bobrow, 1985), design (Xia et al., 2016; Figma, 2026; Xia et al., 2017), and multimedia editing (Kim et al., 2023; Xia, 2020; Han et al., 2022; Yeh and Kim, 2018). These abstractions, however, typically compose objects from elements that share a uniform representation (e.g., code instances, design components, or text passages). In audio dramas, narrative constructs such as characters and scenes comprise heterogeneous asset types (e.g., a character can be expressed through their script, vocal delivery, and accompanying sound effects). A single narrative change, such as altering a character’s personality, must therefore be translated into distinct edits across these asset types. To achieve this, we introduce object-based audio editing and present Dramarrator, an end-to-end pipeline and authoring interface for adapting books into audio dramas. Dramarrator models narrative constructs as editable objects: a Character object such as “Tom” has attributes like voice design, and a Scene object such as “Market Square” has attributes like ambience design. Each generated audio asset on the timeline (e.g., a speech clip, a sound effect) is linked back to the object it belongs to. This structure enables editing at the narrative level: changing Tom’s voice design to be raspier, for example, automatically regenerates all of Tom’s speech clips across the entire project. Dramarrator’s pipeline extracts these objects from the book’s source text, adapts the text into an audio drama script, and orchestrates multiple generative models to produce all audio assets. Creators can then review and refine both objects and individual assets through Dramarrator’s interface (Figure 1). In a user study with 8 professional audio drama producers, Dramarrator significantly lowered the required task load when creating audio dramas compared to existing tools. Professionals attributed time savings to object-based editing: a single object change (e.g., updating a character’s voice) propagates automatically to all dependent assets, replacing up to 30 manual edits at once. Professionals also reported higher creative exploration, since Dramarrator lets them iteratively shape characters and scenes while hearing them in the final audio drama, whereas their existing sequential workflows require all character and scene decisions to be fixed early on in their process. A listener study (N=300N=300) shows that audio dramas refined by professionals in Dramarrator are comparable to those produced with existing tools on engagement, audio and story quality, and character and scene consistency. An exploratory study involving 3 novice creators from adjacent domains (game, RPG, and puzzle design) shows that Dramarrator enables novices without audio-editing expertise to create audio dramas. Their feedback also suggests that object-based editing may generalize to other creative domains organized around narrative constructs such as characters and scenes. In summary, we contribute: • Object-based audio editing, a concept that links narrative objects (e.g., characters, scenes) with editable attributes (e.g., voice, ambience) and object-dependent assets (e.g., speech, sound effects, music) to enable high-level edits that automatically propagate across the entire audio. • Dramarrator, an audio drama authoring system that generates speech, music, and sound effects from a book, automatically assembles them into an audio drama, and allows creators to steer the output through object-based editing. • A user study (N=8) showing that Dramarrator significantly lowers task load when creating audio dramas compared to professionals’ existing tools, complemented by a listener study (N=300) evaluating output quality and an exploratory study (N=3) probing novice perceptions and the generalizability of object-based editing to other storytelling domains. 2. Related Work We build upon prior work on audio editing tools, editing with higher-level abstractions, and automated audio generation. 2.1. Audio Editing Tools Digital audio workstations (DAWs) such as Pro Tools (Avid, 2026), Logic Pro (Apple, 2026), and Adobe Audition (Adobe, 2026a) support multi-track audio editing by organizing speech, sound effects, and music across a linear timeline. While powerful, these tools impose steep learning curves and demand considerable manual effort to navigate and assemble numerous individual audio clips. To mitigate some of these challenges, especially for spoken content, text-based audio editing interfaces (Descript, 2026; Benharrak et al., 2025; Rubin et al., 2013; Shin et al., 2016) enable users to modify content by editing a script that is time-aligned to the corresponding speech. For example, Rubin et al. (Rubin et al., 2013) presented one of the earliest content-based editing tools for audio story editing, and Shin et al. (Shin et al., 2016) bidirectionally linked scripts to audio so that edits in one representation are reflected in the other. This progression from waveform to text-based editing reflects a broader shift toward more semantic, narrative-level representations of audio. Our work features both existing editing paradigms while further advancing the trajectory toward object-based editing, where users manipulate narrative constructs, such as characters and scenes, rather than individual audio clips or words in transcripts. 2.2. Editing via Higher-Level Abstractions In creative projects, a single conceptual change often requires updating many dependent elements. Changing a brand color in a design system or a character’s personality in a story means manually finding and modifying every affected artifact. Prior work facilitates this by grouping related elements into higher-level objects that propagate edits automatically (Li et al., 2023; Beaudouin-Lafon, 2000), as seen in programming (Rentsch, 1982; Stefik and Bobrow, 1985), text (Kim et al., 2023; Han et al., 2022), visual (Xia et al., 2016, 2017; Suh et al., 2025), and 3D tools (Yeh and Kim, 2018). Modifying a class in object-oriented programming updates all its instances; editing a component in Figma (Figma, 2026) propagates changes to every screen that uses it. These abstractions, however, typically group elements that share a uniform representation (e.g., code instances, design components). In audio dramas, a single narrative construct, such as a character’s personality, is realized through heterogeneous assets spanning text (dialogue) and audio (voice style, SFX, music). Editing at this level requires knowing not only which assets to change but how to transform a narrative-level change into appropriate modifications across modalities. To this end, Dramarrator introduces object-based audio editing. Objects represent narrative constructs (e.g., characters, scenes) with editable attributes (e.g., voice or ambience design) and object-dependent assets (e.g., speech, sound effects, music). Dramarrator also gives users freedom to navigate the ladder of abstraction (Li et al., 2023): high-level object changes automatically propagate to all dependent assets, while low-level edits to individual assets can be made via a transcript or timeline editor. 2.3. Automated Audio Generation Recent advances in generative AI have made audio production faster and more accessible by using text prompts to generate speech (ElevenLabs, 2026a; Laban et al., 2022; Google, 2026b), sound effects (Borsos et al., 2023; Kreuk et al., 2023), or music (Copet et al., 2023; Agostinelli et al., 2023). However, these models typically generate individual audio clips rather than structured, multi-track compositions. Recent work orchestrates multiple models to combine speech, sound effects, and music from text instructions (Liu et al., 2025; Guo et al., 2025; Xu et al., 2025), but these systems are limited to short outputs of approximately 2–3 minutes and produce a single flattened audio file, preventing creators from editing or rearranging individual elements after generation. Most closely related to our work, SoundStager (Yoo et al., 2026) generates editable layered soundscapes but requires video input where dialogue and visuals already establish narrative context. In audio dramas, by contrast, the auditory channel is the only medium: speech, sound effects, and music must jointly convey the narrative, and characters and scenes must remain consistent across productions spanning tens of minutes to hours. To address these requirements, Dramarrator employs object-based editing to orchestrate multiple generative models into long-form, multi-track audio dramas, while exposing the editable structure for creators to fine-tune results at both the narrative and asset levels. 3. Design Formulation To understand existing practices and challenges of creating audio dramas, we analyzed literature from industry professionals (Wayland et al., 2020; Toscan, 2023; Bernaerts and Mildorf, 2021; De Fossard, 2005; Crook, 2002; Rattigan, 2002), expert interview videos on YouTube (Writers, 2026a; Junkie, 2026; Writers, 2026b; Media, 2026; Productions, 2026; BBC, 2026; the religious drama society, 2026; McPherson and Storytellers, 2026; Comics, 2026), and online community discussions111w.reddit.com/r/audiodrama/. We analyzed all insights using thematic analysis (Braun and Clarke, 2006, 2019; Mayring, 2021). We first extracted key insights, then moved back and forth through the material using axial coding principles (Corbin and Strauss, 2014), and finally iteratively clustered, split, and merged codes. We collected best practices (Table LABEL:tab:formative_results) and derived 4 design goals (DG1-DG4) for a system that supports creating audio dramas from books: DG1: Adapt book text into scripts that follow audio drama conventions. Books contain internal monologue, descriptive narration, and extended exposition that cannot be conveyed through sound and bore listeners if narrated verbatim (Rattigan, 2002; Bernaerts and Mildorf, 2021; Toscan, 2023; Wayland et al., 2020). Thus, adapting a book into an audio drama requires restructuring the source text so that internal thoughts are externalized into dialogue or sound events, dialogue is short and natural, exposition delivered just-in-time, and narration that restates what sound already conveys is removed (Toscan, 2023; Wayland et al., 2020; Crook, 2002). We aim to support creators by automatically producing scripts that follow existing conventions. DG2: Create distinct and consistent characters and scenes. Listeners identify characters and locations entirely through sound. Because inconsistency across characters or scenes along the story breaks immersion (De Fossard, 2005; Crook, 2002), each character needs a distinctive voice (e.g., pitch, pace, accent, word choice) and associated sound effects (e.g., footsteps) (Wayland et al., 2020; Crook, 2002; De Fossard, 2005; Rattigan, 2002). Similarly, each scene needs a distinctive ambience soundscape (e.g., rain, traffic), sound effects for character actions (e.g., footsteps, a slamming door), and music that matches the scene’s mood (Wayland et al., 2020; Toscan, 2023; Crook, 2002). We aim to support creators in defining characters and scenes as persistent narrative constructs that remain consistent across the full audio drama. DG3: Produce a balanced mix with intelligible dialogue. Because audio dramas consist of multiple tracks mixed together, a challenge is to ensure that listener attention remains on the dialogue, which should be easy to follow, without distracting sound elements (Wayland et al., 2020; Rattigan, 2002). Professionals follow a mixing hierarchy in which they place the volume of tracks relative to the dialogue (i.e., loudest). For example, sound effects are placed slightly below the volume of dialogue, music further below, and ambience at the lowest level (De Fossard, 2005; Crook, 2002; Wayland et al., 2020), with ambience and music reducing its volume when characters speak (Wayland et al., 2020; Toscan, 2023). We aim to produce a multi-track audio drama mix that makes sure dialogue is intelligible without abrupt sounds or transitions. DG4: Support editing of narrative constructs at different production stages. Audio drama production traditionally follows a sequential workflow that moves from high-level to low-level decisions. Creators first make high-level narrative decisions such as writing the script, designing the characters and scene, and then move to lower-level tasks such as audio production, timeline arrangement, and mixing (Crook, 2002; Wayland et al., 2020). Higher-level changes in later production stages are tedious and time-consuming as they may require substantial changes as all dependent audio assets must be manually identified, re-produced, and replaced across the entire project. We aim to support creators in revising narrative constructs such as characters and scenes at any time without having to restart from an earlier production stage. 