LAION · MOSS-8B voice-acting · prompting ablation

Best-of-60 — three rewards compared

Best-of-60, three rewards side by side (for dummies). For every setting we generated 60 spoken takes and here show the single best one chosen by three different scoring rules — so you can hear how the "winner" changes depending on what you optimise:

· all — invWER × (normalised prompt-sim + blend + genuineness): everything, including matching the direction.
· original — invWER × (normalised blend + genuineness): the reward used in the earlier study.
· adherence — prompt-sim × invWER: rewards matching the written "how to say it" above all.

Scores: invWER (1=all words right), WER (0=perfect), prompt-sim (VoiceCLAP direction↔audio cosine, higher=better match), blend (polished), genu (genuine). Expand "original DramaBox prompt" for the source text.

CCcca_voicenet_10979_cc cca_voicenet
SAY: You know, I really do care about you. It's just that you make it so hard sometimes, the way you never listen, always so sure of yourself.
original DramaBox prompt & direction text

Full DramaBox prompt (how to say — with the "quoted" dialogue):
A mature androgynous adult with a deeply throaty, cavernous resonance and a deliberately labored, sedate tempo, delivering this high-quality studio voice recording with no background noise. The same voice is intensely affectionate yet laced with contempt, a loving but scornful tone. (Drawling, half-smile in voice.) "You know, I really do care about you. It's just that you make it so hard sometimes, the way you never listen, always so sure of yourself." CUT TO: The same voice now hollow with sudden

Direction only (dialogue removed — this is what prompt-sim compares against):
A mature androgynous adult with a deeply throaty, cavernous resonance and a deliberately labored, sedate tempo, delivering this high-quality studio voice recording with no background noise. The same voice is intensely affectionate yet laced with contempt, a loving but scornful tone. (Drawling, half-smile in voice.) CUT TO: The same voice now hollow with sudden

What to say (goes in MOSS text field): You know, I really do care about you. It's just that you make it so hard sometimes, the way you never listen, always so sure of yourself.

with_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.849invWER1.000WER0.000prompt-sim+0.164blend6.368genu1.767
original: invWER × (blend + genu)
orig1.104invWER0.964WER0.036prompt-sim-0.020blend6.902genu2.099
adherence: prompt-sim × invWER
ps0.309invWER0.893WER0.107prompt-sim+0.346blend4.052genu0.979
with_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.676invWER1.000WER0.000prompt-sim+0.211blend4.927genu1.284
original: invWER × (blend + genu)
orig1.005invWER0.893WER0.107prompt-sim+0.065blend7.178genu1.794
adherence: prompt-sim × invWER
ps0.211invWER1.000WER0.000prompt-sim+0.211blend4.927genu1.284
no_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.774invWER1.000WER0.000prompt-sim+0.164blend6.082genu1.463
original: invWER × (blend + genu)
orig1.065invWER1.000WER0.000prompt-sim-0.022blend6.158genu2.072
adherence: prompt-sim × invWER
ps0.218invWER1.000WER0.000prompt-sim+0.218blend2.090genu0.644
no_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.698invWER1.000WER0.000prompt-sim+0.114blend5.194genu1.894
original: invWER × (blend + genu)
orig0.971invWER1.000WER0.000prompt-sim+0.070blend5.796genu1.708
adherence: prompt-sim × invWER
ps0.202invWER1.000WER0.000prompt-sim+0.202blend3.175genu1.280
CCcca_voicenet_10276_cc cca_voicenet
SAY: I swear, if that package doesn't show up today, I'll—but no, it's coming. They promised. And my granddaughter painted me a card, you know? She's seven. She uses too much glitter. I'll keep it forever. It's just... the card. That's all I hav
original DramaBox prompt & direction text

Full DramaBox prompt (how to say — with the "quoted" dialogue):
A very old man with a brisk, forward-pushing tempo, a low, resonant but tremulous and breathy voice, delivering this high-quality studio voice recording with no background noise. The same voice is impatient and irritable but threaded with hope and affection, as if waiting for something anticipated. "I swear, if that package doesn't show up today, I'll—but no, it's coming. They promised. And my granddaughter painted me a card, you know? She's seven. She uses too much glitter." (pause) "I'll keep it forever." CUT TO: The same voice now hollowed by grief, the hope drained away, speaking with a fragile, stunned quiet. "It's just... the card. That's all I have now. The glitter's still on my fingers. She was supposed to be here. She was supposed to—" (weakly) "—be here." The performance across both moments should feel like a single natural breath, from bustling expectancy to devastation, with the same old-man tremor and storybook cadence carrying through both.

Direction only (dialogue removed — this is what prompt-sim compares against):
A very old man with a brisk, forward-pushing tempo, a low, resonant but tremulous and breathy voice, delivering this high-quality studio voice recording with no background noise. The same voice is impatient and irritable but threaded with hope and affection, as if waiting for something anticipated. (pause) CUT TO: The same voice now hollowed by grief, the hope drained away, speaking with a fragile, stunned quiet. (weakly) The performance across both moments should feel like a single natural breath, from bustling expectancy to devastation, with the same old-man tremor and storybook cadence carrying through both.

What to say (goes in MOSS text field): I swear, if that package doesn't show up today, I'll—but no, it's coming. They promised. And my granddaughter painted me a card, you know? She's seven. She uses too much glitter. I'll keep it forever. It's just... the card. That's all I have now. The glitter's still on my fingers. She was supposed to be here. She was supposed to— —be here.

with_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.570invWER1.000WER0.000prompt-sim+0.166blend3.393genu1.901
original: invWER × (blend + genu)
orig0.824invWER0.718WER0.282prompt-sim+0.112blend7.169genu1.937
adherence: prompt-sim × invWER
ps0.203invWER1.000WER0.000prompt-sim+0.203blend1.653genu1.381
with_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.639invWER1.000WER0.000prompt-sim+0.196blend4.644genu1.332
original: invWER × (blend + genu)
orig0.884invWER0.831WER0.169prompt-sim+0.227blend6.946genu1.550
adherence: prompt-sim × invWER
ps0.206invWER0.972WER0.028prompt-sim+0.212blend0.694genu0.960
no_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.709invWER0.915WER0.085prompt-sim+0.126blend7.443genu1.429
original: invWER × (blend + genu)
orig1.003invWER0.915WER0.085prompt-sim+0.126blend7.443genu1.429
adherence: prompt-sim × invWER
ps0.292invWER0.915WER0.085prompt-sim+0.319blend2.236genu1.285
no_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.794invWER0.944WER0.056prompt-sim+0.257blend5.865genu1.804
original: invWER × (blend + genu)
orig0.937invWER0.944WER0.056prompt-sim+0.257blend5.865genu1.804
adherence: prompt-sim × invWER
ps0.243invWER0.944WER0.056prompt-sim+0.257blend5.865genu1.804
CCcca_voicenet_2546_cc cca_voicenet
SAY: The door slammed shut behind me and I knew there was no turning back, I found the switch and the whole room flooded with light—I did it. And then nothing, just the silence and the dust settling, I sat down on the floor and couldn't move for
original DramaBox prompt & direction text

Full DramaBox prompt (how to say — with the "quoted" dialogue):
A very old feminine voice with a reedy, breathy quality and a pronounced physiological tremor, the mask resonance overwhelmingly bright yet the entire recording is severely degraded, sounding like a low-bitrate file with heavy background hum. The same voice is trembling with raw fear, yet a fierce thread of triumph cuts through her unsteady breath as she narrates a narrow escape. (She speaks as if reliving a harrowing moment, her breath catching, but a grim smile breaking through the tremor) "The door slammed shut behind me and I knew there was no turning back, I found the switch and the whole room flooded with light—I did it." CUT TO: The same voice now drained of all that fire, heavy with the stunned exhaustion of aftermath, the tremor more pronounced as if each word costs her. (Her voice cracks and slows, the labored breath audible between syllables as she stares at nothing) "And then nothing, just the silence and the dust settling, I sat down on the floor and couldn't move for a long time." The performance across both moments should feel like a survivor recounting a single ordeal from the peak of adrenaline through the collapse into hollow stillness.

Direction only (dialogue removed — this is what prompt-sim compares against):
A very old feminine voice with a reedy, breathy quality and a pronounced physiological tremor, the mask resonance overwhelmingly bright yet the entire recording is severely degraded, sounding like a low-bitrate file with heavy background hum. The same voice is trembling with raw fear, yet a fierce thread of triumph cuts through her unsteady breath as she narrates a narrow escape. (She speaks as if reliving a harrowing moment, her breath catching, but a grim smile breaking through the tremor) CUT TO: The same voice now drained of all that fire, heavy with the stunned exhaustion of aftermath, the tremor more pronounced as if each word costs her. (Her voice cracks and slows, the labored breath audible between syllables as she stares at nothing) The performance across both moments should feel like a survivor recounting a single ordeal from the peak of adrenaline through the collapse into hollow stillness.

What to say (goes in MOSS text field): The door slammed shut behind me and I knew there was no turning back, I found the switch and the whole room flooded with light—I did it. And then nothing, just the silence and the dust settling, I sat down on the floor and couldn't move for a long time.

with_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.464invWER1.000WER0.000prompt-sim+0.095blend3.514genu1.616
original: invWER × (blend + genu)
orig0.819invWER0.827WER0.173prompt-sim-0.041blend6.261genu1.523
adherence: prompt-sim × invWER
ps0.123invWER0.923WER0.077prompt-sim+0.133blend1.725genu0.947
with_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.349invWER1.000WER0.000prompt-sim+0.102blend2.717genu1.355
original: invWER × (blend + genu)
orig0.696invWER0.981WER0.019prompt-sim-0.145blend3.656genu1.430
adherence: prompt-sim × invWER
ps0.102invWER1.000WER0.000prompt-sim+0.102blend2.717genu1.355
no_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.805invWER1.000WER0.000prompt-sim+0.161blend6.456genu1.444
original: invWER × (blend + genu)
orig0.998invWER1.000WER0.000prompt-sim+0.161blend6.456genu1.444
adherence: prompt-sim × invWER
ps0.161invWER1.000WER0.000prompt-sim+0.161blend6.456genu1.444
no_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.461invWER1.000WER0.000prompt-sim-0.026blend4.759genu1.591
original: invWER × (blend + genu)
orig0.847invWER1.000WER0.000prompt-sim-0.026blend4.759genu1.591
adherence: prompt-sim × invWER
ps0.105invWER0.635WER0.365prompt-sim+0.165blend0.227genu1.009
CCcc2c_archetype_8315_cc cc2c_archetype
SAY: I just... I thought the chart was for room three. But the numbers didn't line up. I'm sorry, I should have double-checked. I don't know what I was thinking. No, that's not acceptable. I need the real numbers now. You will explain this to me
original DramaBox prompt & direction text

Full DramaBox prompt (how to say — with the "quoted" dialogue):
A 27-year-old female with a hushed, slightly breathy timbre, delivering this high-quality studio voice recording with no background noise. The same voice is frozen in confusion and embarrassment, speaking with agonizing restraint as she tries to justify a mistake. (voice barely above a whisper) "I just... I thought the chart was for room three." (swallows hard) "But the numbers didn't line up." (glances at the floor) "I'm sorry, I should have double-checked." (trails off) "I don't know what I was thinking." CUT TO: The same voice now sharp with indignation and sudden clarity, speaking with cold, restrained fury. (stares directly ahead) "No, that's not acceptable. I need the real numbers now." (firmly) "You will explain this to me right now." (flat but intense) "I've had enough of the confusion." The performance across both moments should feel like the same person transforming from paralyzed self-doubt to assertive clarity.

Direction only (dialogue removed — this is what prompt-sim compares against):
A 27-year-old female with a hushed, slightly breathy timbre, delivering this high-quality studio voice recording with no background noise. The same voice is frozen in confusion and embarrassment, speaking with agonizing restraint as she tries to justify a mistake. (voice barely above a whisper) (swallows hard) (glances at the floor) (trails off) CUT TO: The same voice now sharp with indignation and sudden clarity, speaking with cold, restrained fury. (stares directly ahead) (firmly) (flat but intense) The performance across both moments should feel like the same person transforming from paralyzed self-doubt to assertive clarity.

What to say (goes in MOSS text field): I just... I thought the chart was for room three. But the numbers didn't line up. I'm sorry, I should have double-checked. I don't know what I was thinking. No, that's not acceptable. I need the real numbers now. You will explain this to me right now. I've had enough of the confusion.

with_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.915invWER1.000WER0.000prompt-sim+0.196blend6.793genu1.694
original: invWER × (blend + genu)
orig1.141invWER1.000WER0.000prompt-sim+0.122blend7.153genu1.907
adherence: prompt-sim × invWER
ps0.227invWER1.000WER0.000prompt-sim+0.227blend4.672genu1.409
with_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.507invWER0.881WER0.119prompt-sim+0.092blend5.516genu1.902
original: invWER × (blend + genu)
orig1.023invWER0.983WER0.017prompt-sim-0.180blend0.967genu5.304
adherence: prompt-sim × invWER
ps0.115invWER0.746WER0.254prompt-sim+0.154blend3.771genu0.974
no_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.946invWER0.983WER0.017prompt-sim+0.147blend7.364genu2.058
original: invWER × (blend + genu)
orig1.166invWER0.983WER0.017prompt-sim+0.147blend7.364genu2.058
adherence: prompt-sim × invWER
ps0.200invWER1.000WER0.000prompt-sim+0.200blend5.554genu1.913
no_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.557invWER0.966WER0.034prompt-sim-0.077blend2.206genu4.559
original: invWER × (blend + genu)
orig1.043invWER1.000WER0.000prompt-sim-0.207blend1.766genu4.803
adherence: prompt-sim × invWER
ps0.089invWER0.983WER0.017prompt-sim+0.091blend2.289genu1.352
CCcc2c_archetype_4348_cc cc2c_archetype
SAY: This bloodwork should've been flagged two hours ago. Who signed off on this? Come on, people. We did everything... everything right. Why does that never feel like enough?
original DramaBox prompt & direction text

Full DramaBox prompt (how to say — with the "quoted" dialogue):
A 34-year-old woman with a tight, slightly clipped voice, delivering this high-quality studio voice recording with no background noise. The same voice is bristling with controlled impatience, barely tolerating a delay in her morning rounds. (tapping a pen against a clipboard) "This bloodwork should've been flagged two hours ago." (glancing at a chart) "Who signed off on this? Come on, people." CUT TO: The same voice now dissolves into raw grief, caught off guard by a patient's sudden death on the table. (a long pause, then a whisper) "We did everything... everything right." (voice cracks) "Why does that never feel like enough?" The performance across both moments should feel like a mask of professionalism cracking open into unguarded sorrow.

Direction only (dialogue removed — this is what prompt-sim compares against):
A 34-year-old woman with a tight, slightly clipped voice, delivering this high-quality studio voice recording with no background noise. The same voice is bristling with controlled impatience, barely tolerating a delay in her morning rounds. (tapping a pen against a clipboard) (glancing at a chart) CUT TO: The same voice now dissolves into raw grief, caught off guard by a patient's sudden death on the table. (a long pause, then a whisper) (voice cracks) The performance across both moments should feel like a mask of professionalism cracking open into unguarded sorrow.

What to say (goes in MOSS text field): This bloodwork should've been flagged two hours ago. Who signed off on this? Come on, people. We did everything... everything right. Why does that never feel like enough?

with_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.443invWER0.897WER0.103prompt-sim+0.040blend6.018genu1.282
original: invWER × (blend + genu)
orig0.832invWER0.897WER0.103prompt-sim+0.040blend6.018genu1.282
adherence: prompt-sim × invWER
ps0.103invWER0.897WER0.103prompt-sim+0.115blend0.709genu0.467
with_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.126invWER0.897WER0.103prompt-sim+0.010blend3.069genu1.142
original: invWER × (blend + genu)
orig0.772invWER0.897WER0.103prompt-sim-0.320blend0.941genu4.171
adherence: prompt-sim × invWER
ps0.011invWER0.172WER0.828prompt-sim+0.061blend2.169genu0.525
no_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all0.984invWER0.586WER0.414prompt-sim+0.008blend6.469genu1.646
original: invWER × (blend + genu)
orig0.604invWER0.586WER0.414prompt-sim+0.008blend6.469genu1.646
adherence: prompt-sim × invWER
ps0.063invWER0.655WER0.345prompt-sim+0.096blend1.308genu0.606
no_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.129invWER0.897WER0.103prompt-sim-0.360blend2.521genu3.982
original: invWER × (blend + genu)
orig0.890invWER0.897WER0.103prompt-sim-0.360blend2.521genu3.982
adherence: prompt-sim × invWER
ps-0.000invWER0.000WER1.000prompt-sim-0.092blend0.715genu0.602
CCcc2c_archetype_5709_cc cc2c_archetype
SAY: It's... still glowing. Even after all this time. I could kneel here forever, you know. But it wouldn't bring her back. Oh, you thought I was done? Darling. I was just remembering what I'm about to take from you.
original DramaBox prompt & direction text

Full DramaBox prompt (how to say — with the "quoted" dialogue):
A 30-year-old woman with a low, velvet-smooth purr that holds a hint of danger, delivering this high-quality studio voice recording with no background noise. The same voice is staring at something beautiful but lost, awe mixing with a deep sorrow that makes her whisper, exhaustion dragging at her words. (slow exhale, almost a sigh) "It's... still glowing. Even after all this time." (a tiny, broken laugh) "I could kneel here forever, you know." (voice cracks slightly) "But it wouldn't bring her back." CUT TO: The same voice now sharpens into a predator's grin, exhaustion gone, replaced by a cold, delighted thrill as she tightens her grip. (a sharp inhale, then a silky low murmur) "Oh, you thought I was done? Darling." (soft chuckle, dangerous) "I was just remembering what I'm about to take from you." The performance across both moments should feel like the same woman is unraveling her grief and then instantly weaponizing it, the transition so seamless it's terrifying.

Direction only (dialogue removed — this is what prompt-sim compares against):
A 30-year-old woman with a low, velvet-smooth purr that holds a hint of danger, delivering this high-quality studio voice recording with no background noise. The same voice is staring at something beautiful but lost, awe mixing with a deep sorrow that makes her whisper, exhaustion dragging at her words. (slow exhale, almost a sigh) (a tiny, broken laugh) (voice cracks slightly) CUT TO: The same voice now sharpens into a predator's grin, exhaustion gone, replaced by a cold, delighted thrill as she tightens her grip. (a sharp inhale, then a silky low murmur) (soft chuckle, dangerous) The performance across both moments should feel like the same woman is unraveling her grief and then instantly weaponizing it, the transition so seamless it's terrifying.