4. Dramarrator Based on our design goals (DG1-DG4), we developed Dramarrator, which introduces object-based audio editing where narrative constructs (e.g., characters, scenes) are represented as objects with editable attributes (e.g., voice, ambience) linked to their dependent audio assets. We introduce object-based audio editing in (1) an editing interface where creators review and refine objects to propagate high-level changes across the project, and (2) in an automatic pipeline that extracts objects from book text to generate multi-track audio drama adaptations. 4.1. Object-Based Audio Editing Dramarrator introduces object-based audio editing based on three concepts: objects, attributes, and object-dependent assets. Objects represent narrative constructs such as characters and scenes. Each object holds attributes that describe properties of that object, for example a character’s voice design or a scene’s ambience design. Finally, each object contains a set of object-dependent assets that keep track of isolated audio assets on the timeline (e.g., speech, sound effects, music) that depend upon the object. Every generated audio asset on the timeline is linked to one or more objects at creation time. A character’s dialogue lines and sound effects (e.g., their footsteps) are linked to that character object. All sound effects, ambience layers, and music cues that occur while a scene is active are linked to that scene object. A sound effect can be linked to both a character and a scene at the same time. For example, Martha’s footsteps in the Market Square scene are linked to both the Martha character object and the Market Square scene object. When a creator modifies an object’s attribute (e.g., “make Sophie nervous”), Dramarrator uses an LLM to identify all dependent assets that require updates and proposes modification suggestions (e.g., rewriting firm spoken lines to sound nervous) for creators to accept or decline (Figure 2). Rather than manually locating and updating all affected audio assets, creators make a single object change that propagates automatically across all object-dependent assets (DG4). Figure 2. Dramarrator’s object-based editing automatically propagates edits on character or scene objects across linked assets that require updates. Specifying a higher-level edit instruction to modify a character (1), triggers Dramarrator to find and suggest edits across all affected assets on the timeline that users can choose to accept or reject (2). 4.2. Dramarrator Interface Dramarrator’s interface consists of three components: an Object Pane for reviewing and editing character and scene objects, a Script Editor for script-based editing, and a Timeline Editor for timeline-based editing (Figure 1). We follow Leila, who creates and edits audio dramas using Dramarrator. The Object Pane displays all extracted Character and Scene objects as editable vignettes with their attributes and object-dependent assets for quick review and editing (DG2, DG4). Character vignettes show a voice design attribute alongside three generated voice previews, and all object-dependent assets are accessible via a Lines and SFX tab. Scene vignettes show an ambience design attribute as an annotated prompt where individual words are linked to their corresponding ambience layers within the Ambience tab to make the prompt-to-sound mapping explicit. An SFX tab lists all sound effects in that scene. Both vignettes support natural-language edit instructions (e.g., “make Sophie nervous”), upon which Dramarrator identifies which object-dependent assets require changes and proposes modifications for Leila to accept or decline before any audio is replaced (DG2, DG4). For example, making Sophie more nervous rewrites her dialogue to sound more anxious and updates her sound effects (e.g., quick and uneven footsteps, door opening slowly with creaking noise) across the full audio drama. The Script Editor visualizes the audio drama as an annotated script with segments grouped by scene. Each scene segment consists of dialogue rows, sound effects, and music cues in script order, where any element can be inserted, moved, edited, removed or re-assigned to a different object (DG4). Voice acting directions can be added inline to any dialogue line (e.g., [angry], [whispered]), and multiple lines can be edited at once via a natural-language instruction with per-line system edit suggestions. For example, Leila adds a “medieval folk music” tag to open the market square scene, inserts a paper-crackling sound effect before Martha’s line, and rewrites a line with acting directions: “[angry] Hey! [sad] But I wanted to keep talking to you!” The Script Editor and Object Pane are linked, so that hovering over an asset in a vignette scrolls the Script Editor to that element and vice versa. It is also linked to the audio drama playback so that the currently playing word or element is always highlighted and in focus. The Timeline Editor visualizes the full composition as clips on four tracks (speech, sound effects, music, ambience) and provides clip-level controls inspired by existing timeline-based editors such as placement, fade-in and fade-out, volume adjustment, and trimming and looping (DG3, DG4). Any clip can be listened to in isolation or regenerated directly from the timeline, and both Script and Timeline Editor are synchronized so that edits in one are immediately reflected in the other. Leila adjusts timings and volumes for a final mix, then exports the composition as an .aaf file to finish mixing and mastering in her preferred DAW. Algorithm 1 Automatic Audio Drama Production 1:Input: book text T 2:Extract objects: ,←LLM(T)C,S (T) characters, scenes 3:Instantiate attributes: ,←LLM(T,,)C,S (T,C,S) voice, ambience design 4:Generate scripts: ←LLM(T,,)25P←\LLM(T,C,S)\^25 5:Score scripts: r(p)←13∑LLMjudge(p)r(p)← 13 _judge(p) 3 judges, 11-item rubric 6:Select script: p∗←arg⁡maxp∈⁡r(p)p^*← _p \ r(p) 7:Link assets: ℒ←Link(p∗,,)L (p^*,C,S) speech, SFX, music, ambience ↦ objects 8:Generate assets: ←AudioGen(ℒ,,)A (L,C,S) using object attributes 9:Arrange & mix: ←Mix()D (A) 10:Output: multi-track audio drama D Figure 3. Dramarrator takes a book as input, then (1) extracts all objects (characters, scenes) and initializes them with attributes (e.g., voice and ambience design), (2) generates 25 script candidates and scores them to choose the best based on a rubric derived from audio drama script best practices, and (3) generates and arranges all assets on timeline tracks to output a multi-track audio drama. 4.3. Automatic Audio Drama Production Dramarrator automatically converts raw book text or audio drama scripts (i.e., input) into a structured, multi-track audio drama (i.e., output) via three steps: (1) extracting objects (i.e., characters, scenes), (2) generating and selecting an adapted audio drama script, and (3) generating all audio assets and placing them onto a multi-track timeline (Figure 3). Extracting Character and Scene objects. Given a raw book text, an LLM analyzes the source text to extract characters and scenes with inferred attributes based on Dramarrator’s object model (Figure 3.1). For each character, Dramarrator infers a voice design attribute describing the character’s vocal traits (e.g., “a middle-aged woman with a measured cadence and a slight regional accent”), which determines the character’s voice across the audio drama (DG2). For each scene, Dramarrator infers an ambience design attribute describing the scene’s acoustic environment (e.g., “a busy market square with a street band playing”), which determines the scene’s soundscape whenever it is active in the audio drama (DG2). In addition, Dramarrator keeps track of dialogue lines in the source text and assigns them to their respective characters derived from their surrounding context. Creating an Audio Drama Script. To adapt the book into a script that follows audio drama conventions, we adopt a best-of-N generation strategy in which N candidate scripts are generated in parallel and the highest-quality script is selected using an LLM-based judge (Hashemi et al., 2024; Gu et al., 2024) (Figure 3.2). We set N=25N=25 based on early empirical experiments. We use an LLM to generate each candidate using a prompt based on the best practices we found in our formative work (Table LABEL:tab:formative_results). For example, the prompt instructs the LLM to remove expository narration and convert it into dialogue, externalize internal thoughts as whispered lines, replace concrete character actions with sound effects, insert voice acting directions based on emotional delivery described in the source text, and remove redundancy between characters’ speech and their sounds (DG1). For example, the source text “Martha slammed the cup down on the counter. ‘I told you not to come here,’ she said angrily.” is converted into: <sfx prompt="ceramic cup slammed hard on wooden counter" duration="1"> [VOICE: MARTHA] "[angry] I told you not to come here." The narration is removed because the sound effects already convey the action, the dialogue is kept but shortened, and the emotional delivery described in the book text (“she said angrily”) becomes a voice acting direction ([angry]). Because not every LLM generation will reliably produce a script that follows all audio drama best practices, we score all 25 candidates using an LLM-as-a-judge approach (Hashemi et al., 2024; Gu et al., 2024). We score each script candidate three times using an 11-item rubric derived from our formative work best practices (Table 2), then average the scores for reliability. We select the script candidate with the highest mean score as the final script. Generating Audio Assets and Assembling the Timeline. We use the selected script and the extracted character and scene objects to generate all audio assets (Figure 3.3). For each character, we synthesize all dialogue lines using a text-to-speech model and the character’s voice design attribute. To ensure voice consistency across the full audio drama, we concatenate all of a character’s dialogue lines into batches and synthesize them together in a single model call, then split them again using word-level alignment timestamps (DG2). We generate 4 variations for each sound effect using a text-to-SFX model and select the first variation by default to be used in the audio drama. Because speech represents the main track in audio dramas, we anchor all music and ambience to word-level timestamps of the scripted dialogue to ensure that music and ambience always align with the dialogue they accompany, even when speech durations vary across multiple generations (DG3). We use an LLM to decompose each scene’s ambience design attribute into individual ambience sound layers (e.g., “city hum”, “church bells”, “distant thunder”) using a minimized version of Schafer’s taxonomy (Schafer, 1994) as used in SoundStager (Yoo et al., 2026) but adapted to the audio drama medium. For example, we remove signal sounds (i.e., foreground sounds, designed to attract attention) as they are more suitable for videos where viewers can see the corresponding visuals, but in audio dramas, they create distractions within scenes and concrete actions already exist as sound effects. We generate each ambience layer using the same text-to-SFX model to allow editing or removing individual layers within scenes (DG2, DG4). All generated assets are assembled into a four-track timeline (speech, sound effects, music, ambience), with scene buffers (i.e., first and last 5 seconds of a scene only play ambience) and crossfades to create smooth transitions between scenes (DG1). We ensure that dialogue remains intelligible as the primary carrier of narrative information (DG3) by placing speech at the reference level, with sound effects, music, and ambience at progressively lower volumes, following existing audio drama mixing practices (Table LABEL:tab:formative_results). We apply sidechaining (i.e., reducing volume when speech is present) to scene ambience and music whenever characters are speaking to prevent other sounds from masking speech (DG3). We normalize all audio and apply a peak limiter based on audiobook standards (ACX, 2026) (DG3). 4.4. Implementation We implemented our pipeline and backend in Python using Flask. We use Gemini 3 Pro (Google, 2026a) for all LLM calls. All speech is synthesized via ElevenLabs’ eleven_v3 model (ElevenLabs, 2026a) using their word-level alignment for per-line slicing. All sound effects and music are generated through proprietary text-to-SFX and text-to-music models. All audio is normalized to −16-16 LUFS and mixed with per-track default volumes for speech (0 dB), sound effect (−6-6 dB), music (−14-14 dB), and ambience (−20-20 dB) based on best practices (Table LABEL:tab:formative_results). We implemented our frontend using React and used the Web Audio API for real-time audio playback. 5. User Study We conducted a user study with 8 professional audio drama producers to compare audio drama creation using object-based audio editing in Dramarrator to their existing tools and investigate two main questions: • How does object-based audio editing support audio drama creation compared to professionals’ existing practice? • What are professionals’ perspectives on Dramarrator’s automatically generated audio dramas? 5.1. Method We used a within-subjects design in which each participant created audio dramas from two different book excerpts, one per condition: Dramarrator and their existing tools. To isolate the effect of the production workflow and editing interface, we ensured participants using their existing tools had access to the same speech, and similar sound effects, and music generation models used by Dramarrator. Book excerpts averaged 1402 words (σ=382σ=382), with 5.7 characters (σ=1.8σ=1.8) and 4.7 scenes (σ=2.6σ=2.6). Participants always completed the existing-tools condition first, followed by Dramarrator. 5.1.1. Participants We recruited 8 participants (77 male, 11 female, ages 2020 to 5656) from Upwork (Upwork, 2026) (Table 3). All participants had prior audio editing experience (μ=6.1μ=6.1 years, σ=5.8σ=5.8 years), created or edited audio dramas or immersive audio storytelling content before, and frequently use generative audio tools in their existing workflows (e.g., ElevenLabs (ElevenLabs, 2026a), Suno (Suno, 2026), Stable Audio (Evans et al., 2024)). We compensated participants at their hourly Upwork rate ($25-$40). 5.1.2. Procedure We ran a 2-stage study. In Stage 1, participants created an audio drama using their existing tools. In Stage 2, participants used Dramarrator in a moderated session. Stage 1. Participants completed a demographic questionnaire, reviewed audio drama reference examples222https://w.audible.com/blog/article-best-classic-audio-dramatizations, and were given a book excerpt with the goal of creating a compelling, engaging, and immersive audio drama within five business days using their own tools. For each step of their workflow, participants tracked the time spent and wrote a short report describing their process. Stage 2. In a 100-minute moderated session, participants began with a 20-minute guided tutorial of Dramarrator using a sample project. Then, participants had 40 minutes to create an audio drama from a new book excerpt following the same goal as in Stage 1, followed by an unlimited time to refine their audio within 5 business days (similar to the existing tools condition). We measured total Dramarrator production time from the start of the creation task to the completion of final mixing and mastering. Participants completed a questionnaire comparing both Dramarrator and their existing tools using 5-point Likert scales drawn from the NASA-TLX (Hart and Staveland, 1988), the Creativity Support Index (Cherry and Latulipe, 2014), and the System Usability Scale (Brooke, 1996), followed by a 20-minute semi-structured interview. 5.1.3. Study limitations Participants always completed the existing-tools condition before Dramarrator. This ordering introduces two opposing effects as participants may have benefited from task-level familiarity when using Dramarrator second, but were also disadvantaged by having years of experience with their own tools compared to only a 20-minute introduction to Dramarrator. We screened participants for prior audio drama experience and assigned different book excerpts between conditions to reduce carry-over effects. Participants further had unlimited time using their existing tools to complete Stage 1 and for the final mixing and mastering at the end of Stage 2, but there was a 40-minute in-session limit using Dramarrator during Stage 2, which may underestimate what participants could accomplish with Dramarrator given more time. Participants’ time estimates were self-reported and should only be used as complementary signals to make conclusions about time savings. Lastly, while common in HCI research, our sample of 8 professionals may limit the generalizability of our findings. Figure 4. Participant ratings (N=8N=8) for questions selected from NASA-TLX, System Usability Scale, and Creativity Support Index comparing Dramarrator and their existing tools. Error bars represent a 95% confidence interval. *p< .05p\,<\,.05, **p< .01p\,<\,.01 5.2. Results All participants preferred the object-based editing paradigm of Dramarrator over their existing tools because Dramarrator significantly reduced the task load and effort required to create audio dramas. Dramarrator reduces task load and is easier to use than existing tools. Participants reported significantly lower mental demand, temporal demand, and effort when working with Dramarrator compared to their traditional tools (p< .05p\,<\,.05, Figure 4 left). Participants rated Dramarrator as significantly easier to use and faster to learn compared to their existing tools (p< .05p\,<\,.05, Figure 4 middle). Participants rated Dramarrator significantly higher on exploration (p< .05p\,<\,.05) while engagement, transparency, expressiveness, and perceived worth of effort remained comparable between conditions (Figure 4, right). We analyzed task load, creativity support, and usability ratings using paired Wilcoxon Signed-Rank tests (Woolson, 2007). We report all means, standard deviations, and significance values in Table 5. Object-based editing amplifies creative control over characters and scenes. Seven out of eight participants used object-level editing features during their sessions with an average of 2.62.6 (σ=1.4σ=1.4) object-based edits that propagated into 25.025.0 (σ=20.5σ=20.5) manual edits, a 9.7×9.7× amplification in editing effort. The amplification was highest for characters with many dialogue lines: P4’s single voice edit on Mr. Utterson propagated to all 30 of his speech lines, and P5’s voice edit on Mr. Prosser propagated to all 18 lines. P6 reimagined the narrator’s voice from “an old man’s voice, sounding like a storyteller” to “a mysterious, low-toned narrator, whispering secrets of a lost kingdom” which P6 described as “a single creative decision that automatically replaces everything across the audio”. P8, the most experienced participant (17 years) used object-based editing most extensively, for example, by iterating on 10 voice designs across 4 characters that saved 56 manual edit operations in total: “changing voices would normally take an hour, and now this was 10 seconds […] if I do this the old school way I have to manually do that on the timeline which would take me 30 minutes and here I can just change that.” While P3, who did not make edits to objects, expressed skepticism that a single AI voice prompt for a character produces consistent results across all lines, and prefers a line-by-line workflow, other participants’ workflows aligned better with Dramarrator’s object model. P1 stated the object-based approach “definitely should be a product — the tedious, annoying work is automated”, P4 found “having the ability to break down scenes and have the control — that’s brilliant”, and P7 highlighted that the object-based approach “allows me to make the story more alive and create contrast between the characters”. P5 contrasts their existing tools to object-based editing: “the objects are the diamond in the tool […] [changing a character’s voice] would take forever and would take weeks especially in a full production with voice artists who then have to re-record […] that would be almost impossible if you already have a whole book in your editor and you just want to change a character, that would be like redoing the whole project all over again.” Participants used object-based editing to support creative exploration (Figure 4, right). Our interaction logs show that participants spent 60%60\% (σ=16%σ=16\%) of their editing time in the script editor, 27%27\% (σ=14%σ=14\%) in the object pane reviewing character and scene objects, and only 9%9\% (σ=7%σ=7\%) in the timeline view. The majority of object pane interactions (μ=138.4μ=138.4, σ=72.9σ=72.9 in total) were browsing character and scene vignettes (μ=110.8μ=110.8, σ=60.8σ=60.8), followed by previewing audio from the vignettes (μ=17.1μ=17.1, σ=12.7σ=12.7), then object-level attribute edits (μ=10.5μ=10.5, σ=7.2σ=7.2). Five participants (P1, P2, P6, P7, P8) started their sessions by browsing character and scene vignettes in the sidebar to get an overview of the generated audio drama before making any edits. Participants repeatedly revisited the same objects throughout their sessions, returning to each unique character an average of 17.317.3 (σ=11.1σ=11.1) times and each scene 6.76.7 (σ=4.9σ=4.9) times, and frequently switching between different characters to compare and ensure contrast between voices. P4 noted: “just the overviews of the objects, I think that’s really good — having the ability to break down scenes and have the control — that’s brilliant”. Participants also used the sidebar to compare characters against one another: P7 highlighted character objects “allow me to make the story more alive and create contrast between the characters.” Figure 5. Self-reported time breakdown for existing tools and Dramarrator. Dramarrator reduces audio drama creation time by a factor of 4, from 16.7 (σ=7.5σ=7.5) to 3.7 (σ=2.3σ=2.3) hours. Exploratory analysis of time savings. We complement participants’ perceived efficiency gains by comparing self-reported times between the existing tools and Dramarrator condition as supporting signals. Dramarrator reduced the time to create audio dramas by a factor of 4, from 16.7 (σ=7.5σ=7.5) to 3.7 (σ=2.3σ=2.3) hours (Figure 5). With their existing tools, participants spent their time across 5 sequential phases: planning and scriptwriting (μ=2.6μ=2.6 h), voice creation and dialogue production (μ=4.2μ=4.2 h), sound effects and music creation (μ=3.9μ=3.9 h), timeline arrangement (μ=3.2μ=3.2 h), and mixing and mastering (μ=2.9μ=2.9 h). Each phase required different tools (e.g., ElevenLabs (ElevenLabs, 2026a) for speech, sound effects libraries (Sound, 2026), Suno (Suno, 2026) for music, and editing tools like Adobe Audition (Adobe, 2026a)), and changing an earlier decision such as a character’s voice required starting over from that phase. With Dramarrator, the automatic audio drama generation pipeline handled scriptwriting, voice casting, sound design, and composition automatically in 7.7 (σ=1.6σ=1.6) minutes on average, after which participants refined the result in the editing interface for approximately 40 minutes, followed by 2.9 hours of final mixing and mastering in their existing tools. All participants described the generated output as a strong starting point: P2 called it “almost all the way done”, P5 summarized it as “a sketch that I can then enhance […] with small adjustments this could be turned into a great audio experience”, and P7 highlighted that “the emotions were right, it had actual characters with personalities”, and P8 noted “it really kept the script well, the writing was not cheesy”. All participants highlighted the time savings: P1 estimated Dramarrator “saves you like 4–5 hours”, P2 noted it “might save me a day of work”, P3 estimated “5 hours just to get the initial composition”, P4 described the savings as “a lot – hours, if not days”, P5 reported it “saved me more than 15 hours”, P7 estimated the process to be “70% faster”, and P8 noted “manually creating this audio drama took me 10 hours or more, now this would just take me an hour”. Dramarrator’s workflow implications. Object-based editing shifted participants from a sequential workflow that consists of pre-production (e.g., scripting, planning characters and scenes), production (e.g., recording or generating assets), and post-production (e.g., assembling and editing assets on a timeline) to an iterative workflow that enabled more experimentation. For example, P8 noted they could make “creative changes on the fly”, a workflow that P6 described as: “[Dramarrator] allows me to listen, change something, listen, change something — otherwise I have to manually create everything first and then it’s hard to change something” Participants also used character and scene objects to shape their narrative depth within the audio drama. For example, P7 gave one character a rougher, lower voice to contrast with another, brighter-voiced character, switching back and forth between editing the voice design and listening to a dialogue sequence between the two to check that they were now clearly distinguishable. P7 described this process as a way to “add depth to each character […] and make the story more alive and create contrast.” P8 described iteratively designing characters to add distinctive sonic “color” to the system-generated result: “I can really be designing these scenes and characters and then always reuse them.” Future use cases for Dramarrator. All participants expressed that they want to use Dramarrator for future projects. P3 and P4 preferred Dramarrator as a plugin integrated in their DAW rather than a standalone tool, while P8 suggested it could also work as standalone software to cater to non-audio-engineers such as publishers and authors. Participants mentioned use cases for Dramarrator beyond converting books to audio dramas such as game design (P1, P2, P4, P6, P7, P8), immersive audio stories (P1, P3, P4), scripted podcasts (P4, P8), YouTube videos with generated audio (P1, P7), animations and trailers (P5, P8), and radio and commercials (P3). P5 envisioned a use case for film where Dramarrator may be used to create a higher-fidelity storyboard to better visualize film sequences (“like a table read […] basically