What to say (goes in MOSS text field): It's... still glowing. Even after all this time. I could kneel here forever, you know. But it wouldn't bring her back. Oh, you thought I was done? Darling. I was just remembering what I'm about to take from you.

with_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.808invWER0.952WER0.048prompt-sim+0.101blend6.927genu2.129
original: invWER × (blend + genu)
orig1.098invWER0.952WER0.048prompt-sim+0.101blend6.927genu2.129
adherence: prompt-sim × invWER
ps0.112invWER0.786WER0.214prompt-sim+0.142blend3.415genu1.953
with_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.527invWER0.929WER0.071prompt-sim+0.069blend5.048genu1.944
original: invWER × (blend + genu)
orig1.078invWER1.000WER0.000prompt-sim-0.371blend2.951genu4.248
adherence: prompt-sim × invWER
ps0.065invWER0.929WER0.071prompt-sim+0.069blend5.048genu1.944
no_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.513invWER1.000WER0.000prompt-sim+0.051blend3.748genu2.076
original: invWER × (blend + genu)
orig0.820invWER1.000WER0.000prompt-sim+0.051blend3.748genu2.076
adherence: prompt-sim × invWER
ps0.116invWER0.952WER0.048prompt-sim+0.122blend0.676genu0.963
no_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.383invWER1.000WER0.000prompt-sim-0.197blend2.232genu3.879
original: invWER × (blend + genu)
orig0.947invWER1.000WER0.000prompt-sim-0.197blend2.232genu3.879
adherence: prompt-sim × invWER
ps0.037invWER0.500WER0.500prompt-sim+0.073blend3.374genu1.572
CCaccc_acting_challenge_7796_cc accc_acting_challenge
SAY: I was seven. I sat by the window every day after school. I thought you'd just been delayed. If I forgive you… does that mean what you did was okay? Was I that difficult to love?
original DramaBox prompt & direction text

Full DramaBox prompt (how to say — with the "quoted" dialogue):
A 34-year-old woman with a voice that is clear and measured, carrying a faint tremor of old hurt, delivering this high-quality studio voice recording with no background noise. The same voice is tightly controlled, as if she has rehearsed these words for years, but the stillness is brittle. (quiet, deliberate) "I was seven. I sat by the window every day after school." (a slow, measured exhale) "I thought you'd just been delayed." CUT TO: The same voice now cracks open into raw, childlike devastation, the adult composure gone. (voice breaking, barely a whisper) "If I forgive you… does that mean what you did was okay?" (a long, shuddering pause) "Was I that difficult to love?" The performance across both moments should feel like a dam breaking—from a rigid facade of composure to the naked, unfiltered ache of a child who never stopped waiting.

Direction only (dialogue removed — this is what prompt-sim compares against):
A 34-year-old woman with a voice that is clear and measured, carrying a faint tremor of old hurt, delivering this high-quality studio voice recording with no background noise. The same voice is tightly controlled, as if she has rehearsed these words for years, but the stillness is brittle. (quiet, deliberate) (a slow, measured exhale) CUT TO: The same voice now cracks open into raw, childlike devastation, the adult composure gone. (voice breaking, barely a whisper) (a long, shuddering pause) The performance across both moments should feel like a dam breaking—from a rigid facade of composure to the naked, unfiltered ache of a child who never stopped waiting.

What to say (goes in MOSS text field): I was seven. I sat by the window every day after school. I thought you'd just been delayed. If I forgive you… does that mean what you did was okay? Was I that difficult to love?

with_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.498invWER1.000WER0.000prompt-sim-0.044blend5.285genu1.599
original: invWER × (blend + genu)
orig0.902invWER1.000WER0.000prompt-sim-0.044blend5.285genu1.599
adherence: prompt-sim × invWER
ps0.116invWER1.000WER0.000prompt-sim+0.116blend3.383genu1.240
with_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.258invWER1.000WER0.000prompt-sim+0.066blend1.591genu1.743
original: invWER × (blend + genu)
orig0.677invWER0.676WER0.324prompt-sim-0.018blend6.249genu1.607
adherence: prompt-sim × invWER
ps0.079invWER0.811WER0.189prompt-sim+0.097blend1.534genu1.122
no_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.688invWER1.000WER0.000prompt-sim+0.003blend6.074genu1.991
original: invWER × (blend + genu)
orig1.044invWER1.000WER0.000prompt-sim+0.003blend6.074genu1.991
adherence: prompt-sim × invWER
ps0.152invWER1.000WER0.000prompt-sim+0.152blend3.086genu1.421
no_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.593invWER1.000WER0.000prompt-sim+0.100blend4.862genu1.528
original: invWER × (blend + genu)
orig0.848invWER1.000WER0.000prompt-sim+0.100blend4.862genu1.528
adherence: prompt-sim × invWER
ps0.100invWER1.000WER0.000prompt-sim+0.100blend4.862genu1.528
CCaccc_acting_challenge_1558_cc accc_acting_challenge
SAY: I thought I'd feel... more. More tears, more ache. Just dust motes in the sun. God, the quiet isn't heavy. It's just... air. I can walk loud now. No one to hear.
original DramaBox prompt & direction text

Full DramaBox prompt (how to say — with the "quoted" dialogue):
A 38-year-old woman with a warm but weary timbre, delivering this high-quality studio voice recording with no background noise. The same voice is struggling to hold onto a practiced sadness, but the familiar silence is making her unexpectedly restless. (letting out a slow breath) "I thought I'd feel... more. More tears, more ache." (pausing, listening to the creak) "Just dust motes in the sun." CUT TO: The same voice now lets out a surprised, private laugh as the emptiness finally feels like permission. (shaking her head, almost amused) "God, the quiet isn't heavy. It's just... air." (hearing her own footsteps) "I can walk loud now. No one to hear." The performance across both moments should feel like a single person slowly untangling grief from relief, each breath a small discovery.

Direction only (dialogue removed — this is what prompt-sim compares against):
A 38-year-old woman with a warm but weary timbre, delivering this high-quality studio voice recording with no background noise. The same voice is struggling to hold onto a practiced sadness, but the familiar silence is making her unexpectedly restless. (letting out a slow breath) (pausing, listening to the creak) CUT TO: The same voice now lets out a surprised, private laugh as the emptiness finally feels like permission. (shaking her head, almost amused) (hearing her own footsteps) The performance across both moments should feel like a single person slowly untangling grief from relief, each breath a small discovery.

What to say (goes in MOSS text field): I thought I'd feel... more. More tears, more ache. Just dust motes in the sun. God, the quiet isn't heavy. It's just... air. I can walk loud now. No one to hear.

with_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.462invWER0.857WER0.143prompt-sim+0.085blend5.453genu1.968
original: invWER × (blend + genu)
orig0.837invWER0.857WER0.143prompt-sim+0.085blend5.453genu1.968
adherence: prompt-sim × invWER
ps0.121invWER0.514WER0.486prompt-sim+0.234blend7.845genu1.936
with_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.324invWER0.971WER0.029prompt-sim-0.141blend4.298genu2.030
original: invWER × (blend + genu)
orig0.891invWER0.943WER0.057prompt-sim-0.362blend2.086genu3.961
adherence: prompt-sim × invWER
ps0.065invWER0.543WER0.457prompt-sim+0.120blend4.000genu1.283
no_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.426invWER0.971WER0.029prompt-sim-0.026blend4.852genu1.576
original: invWER × (blend + genu)
orig0.830invWER0.971WER0.029prompt-sim-0.026blend4.852genu1.576
adherence: prompt-sim × invWER
ps0.099invWER0.714WER0.286prompt-sim+0.139blend7.120genu1.689
no_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.107invWER0.971WER0.029prompt-sim-0.340blend1.149genu3.978
original: invWER × (blend + genu)
orig0.828invWER0.971WER0.029prompt-sim-0.340blend1.149genu3.978
adherence: prompt-sim × invWER
ps0.018invWER0.686WER0.314prompt-sim+0.027blend3.409genu1.526
CCaccc_acting_challenge_8508_cc accc_acting_challenge
SAY: The edges are almost translucent, a pink so pale it's barely there. It smells like rain and something sweet. I can feel every vein, like tiny rivers. I almost died for this? For this stupid, beautiful flower? And I'm supposed to be grateful
original DramaBox prompt & direction text

Full DramaBox prompt (how to say — with the "quoted" dialogue):
A 35-year-old woman with a warm, slightly raspy voice, delivering this high-quality studio voice recording with no background noise. The same voice is delicately exploring a flower with childlike wonder and fragile relief. (gently turning the flower in her palm) "The edges are almost translucent, a pink so pale it's barely there." (inhaling softly) "It smells like rain and something sweet." (brushing a petal) "I can feel every vein, like tiny rivers." CUT TO: The same voice now choked with rage and bitter grief, the wonder turned sour. (voice trembling, jaw tight) "I almost died for this? For this stupid, beautiful flower?" (a raw laugh) "And I'm supposed to be grateful?" (tears in voice) "I lost a year of my life to get here." The performance across both moments should feel like a single, fractured soul remembering how to feel everything at once.

Direction only (dialogue removed — this is what prompt-sim compares against):
A 35-year-old woman with a warm, slightly raspy voice, delivering this high-quality studio voice recording with no background noise. The same voice is delicately exploring a flower with childlike wonder and fragile relief. (gently turning the flower in her palm) (inhaling softly) (brushing a petal) CUT TO: The same voice now choked with rage and bitter grief, the wonder turned sour. (voice trembling, jaw tight) (a raw laugh) (tears in voice) The performance across both moments should feel like a single, fractured soul remembering how to feel everything at once.

What to say (goes in MOSS text field): The edges are almost translucent, a pink so pale it's barely there. It smells like rain and something sweet. I can feel every vein, like tiny rivers. I almost died for this? For this stupid, beautiful flower? And I'm supposed to be grateful? I lost a year of my life to get here.

with_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.306invWER0.982WER0.018prompt-sim+0.059blend2.657genu1.563
original: invWER × (blend + genu)
orig0.642invWER0.600WER0.400prompt-sim-0.007blend7.075genu1.502
adherence: prompt-sim × invWER
ps0.094invWER1.000WER0.000prompt-sim+0.094blend0.690genu1.768
with_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all0.988invWER0.982WER0.018prompt-sim-0.239blend0.063genu3.165
original: invWER × (blend + genu)
orig0.603invWER0.982WER0.018prompt-sim-0.239blend0.063genu3.165
adherence: prompt-sim × invWER
ps0.060invWER0.509WER0.491prompt-sim+0.118blend1.752genu1.130
no_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.282invWER0.927WER0.073prompt-sim+0.043blend4.055genu1.093
original: invWER × (blend + genu)
orig0.713invWER0.691WER0.309prompt-sim+0.005blend6.586genu1.577
adherence: prompt-sim × invWER
ps0.103invWER0.946WER0.054prompt-sim+0.109blend0.078genu0.951
no_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all0.875invWER0.691WER0.309prompt-sim-0.105blend3.799genu1.494
original: invWER × (blend + genu)
orig0.570invWER0.782WER0.218prompt-sim-0.402blend0.000genu3.941
adherence: prompt-sim × invWER
ps0.001invWER0.018WER0.982prompt-sim+0.032blend4.019genu0.593
CCsit_situation_287_cc sit_situation
SAY: ¡Métete con alguien de tu tamaño, pedazo de mierda! Te voy a partir la madre, escúchame bien. Te... te lo juro. Si le tocas un pelo más... te mato.
original DramaBox prompt & direction text

Full DramaBox prompt (how to say — with the "quoted" dialogue):
A 62-year-old man with a gravelly, weathered baritone, delivering this high-quality studio voice recording with no background noise. The same voice is hunched forward, fists up, spit flying as he shoves past a trash can toward a younger attacker, contempt raw and animal in his throat. (spitting between words) "¡Métete con alguien de tu tamaño, pedazo de mierda!" (swinging, breath hissing) "Te voy a partir la madre, escúchame bien." CUT TO: The same voice now crumpled on one knee, ribs on fire, jaw slack with pain as he tastes blood, exhaustion cracking his snarl into something colder, quieter. (coughing, grabbing his side) "Te... te lo juro." (voice dropping, venom trembling) "Si le tocas un pelo más... te mato." The performance across both moments should feel like a single brutal breath—venomous fury collapsing into a wounded, whispered threat, every grunt and pause earned by real damage.

Direction only (dialogue removed — this is what prompt-sim compares against):
A 62-year-old man with a gravelly, weathered baritone, delivering this high-quality studio voice recording with no background noise. The same voice is hunched forward, fists up, spit flying as he shoves past a trash can toward a younger attacker, contempt raw and animal in his throat. (spitting between words) (swinging, breath hissing) CUT TO: The same voice now crumpled on one knee, ribs on fire, jaw slack with pain as he tastes blood, exhaustion cracking his snarl into something colder, quieter. (coughing, grabbing his side) (voice dropping, venom trembling) The performance across both moments should feel like a single brutal breath—venomous fury collapsing into a wounded, whispered threat, every grunt and pause earned by real damage.

What to say (goes in MOSS text field): ¡Métete con alguien de tu tamaño, pedazo de mierda! Te voy a partir la madre, escúchame bien. Te... te lo juro. Si le tocas un pelo más... te mato.

with_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.310invWER0.931WER0.069prompt-sim-0.002blend4.525genu1.239
original: invWER × (blend + genu)
orig0.842invWER0.931WER0.069prompt-sim-0.169blend5.297genu1.603
adherence: prompt-sim × invWER
ps0.035invWER0.897WER0.103prompt-sim+0.039blend2.956genu1.312
with_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.416invWER0.966WER0.035prompt-sim-0.109blend5.544genu1.665
original: invWER × (blend + genu)
orig1.007invWER0.931WER0.069prompt-sim-0.219blend6.769genu1.776
adherence: prompt-sim × invWER
ps0.009invWER0.276WER0.724prompt-sim+0.033blend2.555genu1.446
no_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.759invWER0.931WER0.069prompt-sim+0.193blend7.659genu0.989
original: invWER × (blend + genu)
orig1.098invWER1.000WER0.000prompt-sim-0.168blend6.583genu2.007
adherence: prompt-sim × invWER
ps0.179invWER0.931WER0.069prompt-sim+0.193blend7.659genu0.989
no_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.212invWER0.931WER0.069prompt-sim-0.234blend5.694genu1.348
original: invWER × (blend + genu)
orig0.842invWER0.931WER0.069prompt-sim-0.234blend5.694genu1.348
adherence: prompt-sim × invWER
ps0.006invWER0.345WER0.655prompt-sim+0.017blend2.958genu1.696
CCsit_situation_5459_cc sit_situation
SAY: No te muevas, tonto. Casi, casi está saliendo. ¡Ay, ya! ¿Por qué no te desatas? Me duele la cabeza y esto no se mueve nada.
original DramaBox prompt & direction text

Full DramaBox prompt (how to say — with the "quoted" dialogue):
A child with a breathy, strained voice, delivering this high-quality studio voice recording with no background noise. The same voice is flushed and dizzy from being bent over, focused entirely on a stubborn knot in their shoelace. (breathing hard, words squeezed out) "No te muevas, tonto." (fingers fumbling) "Casi, casi está saliendo." CUT TO: The same voice now hot with frustration, a whine creeping in as the knot tightens and their arms start to shake. (a sharp exhale) "¡Ay, ya! ¿Por qué no te desatas?" (voice cracking with complaint) "Me duele la cabeza y esto no se mueve nada." The performance across both moments should feel like a real child's spiraling mood, from concentrated muttering to helpless whining, all while physically stuck in an uncomfortable pose.

Direction only (dialogue removed — this is what prompt-sim compares against):
A child with a breathy, strained voice, delivering this high-quality studio voice recording with no background noise. The same voice is flushed and dizzy from being bent over, focused entirely on a stubborn knot in their shoelace. (breathing hard, words squeezed out) (fingers fumbling) CUT TO: The same voice now hot with frustration, a whine creeping in as the knot tightens and their arms start to shake. (a sharp exhale) (voice cracking with complaint) The performance across both moments should feel like a real child's spiraling mood, from concentrated muttering to helpless whining, all while physically stuck in an uncomfortable pose.

What to say (goes in MOSS text field): No te muevas, tonto. Casi, casi está saliendo. ¡Ay, ya! ¿Por qué no te desatas? Me duele la cabeza y esto no se mueve nada.

with_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.536invWER0.800WER0.200prompt-sim+0.061blend8.592genu1.451
original: invWER × (blend + genu)
orig0.973invWER0.800WER0.200prompt-sim+0.061blend8.592genu1.451
adherence: prompt-sim × invWER
ps0.049invWER0.800WER0.200prompt-sim+0.061blend8.592genu1.451
with_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.345invWER0.920WER0.080prompt-sim-0.086blend4.305genu2.293
original: invWER × (blend + genu)
orig0.846invWER0.880WER0.120prompt-sim-0.185blend4.144genu2.724
adherence: prompt-sim × invWER
ps0.046invWER0.640WER0.360prompt-sim+0.072blend5.412genu2.081
no_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.642invWER0.960WER0.040prompt-sim-0.104blend7.248genu2.086
original: invWER × (blend + genu)
orig1.131invWER0.960WER0.040prompt-sim-0.104blend7.248genu2.086
adherence: prompt-sim × invWER
ps-0.000invWER0.000WER1.000prompt-sim-0.332blend0.348genu1.857
no_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.069invWER1.000WER0.000prompt-sim-0.175blend2.188genu1.753
original: invWER × (blend + genu)
orig0.747invWER0.960WER0.040prompt-sim-0.384blend2.671genu2.513
adherence: prompt-sim × invWER
ps-0.000invWER0.000WER1.000prompt-sim-0.200blend0.794genu1.741
CCsit_situation_5621_cc sit_situation
SAY: You useless, thoughtless boy. Left this damn thing right in the middle. Now my back. Has to pay for your laziness. But maybe... maybe I can do this. Just one more shove. I'm not as weak as he thinks.
original DramaBox prompt & direction text

Full DramaBox prompt (how to say — with the "quoted" dialogue):
A 70 year old woman with a gravelly, breathy timbre from years of smoking and strain, delivering this high-quality studio voice recording with no background noise. The same voice is furious and physically straining, her body braced against the weight of the oak bookcase, every word forced out through clenched teeth. (heaving a grunt) "You useless, thoughtless boy. Left this damn thing right in the middle. Now my back." (panting, shifting her grip) "Has to pay for your laziness." CUT TO: The same voice now quiet and contemplative, still panting but her anger faded into a fragile, tender hope as she catches her breath. (soft exhale, wiping her forehead) "But maybe... maybe I can do this. Just one more shove." (a small, cracked laugh) "I'm not as weak as he thinks." The performance across both moments should feel like a real, raw struggle—anger dissolving into exhausted hope, with the physical strain grounding every word.