an audioboard”). Participants’ concerns regarding AI audio generation and perspectives on automation. The most common concern raised across participants was AI audio generation quality. Participants highlighted acoustic inconsistency in AI-generated audio assets (speech, sound effects, music) that sometimes sound like they were recorded in different environments, and thus require post-processing such as EQ, reverb matching, and de-essing to unify the acoustic space. Participants were also sometimes frustrated by unreliable sound effect generation that required multiple regenerations to receive a satisfactory result. To address these issues, participants used object-based editing to fix voices at the character-level or remove sounds with artifacts from scene soundscapes via the scene vignettes, rather than having to manually locate and replace each affected clip individually. Participants requested improvements such as connecting audio engineering aspects (e.g., reverb, spatial mixing) to object-based edits so that scene-level acoustic environments propagate automatically (e.g., P8 describes how a cave scene should apply the same reverb to all characters within that scene), chat-based prompt interactions to refine the overall script (P5), more distinction between generation variations (P2, P8), and finer timeline controls such as gap and pacing adjustments (P1, P5). However, all participants noted that being able to export the multi-track project file allows them to repair these issues in their existing tools. Despite these quality concerns, participants viewed Dramarrator as automating tedious production tasks while keeping creative decision-making in their hands. P5 noted that Dramarrator gives them “the opportunity to be the director” rather than a technician, and P8 described the tool as filling the role of a production assistant — a role P8 typically hires for — freeing them to focus on creative work and final mixing. P1 similarly valued that “the tedious, annoying work is automated” while retaining full control over creative choices. A secondary concern expressed by some participants (P1, P8) was long-term job displacement by AI automation, though this was framed in the context of broader industry trends rather than Dramarrator specifically. P8 also raised a concern about artistic homogenization and thus AI tools removing individual creative distinctiveness: “if everybody uses that tool there might be a tendency that everyone’s [audio] sounds the same, so I might want to add my own texture.” P3 raised a concern about copyright and questioned the legal ownership of AI-generated audio when using generated assets in their professional work. Figure 6. Listener ratings (N=300N=300) comparing audio dramas across three conditions: automatically generated by Dramarrator, refined by professionals using Dramarrator, and produced by professionals using their existing tools. Error bars represent a 95% confidence interval. *p< .05p\,<\,.05, **p< .01p\,<\,.01 5.3. Listener Evaluation The user study above helped us understand creators’ perceptions of Dramarrator and its output quality. To complement these findings from the audience side, we further evaluate the output quality through a listening study. Specifically, we collected ratings from 300 listeners on Prolific (Prolific, 2026) comparing the raw output from our pipeline (Dramarrator-Auto), the versions that creators refined based on the pipeline output (Dramarrator-Refine), and the golden examples that creators created using their existing tools during Stage 1 of the user study (Existing-Tools). We randomly chose 6 stories from our user study. Each listener was assigned three stories and rated one randomly chosen condition per story. To assess different dimensions of listener experience, we derived attentional focus, emotional engagement, and narrative presence from the User Engagement Scale (Busselle and Bilandzic, 2009), and asked listeners to rate overall audio and story quality, five audio element arrangement and mixing questions, and character and scene consistency on 7-point Likert scales. We analyzed ratings using Friedman tests with pairwise Wilcoxon tests and Bonferroni correction (Figure 6). Below, we summarize the key findings. Human refinement helps bridge the gap to professional quality. Listeners rated Existing-Tools higher than Dramarrator-Auto on all 11 measures (p< .01p\,<\,.01). However, Dramarrator-Refine significantly improved Dramarrator-Auto on all 11 measures (p< .01p\,<\,.01), achieving ratings comparable to Existing-Tools on 7 measures, including user engagement, overall audio and story quality, and character and scene consistency. Notably, character and scene consistency was rated on par with professionally produced audio dramas. Object-based audio editing in Dramarrator likely has supported maintaining consistency, as attribute changes to character or scene objects automatically propagate to all dependent assets, compared to manual clip-by-clip timeline editing in existing tools. A gap remains in audio element distraction, clarity, placement, and timing. Despite the significant improvement after human refinement, listeners still rated Existing-Tools higher than Dramarrator-Refine on distraction, clarity, placement, and timing of audio elements (p< .05p\,<\,.05). We identified two potential sources for this gap: (1) AI generation artifacts (e.g., unnatural sound effects) that may go unnoticed during refinement but become apparent to listeners, and (2) constraints of Dramarrator (e.g., uniform pauses between speakers, no sound effects overlapping with speech). 6. Exploratory Study with Storytellers Beyond Audio Drama While Dramarrator was designed with audio drama creators in mind, we were also interested in whether the tool could lower the barrier to entry enough that storytellers without audio production expertise might consider creating immersive audio stories, and whether the principles behind Dramarrator may generalize to other domains of storytelling. To this end, we invited three participants (Table 4) to a 90-minute session in which they explored Dramarrator while discussing its usefulness and the potential of object-based editing within their creative domains: an escape room designer (N1), a tabletop RPG enthusiast (N2), and a computer game designer (N3). All three found Dramarrator easy to use and expressed interest in using Dramarrator to add immersive audio stories to their own work. While professionals valued Dramarrator primarily for the time it saved, novices valued that it made audio drama creation accessible to them at all. N3 noted “I definitely would not be able to create audio dramas without experience but now I could”. All three participants connected object-based editing to pain points in their own domains. N1 described how a client requesting a character change in an off-the-shelf escape room they sell, currently requires “basically creating a fully new project doing everything manually”. N2 noted that tabletop RPG game masters often need to adapt scenarios on the fly, and suggested that object-based editing could allow real-time swapping of character voices, scene ambience, or dialogue tone. N3 observed that high-level creative changes in game production grow increasingly costly the later they occur, since every dependent asset must be updated manually. This suggests that object-based editing addresses a broader challenge in narrative content creation that extends beyond audio drama production. 7. Discussion and Future Work Our evaluations show that Dramarrator addressed our design goals. In our user study, Dramarrator lowered professionals’ task load and object-based audio editing enabled iterative shaping of characters and scenes (DG2, DG4). A listener evaluation revealed that automatically generated audio dramas refined via object-based audio editing in Dramarrator were comparable to those created by professionals using their existing tools on engagement, audio and story quality, and character and scene consistency measures (DG1, DG2, DG3). We reflect on our findings and discuss future opportunities to extend and generalize object-based audio editing in Dramarrator. 7.1. Extending the Object Model Dramarrator’s object attributes currently cover narrative properties (e.g., voice design, ambience design), but professionals in our user study requested extensions that would deepen the range of edits a single object change can capture. Participants requested acoustic properties as object attributes. For example, P8 described how a cave scene should automatically apply the same reverb to all characters speaking within it, rather than requiring manual post-processing for each asset. Just as a character’s voice design attribute currently defines their vocal identity and propagates to all their speech assets, scene objects could carry an acoustic environment attribute (e.g., room size, reverb) that is applied to all assets linked to that scene. We can apply Dramarrator to arbitrarily long audio dramas up to the size of the LLM’s context window. However, our object model enforces consistency of characters and scenes, which may change over the course of a story (e.g., a character may age). Borrowing object inheritance from object-oriented programming (Rentsch, 1982; Stefik and Bobrow, 1985) would allow the creation of multiple instances of the same object (e.g., child Alice, adult Alice) that share base attributes but differ in attributes based on the story plot (e.g., adult Alice’s voice is heavier). Future iterations could also extend the usage timeline in Dramarrator’s object vignettes with a visualization that shows when attributes are active, and let creators drag attribute ranges to specific story segments, similar to animation tools (Adobe, 2026b; Blender, 2026; Boom, 2026) where creators define when a property applies via keyframes. For relationship-based conditions (e.g., a character’s tone shifting depending on who they are with), an object could hold conditional attributes that activate based on other objects that co-occur in the same scene. Dramarrator currently relies on an LLM’s latent knowledge to translate high-level object edits across asset modalities (e.g., how a character’s heightened anxiety should modulate their speech, accompanying Foley, and background music simultaneously). While flexible, this approach lacks transparent formal logic that creators can inspect or override. To make this mapping more interpretable, future systems could implement a cross-modal mapping that is grounded in a shared dimension. For example, prior work showed that valence and arousal (Russell, 1980) can be used as an emotional dimension that maps onto individual modalities, such as text (Juslin and Scherer, 2005), speech (Preoţiuc-Pietro et al., 2016), music (Rubin and Agrawala, 2014), and sounds (Fan et al., 2017). This suggests there may be a shared emotion space that could inform a cross-modal mapping to handle dependencies. For example, after a creator makes edits to a scene and therefore shifts its emotional valence, a system could identify all assets which are now out of alignment with the shift in valence to inform what assets need to be updated, rather than relying on an LLM. However, regenerating assets to achieve a desired emotion shift remains an open challenge. 7.2. Generalization of Object-Based Editing Our informal study with 3 novices from adjacent creative domains revealed that the object-based editing paradigm may generalize beyond audio drama production. All three highlighted that their existing tools force them to edit each dependent asset individually, even though they already conceptualize their creative content through characters, scenes, and environments. These observations suggest that wherever a creative domain is organized around narrative constructs with many dependent assets, reifying those constructs as editable objects with propagating attributes (Beaudouin-Lafon, 2000; Xia et al., 2016) could provide a similar effort amplification. For example, in video production, character and scene objects could propagate changes across dependent visual sub-channels such as lighting, camera framing, or character expressions, where the main challenge compared to audio is that regenerating visual assets risks introducing inconsistencies with surrounding frames. 7.3. Implications for Collaborative Work While Dramarrator was designed for individual creators to reflect an emerging practice of individuals expanding their ability to create with AI (Daugherty and Wilson, 2024; Louie et al., 2020; Luo, 2025), we see the potential for object-based audio editing to also expand collaborative workflows. Even without AI asset generation, the object model could serve as a coordination layer among human collaborators. For example, character objects could export per-character casting briefs for voice directors, or scene objects could track which assets are pending or delivered. In the future, we can also enable real-time collaboration in Dramarrator by locking assets linked to objects that are currently modified to prevent edit conflicts. Lastly, instead of replacing voice actors, who offer qualities AI cannot simply replicate (Jiang et al., 2025), Dramarrator’s AI-generated audio drama outputs can serve as high-fidelity prototypes that support creators to concretize story, character, and scene decisions before human talent is engaged — similar to preliminary “table-read” sessions before final studio recordings (Proferes, 2005; Leach, 2013). 