Direction only (dialogue removed — this is what prompt-sim compares against):
A 70 year old woman with a gravelly, breathy timbre from years of smoking and strain, delivering this high-quality studio voice recording with no background noise. The same voice is furious and physically straining, her body braced against the weight of the oak bookcase, every word forced out through clenched teeth. (heaving a grunt) (panting, shifting her grip) CUT TO: The same voice now quiet and contemplative, still panting but her anger faded into a fragile, tender hope as she catches her breath. (soft exhale, wiping her forehead) (a small, cracked laugh) The performance across both moments should feel like a real, raw struggle—anger dissolving into exhausted hope, with the physical strain grounding every word.

What to say (goes in MOSS text field): You useless, thoughtless boy. Left this damn thing right in the middle. Now my back. Has to pay for your laziness. But maybe... maybe I can do this. Just one more shove. I'm not as weak as he thinks.

with_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.511invWER1.000WER0.000prompt-sim+0.100blend4.212genu1.434
original: invWER × (blend + genu)
orig0.904invWER1.000WER0.000prompt-sim-0.038blend5.144genu1.699
adherence: prompt-sim × invWER
ps0.100invWER1.000WER0.000prompt-sim+0.100blend4.212genu1.434
with_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.269invWER1.000WER0.000prompt-sim-0.135blend0.302genu4.001
original: invWER × (blend + genu)
orig0.946invWER1.000WER0.000prompt-sim-0.431blend1.770genu4.177
adherence: prompt-sim × invWER
ps0.009invWER0.225WER0.775prompt-sim+0.042blend1.410genu0.809
no_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.638invWER0.925WER0.075prompt-sim+0.135blend5.921genu1.748
original: invWER × (blend + genu)
orig0.916invWER0.925WER0.075prompt-sim+0.135blend5.921genu1.748
adherence: prompt-sim × invWER
ps0.125invWER0.925WER0.075prompt-sim+0.135blend5.921genu1.748
no_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.357invWER1.000WER0.000prompt-sim+0.049blend3.627genu1.164
original: invWER × (blend + genu)
orig0.875invWER1.000WER0.000prompt-sim-0.235blend1.323genu4.012
adherence: prompt-sim × invWER
ps0.049invWER1.000WER0.000prompt-sim+0.049blend3.627genu1.164
CCextreme_physical_196_cc extreme_physical
SAY: ...warm light on my eyelids... I can almost feel the sun... this is my skin, this is my breath... I am still here... No... NONONO— —it's in my BONES...
original DramaBox prompt & direction text

Full DramaBox prompt (how to say — with the "quoted" dialogue):
A 30-year-old woman with a warm, slightly husky alto, delivering this high-quality studio voice recording with no background noise. The same voice is lying on cold concrete, eyes closed, feeling a single bare bulb's warmth on her face, savoring stillness before the door opens. (soft whisper) "...warm light on my eyelids... I can almost feel the sun..." (barely audible, reverent) "this is my skin, this is my breath... I am still here..." CUT TO: The same voice now feels a metallic clamp on each ankle, then electricity seizing her body—the hum becomes a roar of fire, shattering her voice. (jaw locked, teeth chattering) "No... NONONO—" (high-pitched, breathless shriek) "—it's in my BONES..." (last word dissolves into a sob that becomes a raw, rhythmic grunt) The performance across both moments should feel like a woman being torn from a sacred, silent prayer into a brutal, all-consuming current.

Direction only (dialogue removed — this is what prompt-sim compares against):
A 30-year-old woman with a warm, slightly husky alto, delivering this high-quality studio voice recording with no background noise. The same voice is lying on cold concrete, eyes closed, feeling a single bare bulb's warmth on her face, savoring stillness before the door opens. (soft whisper) (barely audible, reverent) CUT TO: The same voice now feels a metallic clamp on each ankle, then electricity seizing her body—the hum becomes a roar of fire, shattering her voice. (jaw locked, teeth chattering) (high-pitched, breathless shriek) (last word dissolves into a sob that becomes a raw, rhythmic grunt) The performance across both moments should feel like a woman being torn from a sacred, silent prayer into a brutal, all-consuming current.

What to say (goes in MOSS text field): ...warm light on my eyelids... I can almost feel the sun... this is my skin, this is my breath... I am still here... No... NONONO— —it's in my BONES...

with_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.236invWER0.767WER0.233prompt-sim-0.102blend7.030genu1.581
original: invWER × (blend + genu)
orig0.826invWER0.767WER0.233prompt-sim-0.102blend7.030genu1.581
adherence: prompt-sim × invWER
ps0.014invWER0.733WER0.267prompt-sim+0.019blend6.925genu1.094
with_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.140invWER0.767WER0.233prompt-sim+0.022blend5.331genu1.063
original: invWER × (blend + genu)
orig0.701invWER0.733WER0.267prompt-sim-0.141blend5.803genu1.606
adherence: prompt-sim × invWER
ps0.018invWER0.167WER0.833prompt-sim+0.108blend2.965genu1.417
no_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.381invWER0.800WER0.200prompt-sim-0.091blend7.683genu1.807
original: invWER × (blend + genu)
orig0.943invWER0.800WER0.200prompt-sim-0.091blend7.683genu1.807
adherence: prompt-sim × invWER
ps0.014invWER0.100WER0.900prompt-sim+0.145blend5.245genu0.900
no_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.185invWER0.767WER0.233prompt-sim-0.031blend6.126genu1.274
original: invWER × (blend + genu)
orig0.719invWER0.767WER0.233prompt-sim-0.031blend6.126genu1.274
adherence: prompt-sim × invWER
ps0.000invWER0.033WER0.967prompt-sim+0.002blend5.597genu3.427
CCextreme_physical_270_cc extreme_physical
SAY: This lock's been dead for years... the hinges are the only things that remember how to sing. Oh, you think that's funny, do you? Just... just a little janitor?
original DramaBox prompt & direction text

Full DramaBox prompt (how to say — with the "quoted" dialogue):
A 30-year-old woman with a husky, sand-warmed contralto, delivering this high-quality studio voice recording with no background noise. The same voice is soft with melancholy, her palm pressed to cold rippled glass as she hums a slow, forgotten lullaby. (whispering near the glass) "This lock's been dead for years... the hinges are the only things that remember how to sing." (letting the thought dissolve into a breathy sigh) CUT TO: The same voice now yanked into her throat by a sudden face on the glass, her breath stolen, then ripped out as a raw, high-velocity scream that splinters into a trembling, bitter laugh. (stumbling backward, voice cracking) "Oh, you think that's funny, do you? Just... just a little janitor?" (trailing into a jagged sob that she swallows hard) The performance across both moments should feel like a solitary woman's intimate stillness shattered into a feral, wounded animal response.

Direction only (dialogue removed — this is what prompt-sim compares against):
A 30-year-old woman with a husky, sand-warmed contralto, delivering this high-quality studio voice recording with no background noise. The same voice is soft with melancholy, her palm pressed to cold rippled glass as she hums a slow, forgotten lullaby. (whispering near the glass) (letting the thought dissolve into a breathy sigh) CUT TO: The same voice now yanked into her throat by a sudden face on the glass, her breath stolen, then ripped out as a raw, high-velocity scream that splinters into a trembling, bitter laugh. (stumbling backward, voice cracking) (trailing into a jagged sob that she swallows hard) The performance across both moments should feel like a solitary woman's intimate stillness shattered into a feral, wounded animal response.

What to say (goes in MOSS text field): This lock's been dead for years... the hinges are the only things that remember how to sing. Oh, you think that's funny, do you? Just... just a little janitor?

with_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.356invWER0.806WER0.194prompt-sim-0.020blend6.651genu1.728
original: invWER × (blend + genu)
orig0.856invWER0.806WER0.194prompt-sim-0.020blend6.651genu1.728
adherence: prompt-sim × invWER
ps0.031invWER0.419WER0.581prompt-sim+0.073blend7.669genu1.907
with_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.146invWER0.968WER0.032prompt-sim-0.029blend2.207genu1.504
original: invWER × (blend + genu)
orig0.583invWER1.000WER0.000prompt-sim-0.102blend2.424genu1.424
adherence: prompt-sim × invWER
ps0.025invWER0.387WER0.613prompt-sim+0.066blend7.544genu1.467
no_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.480invWER1.000WER0.000prompt-sim+0.001blend4.379genu1.776
original: invWER × (blend + genu)
orig0.850invWER0.839WER0.161prompt-sim-0.094blend6.388genu1.593
adherence: prompt-sim × invWER
ps0.076invWER0.839WER0.161prompt-sim+0.091blend5.432genu1.540
no_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.437invWER1.000WER0.000prompt-sim+0.125blend3.429genu1.301
original: invWER × (blend + genu)
orig0.674invWER1.000WER0.000prompt-sim-0.153blend3.404genu1.363
adherence: prompt-sim × invWER
ps0.125invWER1.000WER0.000prompt-sim+0.125blend3.429genu1.301
CCextreme_physical_556_cc extreme_physical
SAY: Hey... if this wasn't happening, I'd kiss you right now. Right beneath that crack in the wood. No—you don't—get in— —you son of a—
original DramaBox prompt & direction text

Full DramaBox prompt (how to say — with the "quoted" dialogue):
A 20-year-old woman with a warm, honeyed contralto, delivering this high-quality studio voice recording with no background noise. The same voice is pressed against a steel door, heart humming from the boy's nearness, breath shallow and dreamy. (a soft exhale, turning her head) "Hey... if this wasn't happening, I'd kiss you right now." (a slow, lazy smile in her voice) "Right beneath that crack in the wood." CUT TO: The same voice now slammed against the buckled door, wind ripping the air away, muscles screaming as she shoves with her whole spine. (a snarling gasp as she throws her weight) "No—you don't—get in—" (a grinding, hating groan as she holds the seal) "—you son of a—" The performance across both moments should feel like one person ripped from a whispered promise into a snarling, breath-splintered fight for survival.

Direction only (dialogue removed — this is what prompt-sim compares against):
A 20-year-old woman with a warm, honeyed contralto, delivering this high-quality studio voice recording with no background noise. The same voice is pressed against a steel door, heart humming from the boy's nearness, breath shallow and dreamy. (a soft exhale, turning her head) (a slow, lazy smile in her voice) CUT TO: The same voice now slammed against the buckled door, wind ripping the air away, muscles screaming as she shoves with her whole spine. (a snarling gasp as she throws her weight) (a grinding, hating groan as she holds the seal) The performance across both moments should feel like one person ripped from a whispered promise into a snarling, breath-splintered fight for survival.

What to say (goes in MOSS text field): Hey... if this wasn't happening, I'd kiss you right now. Right beneath that crack in the wood. No—you don't—get in— —you son of a—

with_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.494invWER1.000WER0.000prompt-sim-0.064blend5.343genu1.672
original: invWER × (blend + genu)
orig0.928invWER0.828WER0.172prompt-sim-0.032blend7.036genu1.859
adherence: prompt-sim × invWER
ps0.039invWER0.655WER0.345prompt-sim+0.059blend4.760genu1.423
with_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.326invWER0.793WER0.207prompt-sim-0.049blend7.013genu1.618
original: invWER × (blend + genu)
orig0.937invWER0.966WER0.035prompt-sim-0.402blend2.766genu3.683
adherence: prompt-sim × invWER
ps0.059invWER0.655WER0.345prompt-sim+0.090blend1.501genu1.390
no_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.891invWER1.000WER0.000prompt-sim-0.121blend8.964genu2.228
original: invWER × (blend + genu)
orig1.376invWER1.000WER0.000prompt-sim-0.121blend8.964genu2.228
adherence: prompt-sim × invWER
ps0.039invWER0.931WER0.069prompt-sim+0.042blend3.629genu1.426
no_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.141invWER0.862WER0.138prompt-sim-0.018blend3.671genu1.371
original: invWER × (blend + genu)
orig0.878invWER1.000WER0.000prompt-sim-0.586blend2.840genu3.039
adherence: prompt-sim × invWER
ps0.007invWER0.655WER0.345prompt-sim+0.011blend6.553genu1.377
NONCCcca_voicenet_10203_noncc cca_voicenet
SAY: You know, she got the promotion just by being there. I've been here twice as long, but who counts that anymore.
original DramaBox prompt & direction text

Full DramaBox prompt (how to say — with the "quoted" dialogue):
A late-middle-aged female with a grainy, chest-rich alto, delivering this high-quality studio voice recording with no background noise. The same voice is simmering with sour jealousy, each word laced with tight-lipped envy. (Bitterly.) "You know, she got the promotion just by being there. I've been here twice as long, but who counts that anymore." (Voice tightens.) "It's fine.

Direction only (dialogue removed — this is what prompt-sim compares against):
A late-middle-aged female with a grainy, chest-rich alto, delivering this high-quality studio voice recording with no background noise. The same voice is simmering with sour jealousy, each word laced with tight-lipped envy. (Bitterly.) (Voice tightens.) "It's fine.

What to say (goes in MOSS text field): You know, she got the promotion just by being there. I've been here twice as long, but who counts that anymore.

with_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.655invWER1.000WER0.000prompt-sim-0.140blend7.677genu1.684
original: invWER × (blend + genu)
orig1.159invWER1.000WER0.000prompt-sim-0.140blend7.677genu1.684
adherence: prompt-sim × invWER
ps-0.000invWER0.000WER1.000prompt-sim-0.158blend5.514genu1.526
with_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.222invWER1.000WER0.000prompt-sim-0.256blend3.912genu2.146
original: invWER × (blend + genu)
orig0.848invWER1.000WER0.000prompt-sim-0.256blend3.912genu2.146
adherence: prompt-sim × invWER
ps-0.000invWER0.000WER1.182prompt-sim-0.222blend2.029genu1.652
no_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.599invWER1.000WER0.000prompt-sim-0.021blend6.392genu1.374
original: invWER × (blend + genu)
orig0.980invWER1.000WER0.000prompt-sim-0.021blend6.392genu1.374
adherence: prompt-sim × invWER
ps-0.000invWER0.000WER1.000prompt-sim-0.169blend6.549genu1.526
no_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.623invWER1.000WER0.000prompt-sim-0.153blend7.045genu1.980
original: invWER × (blend + genu)
orig1.141invWER1.000WER0.000prompt-sim-0.153blend7.045genu1.980
adherence: prompt-sim × invWER
ps-0.000invWER0.000WER1.000prompt-sim-0.113blend3.732genu1.379
NONCCcca_voicenet_18413_noncc cca_voicenet
SAY: You really didn't have to go and do all this.
original DramaBox prompt & direction text

Full DramaBox prompt (how to say — with the "quoted" dialogue):
A young male with a deep, cartoonishly authoritative voice, tinny and bright yet booming, delivering this high-quality studio voice recording with no background noise. The same voice is overwhelmed with intense gratitude, but the words sound annoyed. (Sighs heavily) "You really didn't have to go and do all this." (Shakes head) "It's just

Direction only (dialogue removed — this is what prompt-sim compares against):
A young male with a deep, cartoonishly authoritative voice, tinny and bright yet booming, delivering this high-quality studio voice recording with no background noise. The same voice is overwhelmed with intense gratitude, but the words sound annoyed. (Sighs heavily) (Shakes head) "It's just

What to say (goes in MOSS text field): You really didn't have to go and do all this.

with_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.768invWER1.000WER0.000prompt-sim-0.130blend8.150genu2.034
original: invWER × (blend + genu)
orig1.262invWER1.000WER0.000prompt-sim-0.130blend8.150genu2.034
adherence: prompt-sim × invWER
ps-0.000invWER0.000WER1.000prompt-sim-0.006blend1.639genu1.291
with_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.605invWER1.000WER0.000prompt-sim-0.003blend5.967genu1.572
original: invWER × (blend + genu)
orig1.076invWER1.000WER0.000prompt-sim-0.154blend5.276genu2.719
adherence: prompt-sim × invWER
ps0.049invWER1.000WER0.000prompt-sim+0.049blend1.803genu1.243
no_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.601invWER1.000WER0.000prompt-sim-0.104blend7.160genu1.444
original: invWER × (blend + genu)
orig1.069invWER1.000WER0.000prompt-sim-0.104blend7.160genu1.444
adherence: prompt-sim × invWER
ps-0.000invWER0.000WER1.546prompt-sim-0.137blend1.437genu1.161
no_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.645invWER0.818WER0.182prompt-sim-0.011blend7.986genu2.903
original: invWER × (blend + genu)
orig1.130invWER0.818WER0.182prompt-sim-0.011blend7.986genu2.903
adherence: prompt-sim × invWER
ps0.060invWER1.000WER0.000prompt-sim+0.060blend4.893genu1.993
NONCCextreme_physical_295_noncc extreme_physical
SAY: Ah, that breeze feels good.
original DramaBox prompt & direction text

Full DramaBox prompt (how to say — with the "quoted" dialogue):
A 40-year-old man with a gravelly, warm baritone, delivering this high-quality studio voice recording with no background noise. The same voice is relaxed and sensual, lips curled in a teasing smirk, tasting sour coffee and evening air through the open window. (soft exhale) "Ah, that breeze feels good." (amused drawl) "Yeah, you can

Direction only (dialogue removed — this is what prompt-sim compares against):
A 40-year-old man with a gravelly, warm baritone, delivering this high-quality studio voice recording with no background noise. The same voice is relaxed and sensual, lips curled in a teasing smirk, tasting sour coffee and evening air through the open window. (soft exhale) (amused drawl) "Yeah, you can

What to say (goes in MOSS text field): Ah, that breeze feels good.

with_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.805invWER1.000WER0.000prompt-sim+0.166blend6.667genu1.272
original: invWER × (blend + genu)
orig1.176invWER1.000WER0.000prompt-sim-0.181blend7.799genu1.712
adherence: prompt-sim × invWER
ps0.166invWER1.000WER0.000prompt-sim+0.166blend6.667genu1.272
with_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.671invWER1.000WER0.000prompt-sim+0.037blend6.185genu1.584
original: invWER × (blend + genu)
orig0.992invWER1.000WER0.000prompt-sim+0.037blend6.185genu1.584
adherence: prompt-sim × invWER
ps0.100invWER1.000WER0.000prompt-sim+0.100blend2.143genu1.715
no_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.726invWER1.000WER0.000prompt-sim+0.066blend6.559genu1.502
original: invWER × (blend + genu)
orig1.254invWER1.000WER0.000prompt-sim-0.175blend7.769genu2.228
adherence: prompt-sim × invWER
ps0.078invWER1.000WER0.000prompt-sim+0.078blend5.467genu0.979
no_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.943invWER1.000WER0.000prompt-sim+0.100blend5.923genu3.083
original: invWER × (blend + genu)
orig1.199invWER1.000WER0.000prompt-sim+0.100blend5.923genu3.083
adherence: prompt-sim × invWER
ps0.127invWER1.000WER0.000prompt-sim+0.127blend4.008genu1.384
NONCCsit_situation_3619_noncc sit_situation
SAY: ¡Por fin, carajo! Después de tanto tiempo peleando, el proyecto pasó. ¿Lo podés creer?
original DramaBox prompt & direction text