7.4. Creative Agency and Ethical Implications Professionals described Dramarrator as freeing them to act as creative directors rather than manual assemblers, because the system handles low-level production so they can focus on shaping characters and scenes. Producers can iterate through many character designs or scene atmospheres and select the best, rather than committing early to save revision costs. For example, P8 iterated over 10 voice designs across 4 characters in a single session, a workflow that previously required hours of manual timeline work. Handing low-level production to the system also increases the abstraction distance from the concrete assets, which introduces new tensions. Some participants found higher-level control liberating (P4, P6, P8), while others found the added flexibility disorienting or felt less ownership over the result (P4, P8). This reflects a known tension in AI-assisted creativity tools between automation that amplifies creative capacity and automation that shifts creative authorship toward the system (Li et al., 2023). Dramarrator shows this tension at the level of object attributes. A character’s voice design or a scene’s ambience is a natural-language prompt that determines how the object appears across the project, yet creators currently cannot inspect or override the normative design guidelines that Dramarrator uses to infer such prompts from the book text (Table LABEL:tab:formative_results). In the future, we should let creators edit, extend, or replace these guidelines for more control. Object-based editing also shifts the creator’s effort from making each edit to reviewing the edits the system proposes, and creators should be able to decide how much review each propagation deserves. For edits where ownership matters, we can introduce selective good friction (Chen and Schmidt, 2024; Cox et al., 2016) to increase creators’ sense of ownership, for example by requiring creators to write the regeneration prompt for each affected asset themselves so that the review effort stays meaningful. For routine edits, we can instead reduce the review burden, for example by learning from previously accepted edits to auto-approve similar ones. Dramarrator uses generative AI tools, such as LLMs (Google, 2026a) and audio generation models (ElevenLabs, 2026a; Adobe, 2026c), which carries creative potential but also ethical concerns. Generative AI tools can augment the creative process by turning creators’ intents into assets almost instantly, and thus significantly speeding up the time between instantiation, assessment, and refinement of ideas (Tseng et al., 2024; Riche et al., 2025). Furthermore, generative AI tools lower barriers to content creation and often make creating content accessible in the first place (Wessel et al., 2025; Constantinides et al., 2018; Huy et al., 2024; Pinski and Benlian, 2024). For example, creating audio dramas traditionally required specialized expertise and significant resources, but our studies showed Dramarrator lowers barriers that once limited who could take part in creating audio drama content. Lowering barriers and automation also raises concerns about long-term job displacement from AI automation, which were also voiced by some participants in our user studies (P1, P8). For example, P8 described using AI in place of a production assistant they used to hire for parts of their work. P3 and P8 also worried about who owns content created this way. P3 questioned the copyright and legal ownership of AI-generated audio used in professional work, because P3 felt the audio was not their own work. P8 worried that if many creators use the same AI tools, audio content could start to sound alike, so creators may need to add their own texture to stay distinctive. 8. Conclusion We presented Dramarrator, an audio drama authoring system that introduces object-based audio editing to support creators in adapting books into audio dramas. Dramarrator extracts narrative constructs (e.g., characters, scenes) from book text or original scripts as objects with attributes (e.g., a character’s voice, a scene’s ambience), automatically generates a multi-track audio drama, and links each asset to its object so that high-level edits propagate automatically across all dependent assets. A user study with 8 professionals showed that Dramarrator significantly lowered task load compared to existing tools. Object-based audio editing amplified editing effort by 9.7×9.7× on average, with a single object change replacing up to 30 manual edits while keeping characters and scenes consistent across the entire audio. A listener study (N=300) revealed that automatically generated audio dramas refined via object-based audio editing in Dramarrator were comparable to those created by professionals using their existing tools on engagement, audio and story quality, and character and scene consistency measures. Finally, an exploratory study with 3 storytellers without audio expertise showed that Dramarrator lowers the barrier to audio drama creation and object-based editing may generalize to other domains organized around narrative constructs with cascading asset dependencies. We hope our work inspires future research into object-based editing abstractions for narrative content creation tools. References ACX (2026) ACX audio submission requirements. Note: Accessed: 2026 External Links: Link Cited by: §4.3. Adobe (2026a) Adobe audition. Note: Accessed: 2026 External Links: Link Cited by: §1, §2.1, §5.2. Adobe (2026b) After effects. Note: Accessed: 2026 External Links: Link Cited by: §7.1. Adobe (2026c) Firefly. Note: Accessed: 2026 External Links: Link Cited by: §1, §7.4. A. Agostinelli, T. I. Denk, Z. Borsos, J. Engel, M. Verzetti, A. Caillon, Q. Huang, A. Jansen, A. Roberts, M. Tagliasacchi, M. Sharifi, N. Zeghidour, and C. Frank (2023) MusicLM: Generating Music From Text. arXiv. External Links: Link, Document Cited by: §2.3. Apple (2026) Logic pro. Note: Accessed: 2026 External Links: Link Cited by: §1, §2.1. Avid (2026) Pro tools. Note: Accessed: 2026 External Links: Link Cited by: §1, §2.1. Y. BBC (2026) Kenneth branagh and david tennant on the art of radio drama (bbc radio 4). Note: Accessed: 2026 External Links: Link Cited by: §3. M. Beaudouin-Lafon (2000) Instrumental interaction: an interaction model for designing post-WIMP user interfaces. In Proceedings of the SIGCHI conference on Human factors in computing systems, p. 446–453. Cited by: §1, §2.2, §7.2. K. Benharrak, P. Peng, and A. Pavel (2025) TalkLess: Blending Extractive and Abstractive Summarization for Editing Speech to Preserve Content and Style. In Proceedings of the 38th Annual ACM Symposium on User Interface Software and Technology, UIST ’25, New York, NY, USA, p. 1–19. External Links: ISBN 979-8-4007-2037-6, Link, Document Cited by: §2.1. L. Bernaerts and J. Mildorf (2021) Audionarratology: lessons from radio drama. Ohio State University Press. Cited by: Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, §3, §3. Blender (2026) Blender. Note: Accessed: 2026 External Links: Link Cited by: §7.1. T. Boom (2026) Harmony. Note: Accessed: 2026 External Links: Link Cited by: §7.1. Z. Borsos, R. Marinier, D. Vincent, E. Kharitonov, O. Pietquin, M. Sharifi, D. Roblek, O. Teboul, D. Grangier, M. Tagliasacchi, and N. Zeghidour (2023) AudioLM: a Language Modeling Approach to Audio Generation. arXiv. External Links: Link, Document Cited by: §1, §2.3. V. Braun and V. Clarke (2006) Using thematic analysis in psychology. Qualitative Research in Psychology 3 (2), p. 77–101. External Links: ISSN 1478-0887, Link, Document Cited by: §3. V. Braun and V. Clarke (2019) Reflecting on reflexive thematic analysis. Qualitative Research in Sport, Exercise and Health 11 (4), p. 589–597. External Links: ISSN 2159-676X, Link, Document Cited by: §3. J. Brooke (1996) SUS: A quick and dirty usability scale. Usability Evaluation in Industry. Cited by: §5.1.2. R. Busselle and H. Bilandzic (2009) Measuring Narrative Engagement. Media Psychology 12 (4), p. 321–347. External Links: ISSN 1521-3269, Link, Document Cited by: §5.3. Z. Chen and R. Schmidt (2024) Exploring a behavioral model of “positive friction” in human-AI interaction. In International Conference on Human-Computer Interaction, p. 3–22. Cited by: §7.4. E. Cherry and C. Latulipe (2014) Quantifying the Creativity Support of Digital Tools through the Creativity Support Index. ACM Trans. Comput.-Hum. Interact. 21 (4), p. 21:1–21:25. External Links: ISSN 1073-0516, Link, Document Cited by: §5.1.2. Y. W. Comics (2026) How to write an audio drama and the collaborations that come with it.. Note: Accessed: 2026 External Links: Link Cited by: Table 1, Table 1, Table 1, §3. P. Constantinides, O. Henfridsson, and G. G. Parker (2018) Introduction—platforms and infrastructures in the digital age. Vol. 29, Informs. Cited by: §7.4. J. Copet, F. Kreuk, I. Gat, T. Remez, D. Kant, G. Synnaeve, Y. Adi, and A. Défossez (2023) Simple and controllable music generation. In Proceedings of the 37th International Conference on Neural Information Processing Systems, NIPS ’23, Red Hook, NY, USA, p. 47704–47720. Cited by: §2.3. J. Corbin and A. Strauss (2014) Basics of Qualitative Research: Techniques and Procedures for Developing Grounded Theory. SAGE Publications (en). External Links: ISBN 978-1-4833-0086-3 Cited by: §3. A. L. Cox, S. J. Gould, M. E. Cecchinato, I. Iacovides, and I. Renfree (2016) Design frictions for mindful interactions: The case for microboundaries. In Proceedings of the 2016 CHI conference extended abstracts on human factors in computing systems, p. 1389–1397. Cited by: §7.4. T. Crook (2002) Radio drama. Routledge. Cited by: Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, §1, §3, §3, §3, §3, §3. P. R. Daugherty and H. J. Wilson (2024) Human+ machine, updated and expanded: reimagining work in the age of AI. Harvard Business Press. Cited by: §7.3. E. De Fossard (2005) Writing and producing radio dramas. Vol. 1, Sage. Cited by: Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, §1, §3, §3, §3. Descript (2026) Descript. Note: Accessed: 2026 External Links: Link Cited by: §2.1. ElevenLabs (2026a) Eleven v3. Note: Accessed: 2026 External Links: Link Cited by: §1, §2.3, §4.4, §5.1.1, §5.2, §7.4. ElevenLabs (2026b) ElevenLabs. Note: Accessed: 2026 External Links: Link Cited by: §1. Z. Evans, J. D. Parker, C. J. Carr, Z. Zukowski, J. Taylor, and J. Pons (2024) Stable Audio Open. arXiv. External Links: Link, Document Cited by: §5.1.1. J. Fan, M. Thorogood, and P. Pasquier (2017) Emo-soundscapes: A dataset for soundscape emotion recognition. In 2017 Seventh International Conference on Affective Computing and Intelligent Interaction (ACII), p. 196–201. External Links: ISSN 2156-8111, Link, Document Cited by: §7.1. Figma (2026) Figma. Note: Accessed: 2026 External Links: Link Cited by: §1, §2.2. Google (2026a) Gemini 3 pro. Note: Accessed: 2026 External Links: Link Cited by: §4.4, §7.4. Google (2026b) NotebookLM. Note: Accessed: 2026 External Links: Link Cited by: §2.3. J. Gu, X. Jiang, Z. Shi, H. Tan, X. Zhai, C. Xu, W. Li, Y. Shen, S. Ma, H. Liu, et al. (2024) A survey on llm-as-a-judge. The Innovation. Cited by: §4.3, §4.3. Y. Guo, T. Wang, Y. Ge, S. Ma, Y. Ge, W. Zou, and Y. Shan (2025) AudioStory: Generating Long-Form Narrative Audio with Large Language Models. arXiv. External Links: Link, Document Cited by: §2.3. H. L. Han, J. Yu, R. Bournet, A. Ciorascu, W. E. Mackay, and M. Beaudouin-Lafon (2022) Passages: Interacting with Text Across Documents. In Proceedings of the 2022 CHI Conference on Human Factors in Computing Systems, CHI ’22, New York, NY, USA, p. 1–17. External Links: ISBN 978-1-4503-9157-3, Link, Document Cited by: §1, §2.2. S. G. Hart and L. E. Staveland (1988) Development of NASA-TLX (Task Load Index): Results of empirical and theoretical research. In Advances in psychology, Vol. 52, p. 139–183. Cited by: §5.1.2. H. Hashemi, J. Eisner, C. Rosset, B. Van Durme, and C. Kedzie (2024) Llm-rubric: A multidimensional, calibrated approach to automated evaluation of natural language texts. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), p. 13806–13834. Cited by: §4.3, §4.3. L. V. Huy, H. T. Nguyen, T. Vo-Thanh, N. H. T. Thinh, T. Thi Thu Dung, et al. (2024) Generative AI, why, how, and outcomes: A user adoption study. AIS Transactions on Human-Computer Interaction 16 (1), p. 1–27. Cited by: §7.4. H. H. Jiang, W. Agnew, T. Friedlander, Z. Yang, S. E. Fox, M. S. Bernstein, J. C. Passananti, M. Ogata, and K. Ortiz (2025) Forging an HCI Research Agenda with Artists Impacted by Generative AI. In Proceedings of the Extended Abstracts of the CHI Conference on Human Factors in Computing Systems, p. 1–4. Cited by: §7.3. Y. B. Junkie (2026) Talking audio drama sound design with kenny neal — booth junkie. Note: Accessed: 2026 External Links: Link Cited by: Table 1, Table 1, Table 1, Table 1, Table 1, §3. P. N. Juslin and K. R. Scherer (2005) Vocal expression of affect. In The New Handbook of Methods in Nonverbal Behavior Research, J. A. Harrigan, R. Rosenthal, and K. R. Scherer (Eds.), p. 0. External Links: ISBN 978-0-19-852961-3, Link, Document Cited by: §7.1. T. S. Kim, Y. Lee, M. Chang, and J. Kim (2023) Cells, Generators, and Lenses: Design Framework for Object-Oriented Interaction with Large Language Models. In Proceedings of the 36th Annual ACM Symposium on User Interface Software and Technology, UIST ’23, New York, NY, USA, p. 1–18. External Links: ISBN 979-8-4007-0132-0, Link, Document Cited by: §1, §2.2. F. Kreuk, G. Synnaeve, A. Polyak, U. Singer, A. Défossez, J. Copet, D. Parikh, Y. Taigman, and Y. Adi (2023) AudioGen: Textually Guided Audio Generation. arXiv. External Links: Link, Document Cited by: §2.3. P. Laban, E. Ye, S. Korlakunta, J. Canny, and M. Hearst (2022) NewsPod: automatic and interactive news podcasts. In Proceedings of the 27th International Conference on Intelligent User Interfaces, IUI ’22, New York, NY, USA, p. 691–706. External Links: ISBN 9781450391443, Link, Document Cited by: §2.3. R. Leach (2013) Theatre studies: The basics. Routledge. Cited by: §7.3. J. Li, E. Rawn, J. Ritchie, J. Tran O’Leary, and S. Follmer (2023) Beyond the Artifact: Power as a Lens for Creativity Support Tools. In Proceedings of the 36th Annual ACM Symposium on User Interface Software and Technology, UIST ’23, New York, NY, USA, p. 1–15. External Links: ISBN 979-8-4007-0132-0, Link, Document Cited by: §1, §2.2, §7.4. X. Liu, Z. Zhu, H. Liu, Y. Yuan, Q. Huang, M. Cui, J. Liang, Y. Cao, Q. Kong, M. D. Plumbley, and W. Wang (2025) WavJourney: Compositional Audio Creation With Large Language Models. IEEE Transactions on Audio, Speech and Language Processing 33, p. 2830–2844. External Links: Document Cited by: §2.3. R. Louie, A. Coenen, C. Z. Huang, M. Terry, and C. J. Cai (2020) Novice-AI music co-creation via AI-steering tools for deep generative models. In Proceedings of the 2020 CHI conference on human factors in computing systems, p. 1–13. Cited by: §7.3. Y. Luo (2025) Designing with AI: A systematic literature review on the use, development, and perception of AI-enabled UX design tools. Advances in Human-Computer Interaction 2025 (1), p. 3869207. Cited by: §7.3. P. Mayring (2021) Qualitative content analysis: A step-by-step guide. Qualitative Content Analysis, p. 1–100. Cited by: §3. Y. R. McPherson and S. A. Storytellers (2026) How to start your audio drama podcast with producer brooke pillifant. Note: Accessed: 2026 External Links: Link Cited by: Table 1, Table 1, Table 1, Table 1, §3. Y. C. C. Media (2026) HOW to create an audio drama — fictional podcast. Note: Accessed: 2026 External Links: Link Cited by: Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, §3. M. Pinski and A. Benlian (2024) AI literacy for users–A comprehensive review and future research directions of learning methods, components, and effects. Computers in Human Behavior: Artificial Humans 2 (1), p. 100062. Cited by: §7.4. D. Preoţiuc-Pietro, H. A. Schwartz, G. Park, J. Eichstaedt, M. Kern, L. Ungar, and E. Shulman (2016) Modelling Valence and Arousal in Facebook posts. In Proceedings of the 7th Workshop on Computational Approaches to Subjectivity, Sentiment and Social Media Analysis, A. Balahur, E. van der Goot, P. Vossen, and A. Montoyo (Eds.), San Diego, California, p. 9–15. External Links: Link, Document Cited by: §7.1. Y. F. &. S. Productions (2026) Writing for audio fiction - audio drama podcasting with k. a. statz. Note: Accessed: 2026 External Links: Link Cited by: §3. N.T. Proferes (2005) Film Directing Fundamentals: See Your Film Before Shooting. Focal Press. External Links: ISBN 978-0-240-80562-7, Link, LCCN 2004019069 Cited by: §7.3. Prolific (2026) Prolific. Note: Accessed: 2026 External Links: Link Cited by: §5.3. D. Rattigan (2002) Theatre of Sound: Radio and the Dramatic Imagination. Carysfort Press. Cited by: Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, §1, §3, §3, §3, §3. T. Rentsch (1982) Object oriented programming. ACM Sigplan Notices 17 (9), p. 51–57. Cited by: §1, §2.2, §7.1. N. Riche, A. Offenwanger, F. Gmeiner, D. Brown, H. Romat, M. Pahud, N. Marquardt, K. Inkpen, and K. Hinckley (2025) AI-Instruments: Embodying Prompts as Instruments to Abstract & Reflect Graphical Interface Commands as General-Purpose Tools. In Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems, CHI ’25, New York, NY, USA, p. 1–18. External Links: ISBN 979-8-4007-1394-1, Link, Document Cited by: §7.4. S. Rubin and M. Agrawala (2014) Generating emotionally relevant musical scores for audio stories. In Proceedings of the 27th annual ACM symposium on User interface software and technology, UIST ’14, New York, NY, USA, p. 439–448. External Links: ISBN 978-1-4503-3069-5, Link, Document Cited by: §7.1. S. Rubin, F. Berthouzoz, G. J. Mysore, W. Li, and M. Agrawala (2013) Content-based tools for editing audio stories. In Proceedings of the 26th annual ACM symposium on User interface software and technology, UIST ’13, New York, NY, USA, p. 113–122. External Links: ISBN 978-1-4503-2268-3, Link, Document Cited by: §2.1. J. A. Russell (1980) A circumplex model of affect.. Journal of personality and social psychology 39 (6), p. 1161. Cited by: §7.1. R. M. Schafer (1994) The Soundscape: Our Sonic Environment and the Tuning of the World. Destiny Books, Rochester (English). External Links: ISBN 978-1-59477-668-7 Cited by: §4.3. H. V. Shin, W. Li, and F. Durand (2016) Dynamic Authoring of Audio with Linked Scripts. In Proceedings of the 29th Annual Symposium on User Interface Software and Technology, UIST ’16, New York, NY, USA, p. 509–516. External Links: ISBN 978-1-4503-4189-9, Link, Document Cited by: §2.1. E. Sound (2026) Sound effects. Note: Accessed: 2026 External Links: Link Cited by: §5.2. M. Stefik and D. G. Bobrow (1985) Object-oriented programming: Themes and variations. AI magazine 6 (4), p. 40–40. Cited by: §1, §2.2, §7.1. S. Suh, M. Lai, K. Pu, S. P. Dow, and T. Grossman (2025) StoryEnsemble: Enabling Dynamic Exploration & Iteration in the Design Process with AI and Forward-Backward Propagation. In Proceedings of the 38th Annual ACM Symposium on User Interface Software and Technology, UIST ’25, New York, NY, USA, p. 1–36. External Links: ISBN 979-8-4007-2037-6, Link, Document Cited by: §2.2. Suno (2026) AI music generator. Note: Accessed: 2026 External Links: Link Cited by: §1, §5.1.1, §5.2. Y. R. the religious drama society (2026) How to write a radio drama. Note: Accessed: 2026 External Links: Link Cited by: §3. R. Toscan (2023) Writing audio drama: making scripts that work for fiction & true crime podcasts. Independently published. External Links: ISBN 979-8378489947, Link Cited by: Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, §1, §3, §3, §3, §3. T. Tseng, R. Cheng, and J. Nichols (2024) Keyframer: Empowering animation design using large language models. arXiv preprint arXiv:2402.06071. Cited by: §7.4. Upwork (2026) Upwork. Note: Accessed: 2026 External Links: Link Cited by: §5.1.1. K. Wayland, W. Lucas, and S. Ryle (2020) Bombs always beep - 2nd edition - revenge of the beep: creating modern audio theater. Amazon Digital Services LLC - KDP Print US. External Links: ISBN 9781648585388, Link Cited by: Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, §1, §1, §3, §3, §3, §3, §3. M. Wessel, M. Adam, A. Benlian, A. Majchrzak, and F. Thies (2025) Generative AI and its transformative value for digital platforms. Journal of Management Information Systems 42 (2), p. 346–369. Cited by: §7.4. R. F. Woolson (2007) Wilcoxon signed-rank test. Wiley encyclopedia of clinical trials, p. 1–3. Cited by: §5.2. Y. B. Writers (2026a) Scriptwriting advice from the bbc radio drama north team. Note: Accessed: 2026 External Links: Link Cited by: Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, §3. Y. B. Writers (2026b) Writing for radio - advice from the team in bbc radio drama north. Note: Accessed: 2026 External Links: Link Cited by: Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, Table 1, §3. H. Xia, B. Araujo, T. Grossman, and D. Wigdor (2016) Object-Oriented Drawing. In Proceedings of the 2016 CHI Conference on Human Factors in Computing Systems, CHI ’16, New York, NY, USA, p. 4610–4621. External Links: ISBN 978-1-4503-3362-7, Link, Document Cited by: §1, §2.2, §7.2. H. Xia, B. Araujo, and D. Wigdor (2017) Collection Objects: Enabling Fluid Formation and Manipulation of Aggregate Selections. In Proceedings of the 2017 CHI Conference on Human Factors in Computing Systems, CHI ’17, New York, NY, USA, p. 5592–5604. External Links: ISBN 978-1-4503-4655-9, Link, Document Cited by: §1, §2.2. H. Xia (2020) Crosspower: Bridging Graphics and Linguistics. In Proceedings of the 33rd Annual ACM Symposium on User Interface Software and Technology, UIST ’20, New York, NY, USA, p. 722–734. External Links: ISBN 978-1-4503-7514-6, Link, Document Cited by: §1. X. Xu, J. Mei, C. Li, Y. Wu, M. Yan, S. Lai, J. Zhang, and M. Wu (2025) M-StoryAgent: Immersive Narrated Storybook Video Generation with a Multi-Agent Paradigm across Text, Image and Audio. arXiv. External Links: Link, Document Cited by: §2.3. T. Yeh and J. Kim (2018) CraftML: 3D Modeling is Web Programming. In Proceedings of the 2018 CHI Conference on Human Factors in Computing Systems, CHI ’18, New York, NY, USA, p. 1–12. External Links: ISBN 978-1-4503-5620-6, Link, Document Cited by: §1, §2.2. S. Yoo, A. H. Santisteban, P. Seetharaman, J. Salamon, O. Nieto, and A. Truong (2026) SoundStager: Interactive Design of Story-Driven GenAI Soundscapes for Video. (en). Cited by: §2.3, §4.3. Appendix A Appendix Table 1. Guidelines for creating audio dramas, derived from qualitative content analysis of professional audio drama literature. Guidelines are organized by the four design goals (DG1–DG4) they inform. Theme Goal How Why Sources DG1: Produce an effective audio drama script Script Structure & Pacing Keep scenes short and momentum-driven Target individual scenes of under 1–2 minutes of audio (≈ 1–2 script pages) and complete episodes of 25–30 minutes (22–27 pages). After drafting, ask of every scene: “Does this scene advance the central conflict?” Merge adjacent scenes if combining them loses nothing. Cut any scene that only recaps prior events without advancing conflict. Audio listeners have no visual anchor; a prolonged scene without action or conflict causes attention to drift. Short scenes maintain urgency and momentum. (Toscan, 2023; Wayland et al., 2020; Writers, 2026a, b) Deliver exposition just-in-time; never front-load it Scatter all background information (character history, relationships, world-building) as small fragments within natural dialogue exchanges across multiple scenes. Never open an episode with an expository monologue or narrator backstory dump. Apply the rule: there are no introductions in dramatic writing—begin every scene mid-action and let exposition emerge organically from what characters say to each other. Front-loading background information overwhelms listeners and kills engagement before the story can grip them; listeners disengage before the drama has a chance to begin. (Toscan, 2023; Wayland et al., 2020; Writers, 2026a, b) Open each episode with a compelling hook; close to invite re-engagement Begin every episode in medias res—at a moment of conflict, tension, or mystery—within the first 30 seconds. Do not open with scene-setting narration or pleasantries. End each episode with an unresolved question, a cliffhanger, or a high emotional moment that leaves the listener in suspense. Never use narration to recap the previous episode. Audio is a passive medium; listeners can disengage effortlessly. The opening hook secures their commitment; the closing hook drives return engagement across episodes. (De Fossard, 2005; Writers, 2026a, b) Dialogue Craft Write tight, natural, conversational lines Enforce the right-margin rule: each spoken line should be no longer than ≈ 17 words. Avoid multiple consecutive sentences in one character’s turn. Read every line aloud—if it sounds unnatural when spoken, rewrite it. Cut filler phrases, pleasantries, and any line that neither advances the plot nor reveals character. Avoid literary constructions that work on the page but confuse listeners when heard. Dense or literary dialogue crowds out the SFX and ambience that make audio drama immersive; it sounds stilted when performed and strains listener comprehension. (Toscan, 2023; Wayland et al., 2020; Writers, 2026a, b) Limit each scene to two characters (two-handers) Design every scene as a dialogue between exactly two characters. When a third character must briefly appear, give them a single functional line and remove them immediately. Audit the entire script and cut any character who: (a) appears for less than one full script page in total, (b) has no direct connection to the central conflict, (c) exists solely as a technical device (e.g., a server who