Full DramaBox prompt (how to say — with the "quoted" dialogue):
A 45-year-old man with a gravelly, whiskey-roughened voice, delivering this high-quality studio voice recording with no background noise. The same voice is elated and heavily intoxicated, swaying on a barstool as he triumphantly celebrates his long-awaited project approval. (stumbling, gripping the bar) "¡Por fin, carajo! Después de tanto tiempo peleando, el proyecto pasó. ¿Lo podés creer?" (laughing at his own feet) "Mirá, hasta las luces del

Direction only (dialogue removed — this is what prompt-sim compares against):
A 45-year-old man with a gravelly, whiskey-roughened voice, delivering this high-quality studio voice recording with no background noise. The same voice is elated and heavily intoxicated, swaying on a barstool as he triumphantly celebrates his long-awaited project approval. (stumbling, gripping the bar) (laughing at his own feet) "Mirá, hasta las luces del

What to say (goes in MOSS text field): ¡Por fin, carajo! Después de tanto tiempo peleando, el proyecto pasó. ¿Lo podés creer?

with_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.417invWER1.000WER0.000prompt-sim-0.192blend6.524genu1.260
original: invWER × (blend + genu)
orig0.976invWER1.000WER0.000prompt-sim-0.192blend6.524genu1.260
adherence: prompt-sim × invWER
ps-0.000invWER0.000WER1.286prompt-sim-0.298blend3.339genu1.720
with_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.575invWER1.000WER0.000prompt-sim-0.327blend8.159genu2.102
original: invWER × (blend + genu)
orig1.274invWER1.000WER0.000prompt-sim-0.327blend8.159genu2.102
adherence: prompt-sim × invWER
ps0.003invWER0.786WER0.214prompt-sim+0.003blend4.205genu2.567
no_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.592invWER1.000WER0.000prompt-sim-0.253blend8.113genu1.751
original: invWER × (blend + genu)
orig1.214invWER1.000WER0.000prompt-sim-0.253blend8.113genu1.751
adherence: prompt-sim × invWER
ps-0.000invWER0.000WER1.143prompt-sim-0.201blend5.175genu2.494
no_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.281invWER1.000WER0.000prompt-sim-0.119blend4.169genu1.442
original: invWER × (blend + genu)
orig0.820invWER1.000WER0.000prompt-sim-0.214blend3.505genu2.233
adherence: prompt-sim × invWER
ps0.012invWER0.071WER0.929prompt-sim+0.164blend1.958genu2.119
NONCCcca_voicenet_11461_noncc cca_voicenet
SAY: You— You don't get it. You never did. I'm drowning.
original DramaBox prompt & direction text

Full DramaBox prompt (how to say — with the "quoted" dialogue):
A 35-year-old male with a hyper-masculine, booming chest resonance and a fast, compressed delivery, yet marked by disfluent stutters and a flat, unstressed intonation, delivering this high-quality studio voice recording with no background noise. The same voice is vibrating with barely contained rage, jaw tight, words slamming together in a rapid, choked stream of anger and distress. (He presses his palm against the table, knuckles white, the low rumble of contempt saturating each syllable.) "You—" (He breaks off, swallows, starts again faster.) "You don't get it." (A hard, bitter exhale through his nose.) "You never did." (He shakes his head, voice cracking under the pressure, the flat emphasis making each word a dead weight.) "I'm drowning." (He stares at the far wall

Direction only (dialogue removed — this is what prompt-sim compares against):
A 35-year-old male with a hyper-masculine, booming chest resonance and a fast, compressed delivery, yet marked by disfluent stutters and a flat, unstressed intonation, delivering this high-quality studio voice recording with no background noise. The same voice is vibrating with barely contained rage, jaw tight, words slamming together in a rapid, choked stream of anger and distress. (He presses his palm against the table, knuckles white, the low rumble of contempt saturating each syllable.) (He breaks off, swallows, starts again faster.) (A hard, bitter exhale through his nose.) (He shakes his head, voice cracking under the pressure, the flat emphasis making each word a dead weight.) (He stares at the far wall

What to say (goes in MOSS text field): You— You don't get it. You never did. I'm drowning.

with_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.799invWER1.000WER0.000prompt-sim-0.056blend7.928genu1.889
original: invWER × (blend + genu)
orig1.221invWER1.000WER0.000prompt-sim-0.290blend7.354genu2.289
adherence: prompt-sim × invWER
ps0.097invWER0.917WER0.083prompt-sim+0.106blend5.413genu1.360
with_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.825invWER1.000WER0.000prompt-sim-0.141blend8.835genu2.028
original: invWER × (blend + genu)
orig1.391invWER1.000WER0.000prompt-sim-0.248blend8.900genu2.369
adherence: prompt-sim × invWER
ps0.001invWER0.083WER0.917prompt-sim+0.008blend3.930genu2.264
no_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.765invWER0.917WER0.083prompt-sim+0.003blend8.394genu2.002
original: invWER × (blend + genu)
orig1.261invWER0.917WER0.083prompt-sim-0.128blend8.878genu2.285
adherence: prompt-sim × invWER
ps0.142invWER0.750WER0.250prompt-sim+0.189blend1.866genu1.109
no_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.911invWER1.000WER0.000prompt-sim+0.034blend8.529genu1.613
original: invWER × (blend + genu)
orig1.296invWER1.000WER0.000prompt-sim-0.110blend8.702genu1.886
adherence: prompt-sim × invWER
ps0.034invWER1.000WER0.000prompt-sim+0.034blend8.529genu1.613
NONCCcca_voicenet_11643_noncc cca_voicenet
SAY: Well, that's one way to break the ice. I guess I should've seen that coming. But honestly, I'm just glad it's over.
original DramaBox prompt & direction text

Full DramaBox prompt (how to say — with the "quoted" dialogue):
A young man with a glacially slow, soft-onset voice that snaps into percussive emphasis on key words, delivering this high-quality studio voice recording with no background noise. The same voice is a strained blend of amusement and relief, despite a sharp, grinding pain that makes each syllable an effort. (He touches the wound with trembling fingers, a pained but amused exhale escaping.) "Well, that's one way to break the ice." (He shifts his weight, then lets out a short, ragged chuckle.) "I guess I should've seen that coming." (He pauses, breathing slowly.) "But honestly, I'm just glad it's over." (A soft, wheezing laugh.) "At least now I know

Direction only (dialogue removed — this is what prompt-sim compares against):
A young man with a glacially slow, soft-onset voice that snaps into percussive emphasis on key words, delivering this high-quality studio voice recording with no background noise. The same voice is a strained blend of amusement and relief, despite a sharp, grinding pain that makes each syllable an effort. (He touches the wound with trembling fingers, a pained but amused exhale escaping.) (He shifts his weight, then lets out a short, ragged chuckle.) (He pauses, breathing slowly.) (A soft, wheezing laugh.) "At least now I know

What to say (goes in MOSS text field): Well, that's one way to break the ice. I guess I should've seen that coming. But honestly, I'm just glad it's over.

with_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.648invWER0.962WER0.038prompt-sim+0.018blend6.514genu1.771
original: invWER × (blend + genu)
orig1.090invWER0.962WER0.038prompt-sim-0.146blend7.083genu1.904
adherence: prompt-sim × invWER
ps0.076invWER0.962WER0.038prompt-sim+0.079blend4.460genu1.721
with_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.859invWER0.962WER0.038prompt-sim-0.007blend3.308genu5.441
original: invWER × (blend + genu)
orig1.250invWER0.962WER0.038prompt-sim-0.007blend3.308genu5.441
adherence: prompt-sim × invWER
ps0.094invWER0.962WER0.038prompt-sim+0.098blend2.972genu4.494
no_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.733invWER0.962WER0.038prompt-sim+0.034blend7.051genu1.877
original: invWER × (blend + genu)
orig1.082invWER0.962WER0.038prompt-sim+0.034blend7.051genu1.877
adherence: prompt-sim × invWER
ps0.167invWER0.962WER0.038prompt-sim+0.174blend1.295genu1.153
no_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.691invWER0.962WER0.038prompt-sim+0.039blend1.981genu4.880
original: invWER × (blend + genu)
orig1.116invWER0.923WER0.077prompt-sim-0.073blend3.839genu4.512
adherence: prompt-sim × invWER
ps0.038invWER0.962WER0.038prompt-sim+0.039blend1.981genu4.880
NONCCcca_voicenet_9837_noncc cca_voicenet
SAY: It's just... the way the light hits those leaves in the afternoon. I always thought that was the most peaceful thing.
original DramaBox prompt & direction text

Full DramaBox prompt (how to say — with the "quoted" dialogue):
A middle-to-late-aged speaker with a grainy, androgynous timbre, delivering this high-quality studio voice recording with no background noise. The same voice is caught in a tight spiral of helpless rage, forcing a veneer of calm contemplation. (A slow, deliberate exhale, jaw tight) "It's just... the way the light hits those leaves in the afternoon." (A pause, swallowing) "I always thought that was the most peaceful thing." (Fingers drumming