takes an order and is never heard from again). More than two simultaneous voices creates confusion about character identity in a purely auditory medium, undermining story comprehension. (Toscan, 2023; Wayland et al., 2020; Crook, 2002; Writers, 2026a, b) Write for the ear, not the eye Favor short, concrete, everyday words over long or abstract ones. Test each sentence by reading aloud at performance pace and mark any phrase that feels awkward or trips the tongue—rewrite it. Prefer active voice and simple sentence structures. Avoid tongue-twisters, complex nested clauses, and multi-part lists within a single sentence. If a sentence requires the listener to hold too many concepts before resolving, break it into two. Listeners cannot pause and re-read; every sentence must land on the first pass. Phonetically difficult or structurally complex lines cause comprehension failure that cannot be corrected in a live listening experience. (Crook, 2002; Junkie, 2026) Narration & Prose Adaptation Minimize narration; use it only when no other option exists Before accepting any narrator line as final, ask: “Can this be conveyed through dialogue or sound?” Retain narration only for four cases where audio has no alternative: (1) character physical appearance, (2) colors, (3) a specific named location that cannot be established through ambience alone, (4) an action that would be genuinely ambiguous without verbal clarification. Cut or convert every other narrator line. No single narrator passage should exceed ≈ 30 seconds of spoken audio. Narration is described as “diluted poison” for audio drama: small amounts are tolerable, but excess stalls narrative momentum and regresses the production toward the audiobook format, destroying immersive potential. (Wayland et al., 2020; Toscan, 2023; Bernaerts and Mildorf, 2021) When narration is used, make it intimate and conversational Cast the central character as the internal narrator (first-person homodiegetic). Write narrator lines as if the character is confiding privately to a trusted listener: use contractions, casual vocabulary, and natural pauses. Avoid formal or literary sentence structures in narrator lines. Keep narrator monologues to 15–30 seconds before returning to scene action. Record the narrator voice dry (no reverb, close-mic) to reinforce intimacy with the listener. A formal, literary narrator voice breaks the immersive illusion of a performed scene and reminds the listener they are hearing a narration; an informal narrator creates an intimate bond that sustains engagement. (Toscan, 2023; Bernaerts and Mildorf, 2021; Media, 2026) Externalize internal monologue and prose description into dialogue or sound When adapting a prose source, identify every passage of: (a) internal thought (e.g., “She wondered if he knew”), (b) descriptive narration (e.g., “The room smelled of old books”), (c) reported speech (e.g., “He told her he was leaving”). For each passage, apply one of three conversion strategies: (1) Dialogue conversion—rewrite internal thoughts as direct speech between characters or brief first-person narrator asides; (2) Sound conversion—replace environmental description with an ambience or SFX annotation (e.g., [AMBIENCE: dusty, close acoustic space; faint creaking of shelves]); (3) Delivery conversion—encode subtext in the actor’s delivery direction rather than in narration. Audio drama is a performative, not descriptive, medium. Internal thought and prose description are invisible to listeners and must be externalized into the audible world to exist. This is the central challenge of adapting prose fiction for audio. (Rattigan, 2002; Bernaerts and Mildorf, 2021; Toscan, 2023) Avoid cross-modal redundancy between narration and sound After assembling a scene with all SFX and ambience, re-read every narrator and character line while listening to the mix. Flag any line describing an action the listener can already hear (e.g., remove “just a minute, while I open this crate” if the crate-opening SFX is audible in the mix). Cut or substantially rephrase all flagged lines. Deliberate redundancy—for comic effect or critical emphasis—is only permissible as a conscious, purposeful choice. When sound already tells the listener what is happening, narrating the same information creates boredom, wastes scene time, and signals a lack of trust in the listener’s intelligence, damaging narrative engagement. (Wayland et al., 2020; Toscan, 2023; Bernaerts and Mildorf, 2021; Media, 2026) DG2: Give characters and scenes distinct, consistent acoustic identities Character Design & Voice Give each character an immediately identifiable voice Before scripting, write a one-page voice profile for each character defining: (1) pitch range relative to other characters (high/low), (2) speaking pace (fast, measured, or halting), (3) accent or dialect, (4) vocabulary level (formal, colloquial, or technical), (5) habitual phrases or verbal tics (e.g., always beginning sentences with “Look,” or trailing off mid-sentence). Ensure no two main characters share more than one of these features. Actors must internalize and maintain their profiles throughout the entire production. Voice is the sole identifier of character in audio drama; indistinct voices cause listeners to confuse characters and lose track of the story. (Wayland et al., 2020; Crook, 2002; De Fossard, 2005; Rattigan, 2002; Writers, 2026a, b; McPherson and Storytellers, 2026) Ensure listeners always know who is speaking Use character names frequently and naturally within dialogue—at least once every 3–4 exchanges (e.g., “I hear you, Sarah, but…” rather than “I hear you, but…”). Dedicate a new character’s entire first scene to establishing their voice; never introduce a character in a single line. Use recurring story devices to re-anchor identity: a catchphrase, a distinctive sound cue on entry, or a structural habit (e.g., one character always asks questions, another always gives commands). In a purely auditory medium, name repetition and consistent character habits are the primary anchors for listener comprehension; without them, listeners lose track of who is driving the scene. (Wayland et al., 2020; Toscan, 2023; Comics, 2026) Use voice texture to convey personality and emotion beyond the words Direct actors on the physical qualities of their voice: breath patterns (controlled and even = authority; ragged and shallow = fear or excitement), resonance placement (deep chest voice = calm authority; high head voice = vulnerability), and pace modulation (deliberate slowing = gravity; accelerating = anxiety). Write explicit delivery directions in the script alongside each dialogue line (e.g., [barely above a whisper, breath catching]). Reinforce these qualities with VOX processing in post-production. The texture, grain, and physicality of a voice communicates character depth that no amount of descriptive narration can match; listeners form emotional connections to voice quality before they consciously process the words. (Rattigan, 2002; Crook, 2002; Bernaerts and Mildorf, 2021; Writers, 2026a, b; McPherson and Storytellers, 2026) Sound Design Layer all four sound categories in every scene Treat each scene as four independent tracks, all of which must be consciously addressed: (1) VOX—any processing applied to character voices to reflect acoustic context (e.g., telephone band-pass filter, room reverb, megaphone effect, whisper treatment); (2) SFX—discrete sound events triggered by specific actions (e.g., a door opening, glass shattering, footsteps); (3) Ambience—the continuous environmental background running under the entire scene (e.g., wind, traffic, crowd murmur, rain); (4) Music—the emotional underscore. Before producing each scene, confirm that each track has at least one intentional element assigned to it. Each layer contributes a distinct dimension of the story world; missing any single layer creates an incomplete aural image—a scene without ambience sounds staged and artificial; a scene without music lacks emotional direction. (Wayland et al., 2020; Toscan, 2023; Crook, 2002; Writers, 2026a, b; McPherson and Storytellers, 2026) Foley every character–object interaction with character-specific sounds For every moment a character physically handles or interacts with an object, generate or record a corresponding close-proximity Foley sound. This applies to: picking up or setting down any object, footsteps (distinct per character), using a phone, handling paper, opening or closing a door, sitting in or rising from a chair, writing. Do not use generic SFX library sounds for on-scene character actions—library sounds lack intimacy and acoustic consistency with the scene. The manner of interaction must be encoded in the sound: a character slamming a cup down versus placing it gently are two different Foley takes that communicate two different emotional states. The way a character physically interacts with objects expresses their emotional state through sound—the primary form of physical emotional expression in a medium without visuals. (Wayland et al., 2020) Establish location with ambience before any character speaks Begin every scene by fading in the ambience track for 2–5 seconds before dialogue starts. Select a combination of 2–3 specific environmental sounds that unambiguously signal the location (e.g., hospital: HVAC hum + distant PA announcement + squeaking wheels; beach: ocean waves + seagulls + wind). Once location is established and dialogue begins, reduce ambience by 6–10 dB so it sits beneath speech for the rest of the scene. Maintain the ambience continuously—never cut it mid-scene unless the location changes. Listeners need a spatial anchor before they can process dialogue effectively; beginning a scene mid-dialogue with no acoustic environment leaves the listener spatially disoriented, reducing both comprehension and immersion. (Wayland et al., 2020; Toscan, 2023; Comics, 2026; Junkie, 2026) Deliberately classify each sound element as diegetic or non-diegetic For every sound element, ask: “Can the characters in this scene hear this?” Diegetic [D] sounds (characters can hear them): SFX, Foley, and source music (e.g., a record player in the room). Non-diegetic [ND] sounds (only the listener hears them): emotional underscore music and stylized sound design. Label each element in the script as [D] or [ND] and strictly enforce the distinction: if the underscore is non-diegetic, characters must never react to it; if a sound is diegetic, characters may respond to it in dialogue or action. Accidentally blurring diegetic and non-diegetic breaks the internal logic of the story world; listeners unconsciously track this consistency, and a violation produces a subliminal sense that something is wrong, eroding immersion. (Bernaerts and Mildorf, 2021; Crook, 2002; McPherson and Storytellers, 2026; Junkie, 2026) Trust the listener’s imagination; avoid over-specifying every sound After finalizing a scene’s sound design, remove each SFX element one at a time and ask: “Does removing this sound genuinely confuse the listener about what is happening?” If the answer is no, remove it. Avoid stacking multiple simultaneous SFX for a single narrative moment. Leave deliberate sonic gaps for the listener’s imagination to fill in. Guiding rule: not every physical action needs a corresponding SFX—a character walking to a window may need only a subtle ambience shift, not explicit footstep Foley. Audio drama uniquely activates the listener’s imagination in a way no visual medium can; over-specifying every sound replaces imaginative engagement with passive reception, reducing immersion rather than increasing it. (Crook, 2002; Rattigan, 2002; Wayland et al., 2020; Media, 2026) Music Underscore each scene’s emotional tone with scene-specific music Before selecting or composing music, write a single-word emotional descriptor for each scene (e.g., “dread,” “tenderness,” “urgency,” “relief”). Choose music that precisely matches this descriptor—not generic background music. Introduce the music cue at scene start or at the moment of an emotional shift mid-scene. Set music level ≈ 15–18 dB below dialogue. Attenuate music in the 250–3000 Hz range to prevent frequency masking of speech. Change the music cue whenever the scene’s emotional tone shifts substantially. Music operates on listeners’ emotions faster and more reliably than dialogue; the right underscore ensures listeners feel a scene’s emotional intent before they consciously process the words, making emotional engagement more immediate. (Wayland et al., 2020; Toscan, 2023; De Fossard, 2005; Crook, 2002; Writers, 2026a, b) Use music to establish acoustic presence in interior scenes with no natural ambience For interior scenes in acoustically neutral settings (e.g., an empty office, a bare room) where no recognizable environmental sound is available, use a sustained, low-energy musical pad or synthesized electronic texture as a substitute ambience. Place it at ambience level (≈ −-45 to −-50 LUFS). Match the tone to the scene’s emotional register: a warm, slow pad for intimacy; a cold, high-frequency drone for unease. If any faint location-specific sound exists (e.g., HVAC hum, distant traffic), layer it with the musical pad. Complete acoustic silence beneath dialogue reads as a production error rather than an artistic choice; it creates listener discomfort and collapses the sense of spatial presence within the scene. (Toscan, 2023; Wayland et al., 2020; Media, 2026) Establish recurring musical motifs for each major character and the show itself Compose or select a distinctive 5–10 second melodic or rhythmic motif for each major character. Play it at underscore level whenever the character is introduced or is the scene’s emotional focus. Create a 20–30 second signature theme tune for the show’s opening. Use a 2–3 second transition sting (a short, resolving musical phrase) as the standard sonic device between scenes. Repeat each motif consistently and never substitute it—listeners must form an unconscious association between motif and character. Recurring motifs create emotional shorthand: after a few scenes, hearing a character’s motif alone triggers the listener’s full emotional association with that character, deepening narrative investment without any additional dialogue. (De Fossard, 2005; Crook, 2002) Scene Transitions & Acoustic Space Communicate every scene change through sound, not dialogue Use two techniques together: (1) Contrasting ambiences—fade out Scene A’s ambience, then fade in Scene B’s distinctly different ambience (e.g., busy café → quiet forest); (2) Carrier sounds—use a sound that physically performs the transition (footsteps walking out, a door closing, a car driving away, a phone hanging up). Insert a brief 1–2 second silence or a transition sting between the two ambiences to mark the cut. Never write dialogue that telegraphs a scene change (e.g., “Well, I should be going now”). Dialogue-telegraphed transitions are clunky and unnatural; sound-based transitions are seamless and preserve dramatic momentum without requiring the listener to consciously acknowledge a location change. (Wayland et al., 2020; Toscan, 2023; Crook, 2002; Rattigan, 2002; Comics, 2026; Media, 2026) Signal temporal jumps through a consistent, distinctive sound device Design a single unique temporal-shift audio signature and use it every time the narrative moves forward or backward in time. Options include: a distinct musical sting with a long reverb tail; an abrupt cut to silence followed by a different acoustic texture; a pitch-shifted or time-stretched sound sweep; or a shift from stereo to mono for scenes set in the past. For flashback sequences, apply a consistent audio treatment throughout (e.g., warmer EQ, more compression, longer reverb decay). Establish this device in the first episode and never deviate from it. Audio drama has no visual chapter breaks or title cards; without an explicit audio cue, a temporal jump leaves listeners confused about where they are in the narrative timeline—confusion that compounds if left unresolved. (Bernaerts and Mildorf, 2021) Use acoustic perspective to place each character physically in the scene space Apply consistent acoustic positioning: Close/foreground (arm’s reach): dry voice, no reverb, full frequency range. Across the room: add medium room reverb, slightly reduce presence frequencies (2–5 kHz), lower volume 4–6 dB. Through a wall: add heavy reverb, high-frequency rolloff above 4 kHz, lower volume 8–12 dB. On a telephone: apply a band-pass filter (300–3400 Hz), add mild distortion, compress dynamic range. Maintain each character’s assigned acoustic position consistently within a scene; shift it only if they physically move in the narrative. Without visual cues, all characters default to sounding equally close to the listener; acoustic positioning is the only tool for creating spatial realism and preventing the scene from sounding like a flat, stageless recording. (Crook, 2002; Bernaerts and Mildorf, 2021) Acoustic Focalization & Listener Agency Choose whose acoustic point of view the listener inhabits in each scene At the top of each scene in the script, write a focalization note: “Listener hears through [Character]’s perspective.” Implement this by: (a) including only sounds that character could physically hear from their position in the scene; (b) using that character’s acoustic location as the spatial reference point for all other sounds; (c) withholding sounds from events happening elsewhere to create deliberate information asymmetry. For dramatic irony, momentarily shift focalization to include a sound the focal character cannot hear (e.g., the quiet sound of a door opening behind them while they speak). Acoustic focalization controls information flow, builds suspense, and enables dramatic irony—the audio equivalent of film’s camera POV. Used deliberately, it creates a uniquely intimate listener–character bond. (Bernaerts and Mildorf, 2021) Use deliberate silence as an active dramatic tool At moments of high dramatic weight—a revelation, a key decision, an unexpected twist—simultaneously cut all ambience, music, and SFX tracks, inserting 2–4 seconds of near-complete silence. Mark this explicitly in the script as [SILENCE: 3 sec]. Build toward the silence by gradually removing sound elements beforehand (remove music first, then SFX, then ambience) to amplify anticipation. Distinguish planned silence from a production gap by ensuring it is precisely timed and followed by a deliberate, impactful sound event. The sudden withdrawal of all sound forces the listener’s full attention, creates a physical sense of anticipation, and makes the subsequent sound—a single word, a gunshot, one musical note—maximally impactful. (Crook, 2002; Rattigan, 2002; Bernaerts and Mildorf, 2021) DG3: Ensure production-quality audio mixing Production Consistency Maintain acoustic and character consistency throughout the entire production Create and maintain a production bible documenting for each character: voice actor, microphone model and position, recording room, and EQ/compression settings. Document for each recurring location: the ambience file name and version, and reverb settings (room size, pre-delay, decay). Reuse these exact settings every time a character or location reappears. Never substitute a different voice actor mid-production; if a re-recording is necessary, precisely match the original session’s acoustic conditions. Acoustic inconsistency—a character sounding different between scenes, or the same location sounding different across episodes—immediately breaks immersion; listeners detect these mismatches unconsciously even when they cannot articulate what is wrong. (De Fossard, 2005; Crook, 2002; Junkie, 2026) Audio Mixing Apply a consistent audio mixing hierarchy so that dialogue always remains intelligible Target these integrated loudness levels (LUFS) in the final mix: Dialogue/Voice: −-24 LUFS (reference level; must never be masked by other elements). Music underscore: −-30 LUFS (≈ 6 dB below voice). SFX: −-33 to −-39 LUFS (adjusted by narrative importance). Ambience: −-45 to −-60 LUFS (background only). Additionally, apply a high-pass filter to music at 200–300 Hz and a low-pass at 8–10 kHz to prevent music frequencies from masking speech. Verify the final mix on at least two playback systems (headphones and loudspeakers). If music or SFX mask dialogue at any moment, the primary content of the drama—what characters say—is irreversibly lost for the listener; the mixing hierarchy ensures each sound layer serves its role without interfering with the others. (De Fossard, 2005; Crook, 2002; Media, 2026) DG4: Support iterative refinement across script and audio Production Workflow Write the script as a complete production blueprint, not just a dialogue document Use a standardized script format throughout: character names in Small Caps flush left, dialogue below, and all sound cues in [SQUARE BRACKETS] at the exact script position where they occur. Each sound cue annotation must specify: (a) type (SFX / AMBIENCE / MUSIC / VOX), (b) a precise description—not “car sound” but “[SFX: heavy diesel engine idling, then cutting out abruptly]”, (c) its approximate duration or whether it is continuous, (d) its emotional function (e.g., “[MUSIC: low tension underscore, establishes dread—continues under dialogue]”). Include a Character Voice Profile page and a Locations & Ambience reference page at the front of the script. Ambiguous scripts require verbal clarification between writers, directors, and sound designers, introducing inconsistencies and rework; a fully annotated script serves as a complete production specification that enables direct translation from script to audio. (De Fossard, 2005; Media, 2026; Junkie, 2026) Dimension What is scored Dialogue efficiency Lines are short (≤ 17 words), natural, one thought per turn Narration discipline Narrator used only when information cannot be conveyed using other sounds Exposition handling Backstory scattered in dialogue, never front-loaded Narrative clarity Zero-background listener can follow; each line responds to the prior one Scene propulsion Every scene opens mid-action and ends on unresolved tension SFX specificity Prompts name material and manner (e.g., “leather soles scraping bark”) SFX discipline SFX are spaced, not stacked; each is necessary Sound design coverage Every scene has at least one SFX or music cue Music emotional precision Cue prompts name the emotion; placement aligns with tonal shifts Adaptation fidelity All key events preserved; internal thought externalized Cross-modal redundancy No text restates what a co-located SFX already conveys Table 2. 11 rubric dimensions used by the LLM judge to score script candidates. Each dimension is rated on a 1–5 scale. ID Age Gender Country Experience Expertise* Professional Background Frequently used Generative AI Tools P1 31 M Morocco 12 years 3 Audio series, film production, and dubbing ElevenLabs P2 49 F USA 1 year 2 Audio editing, noise removal, and mixing ElevenLabs, Descript P3 47 M Greece 8 years 3 Music composition and sound design (children’s, sci-fi) ElevenLabs P4 56 M UK 3 years 3 Audiobook narration and production (children’s) ElevenLabs P5 39 M N. Macedonia 3 years 3 Sound design, music integration, and voiceover audiobooks ElevenLabs, Cartesia, Stable Audio P6 20 M Costa Rica 1 year 2 Mixing and sound design (thrillers, mystery) ElevenLabs, Suno P7 29 M Montenegro 4 years 3 Audio production and voiceover ElevenLabs P8 44 M Philippines 17 years 3 Directing, sound design for radio drama, podcasts, and commercials (sci-fi, action, crime, horror, fantasy) ElevenLabs, Suno, Udio, Hume, Gemini Table 3. User study participants (P1–P8). All participants were recruited from Upwork with 6.1 years of audio editing experience (σ=5.8σ=5.8) and all have created or edited audio dramas or immersive audio content before. *Self-reported audio editing expertise (1 = Beginner, 2 = Intermediate, 3 = Advanced). Table listing the 8 user study participants with their age, gender, country, years of audio drama experience, self-reported expertise level, professional background, and generative AI tools used in their existing workflows. ID Background Relevant Projects N1 Escape room designer Digital and physical escape rooms, gamified e-books, interactive learning content N2 Tabletop RPG enthusiast DnD campaigns, narrative RPG scenarios, interactive fiction N3 Game designer Game production, dialogue scripting, level and scene design Table 4. Informal exploratory study participants (N1–N3). None had prior audio production experience. Table listing the 3 informal exploratory study participants with their background and relevant project domains. Measure Existing Practice Dramarrator p NASA-TLX Mental Demand μ=4.25μ=4.25, σ=0.89σ=0.89 μ=2.00μ=2.00, σ=1.07σ=1.07 .0078.0078 ∗⁣∗** Temporal Demand μ=3.88μ=3.88, σ=0.35σ=0.35 μ=2.25μ=2.25, σ=0.89σ=0.89 .0156.0156 ∗* Performance μ=4.38μ=4.38, σ=1.06σ=1.06 μ=3.88μ=3.88, σ=0.83σ=0.83 .3438.3438 Effort μ=4.62μ=4.62, σ=0.74σ=0.74 μ=1.75μ=1.75, σ=0.89σ=0.89 .0078.0078 ∗⁣∗** Frustration μ=3.00μ=3.00, σ=1.31σ=1.31 μ=1.88μ=1.88, σ=0.83σ=0.83 .1250.1250 Creativity Support Index Exploration μ=2.62μ=2.62, σ=0.92σ=0.92 μ=4.12μ=4.12, σ=1.13σ=1.13 .0391.0391 ∗* Engagement μ=4.62μ=4.62, σ=1.06σ=1.06 μ=4.38μ=4.38, σ=0.92σ=0.92 1.0001.000 Worth the Effort μ=4.38μ=4.38, σ=0.92σ=0.92 μ=4.88μ=4.88, σ=0.35σ=0.35 .3750.3750 Transparency μ=3.00μ=3.00, σ=1.31σ=1.31 μ=3.88μ=3.88, σ=1.36σ=1.36 .0625.0625 Expressiveness μ=4.62μ=4.62, σ=0.52σ=0.52 μ=4.00μ=4.00, σ=1.07σ=1.07 .2500.2500 System Usability Scale Frequent Use μ=4.00μ=4.00, σ=0.76σ=0.76 μ=4.62μ=4.62, σ=0.74σ=0.74 .2812.2812 Easy to use μ=2.88μ=2.88, σ=1.25σ=1.25 μ=4.62μ=4.62, σ=0.52σ=0.52 .0312.0312 ∗* Quick to learn μ=1.62μ=1.62, σ=0.74σ=0.74 μ=4.38μ=4.38, σ=0.74σ=0.74 .0078.0078 ∗⁣∗** Confident Use μ=4.50μ=4.50, σ=0.76σ=0.76 μ=4.38μ=4.38, σ=1.06σ=1.06 1.0001.000 Lot to learn μ=4.25μ=4.25, σ=1.49σ=1.49 μ=2.00μ=2.00, σ=1.07σ=1.07 .0156.0156 ∗* Table 5. Within-subject comparison of existing practice and Dramarrator (n=8n=8). We report means (μ) and sample standard deviations (σ) per condition, and p-values from two-tailed paired Wilcoxon Signed Rank tests. p∗<.05^*p<.05, p∗<.01^**p<.01.

Instant research discovery