Direction only (dialogue removed — this is what prompt-sim compares against):
A middle-to-late-aged speaker with a grainy, androgynous timbre, delivering this high-quality studio voice recording with no background noise. The same voice is caught in a tight spiral of helpless rage, forcing a veneer of calm contemplation. (A slow, deliberate exhale, jaw tight) (A pause, swallowing) (Fingers drumming

What to say (goes in MOSS text field): It's just... the way the light hits those leaves in the afternoon. I always thought that was the most peaceful thing.

with_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.778invWER1.000WER0.000prompt-sim+0.015blend7.217genu1.742
original: invWER × (blend + genu)
orig1.223invWER1.000WER0.000prompt-sim-0.197blend7.812genu2.004
adherence: prompt-sim × invWER
ps0.113invWER1.000WER0.000prompt-sim+0.113blend6.475genu1.529
with_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all2.015invWER1.000WER0.000prompt-sim+0.094blend8.636genu1.809
original: invWER × (blend + genu)
orig1.347invWER1.000WER0.000prompt-sim-0.077blend8.856genu2.117
adherence: prompt-sim × invWER
ps0.166invWER1.000WER0.000prompt-sim+0.166blend5.161genu1.063
no_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.756invWER1.000WER0.000prompt-sim+0.060blend6.500genu1.770
original: invWER × (blend + genu)
orig1.119invWER1.000WER0.000prompt-sim-0.065blend7.085genu1.814
adherence: prompt-sim × invWER
ps0.134invWER1.000WER0.000prompt-sim+0.134blend4.465genu0.900
no_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.973invWER1.000WER0.000prompt-sim+0.077blend8.194genu1.944
original: invWER × (blend + genu)
orig1.253invWER1.000WER0.000prompt-sim+0.077blend8.194genu1.944
adherence: prompt-sim × invWER
ps0.172invWER0.955WER0.045prompt-sim+0.180blend5.062genu1.029
NONCCcc2c_archetype_265_noncc cc2c_archetype
SAY: I told you I don't have time for this.
original DramaBox prompt & direction text

Full DramaBox prompt (how to say — with the "quoted" dialogue):
A 28-year-old woman with a husky, smoky voice, delivering this high-quality studio voice recording with no background noise. The same voice is brittle and fast, her words clipped with a low, agitated hum of numbness and irritation. (breathing out sharply) "I told you I don't have time for this." (tapping a nail) "Just say what you want and get it over with

Direction only (dialogue removed — this is what prompt-sim compares against):
A 28-year-old woman with a husky, smoky voice, delivering this high-quality studio voice recording with no background noise. The same voice is brittle and fast, her words clipped with a low, agitated hum of numbness and irritation. (breathing out sharply) (tapping a nail) "Just say what you want and get it over with

What to say (goes in MOSS text field): I told you I don't have time for this.

with_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.893invWER1.000WER0.000prompt-sim-0.008blend8.416genu1.850
original: invWER × (blend + genu)
orig1.316invWER1.000WER0.000prompt-sim-0.091blend7.927genu2.527
adherence: prompt-sim × invWER
ps0.081invWER1.000WER0.000prompt-sim+0.081blend5.493genu1.344
with_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.735invWER1.000WER0.000prompt-sim-0.007blend7.156genu1.653
original: invWER × (blend + genu)
orig1.102invWER1.000WER0.000prompt-sim-0.007blend7.156genu1.653
adherence: prompt-sim × invWER
ps-0.000invWER0.000WER1.000prompt-sim-0.106blend0.659genu0.651
no_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.791invWER1.000WER0.000prompt-sim-0.066blend8.450genu1.555
original: invWER × (blend + genu)
orig1.252invWER1.000WER0.000prompt-sim-0.243blend8.654genu1.638
adherence: prompt-sim × invWER
ps0.053invWER1.000WER0.000prompt-sim+0.053blend5.860genu1.883
no_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.547invWER1.000WER0.000prompt-sim-0.281blend7.141genu2.280
original: invWER × (blend + genu)
orig1.198invWER1.000WER0.000prompt-sim-0.281blend7.141genu2.280
adherence: prompt-sim × invWER
ps0.012invWER0.600WER0.400prompt-sim+0.020blend2.684genu2.249
NONCCsit_situation_2127_noncc sit_situation
SAY: Y entonces el tipo me dice, 'estás buenísima, ¿cómo es que estás sola?'
original DramaBox prompt & direction text

Full DramaBox prompt (how to say — with the "quoted" dialogue):
A 25-year-old woman with a bright but fraying vocal quality, delivering this high-quality studio voice recording with no background noise. The same voice is riding a second wind at a 3 AM after-party, forcing laughter as she recounts a crude comment, but disgust is already tightening her throat. (laughing too loudly) "Y entonces el tipo me dice, 'estás buenísima, ¿cómo es que estás sola?'" (voice cracking, swallowing) "

Direction only (dialogue removed — this is what prompt-sim compares against):
A 25-year-old woman with a bright but fraying vocal quality, delivering this high-quality studio voice recording with no background noise. The same voice is riding a second wind at a 3 AM after-party, forcing laughter as she recounts a crude comment, but disgust is already tightening her throat. (laughing too loudly) (voice cracking, swallowing) "

What to say (goes in MOSS text field): Y entonces el tipo me dice, 'estás buenísima, ¿cómo es que estás sola?'

with_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.574invWER1.000WER0.000prompt-sim-0.108blend5.965genu2.073
original: invWER × (blend + genu)
orig1.229invWER1.000WER0.000prompt-sim-0.307blend7.842genu2.019
adherence: prompt-sim × invWER
ps-0.000invWER0.000WER1.000prompt-sim-0.037blend1.147genu1.464
with_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.346invWER0.923WER0.077prompt-sim-0.172blend2.593genu3.961
original: invWER × (blend + genu)
orig0.920invWER0.923WER0.077prompt-sim-0.172blend2.593genu3.961
adherence: prompt-sim × invWER
ps-0.000invWER0.000WER1.000prompt-sim-0.145blend1.875genu0.884
no_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.644invWER1.000WER0.000prompt-sim-0.283blend8.391genu2.104
original: invWER × (blend + genu)
orig1.298invWER1.000WER0.000prompt-sim-0.283blend8.391genu2.104
adherence: prompt-sim × invWER
ps-0.000invWER0.000WER1.000prompt-sim-0.061blend2.522genu2.124
no_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.284invWER1.000WER0.000prompt-sim-0.309blend4.221genu2.692
original: invWER × (blend + genu)
orig1.083invWER0.923WER0.077prompt-sim-0.450blend7.614genu1.810
adherence: prompt-sim × invWER
ps-0.000invWER0.000WER1.000prompt-sim-0.229blend3.767genu4.577
NONCCcc2c_archetype_4265_noncc cc2c_archetype
SAY: I still scan every window. Every shadow.
original DramaBox prompt & direction text

Full DramaBox prompt (how to say — with the "quoted" dialogue):
A 30s male with a tightly controlled, gravelly tenor, delivering this high-quality studio voice recording with no background noise. The same voice is a watchful protector, sour with unresolved betrayal, prideful in his vigilance, and aching with longing for a lost partner. (breathing slow, jaw tight) "I still scan every window. Every shadow." (shifting weight, glancing) "She'd tell me to let it go. I can

Direction only (dialogue removed — this is what prompt-sim compares against):
A 30s male with a tightly controlled, gravelly tenor, delivering this high-quality studio voice recording with no background noise. The same voice is a watchful protector, sour with unresolved betrayal, prideful in his vigilance, and aching with longing for a lost partner. (breathing slow, jaw tight) (shifting weight, glancing) "She'd tell me to let it go. I can

What to say (goes in MOSS text field): I still scan every window. Every shadow.

with_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.561invWER1.000WER0.000prompt-sim-0.190blend7.365genu1.623
original: invWER × (blend + genu)
orig1.118invWER1.000WER0.000prompt-sim-0.190blend7.365genu1.623
adherence: prompt-sim × invWER
ps-0.000invWER0.000WER2.000prompt-sim-0.225blend3.628genu1.327
with_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.466invWER1.000WER0.000prompt-sim-0.207blend6.633genu1.603
original: invWER × (blend + genu)
orig1.040invWER1.000WER0.000prompt-sim-0.207blend6.633genu1.603
adherence: prompt-sim × invWER
ps-0.000invWER0.000WER1.143prompt-sim-0.194blend2.059genu0.247
no_dialogue · no_ref
all: invWER × (prompt-sim + blend + genu)
all1.777invWER1.000WER0.000prompt-sim-0.086blend7.550genu2.192
original: invWER × (blend + genu)
orig1.226invWER1.000WER0.000prompt-sim-0.086blend7.550genu2.192
adherence: prompt-sim × invWER
ps-0.000invWER0.000WER1.571prompt-sim-0.182blend3.406genu1.349
no_dialogue · emo_ref
all: invWER × (prompt-sim + blend + genu)
all1.672invWER1.000WER0.000prompt-sim-0.235blend4.390genu4.577
original: invWER × (blend + genu)
orig1.297invWER1.000WER0.000prompt-sim-0.295blend5.537genu3.965
adherence: prompt-sim × invWER
ps-0.000invWER0.000WER1.286prompt-sim-0.237blend0.309genu0.704