Two frozen models. Exact instructions, matched seeds and complete take distributions. Find a promising recipe quickly, then inspect its limitations and listen.
Emotion/style only. Content Enjoyment: 0%. Genuineness: 0%. These remain visible diagnostics, not reward terms. WER and speaker identity are separate qualification gates, not terms in this score.
Observed head means below refer to g0-p06 at CFG 3; every recipe also shows its own values below. These are recipe means, not one hand-picked audio.
Φ is a Gaussian CDF using the frozen EARS training statistics. This makes mixed emotion/style heads comparable; it is not an emotion probability. It compresses extreme scores, so raw means remain visible too. A pure-emotion condition has only its named emotion head, at 100%. A mixed anger/ranting scene uses 50% Anger + 50% Ranting. Each recipe score is the mean over all its takes, including failed takes.
Unchanged gates: ≥90% valid terminated takes; per-take WER ≤15% to count as valid; candidate mean WER ≤10%; paired reference cosine decline no worse than −0.05. Existing family-aware burst conditions still need all requested families detected in ≥60% of takes to qualify as reliable burst recipes. A high emotion/style score alone does not establish a successful scream or a correctly timed transition.
Re-scored, not regenerated: the original raw EARS predictions were reused on CPU. Old fitness, audio, prompts and scientific gates are retained. Old held-out sets tested only the previously selected wording; a newly re-ranked search winner does not inherit that validation. No universal best weighting or human-preference optimum is claimed.
Historical quality-fitness diagnostic — not the current steering ranking
Exact score for this condition
This is the prompt-search fitness, not the training RL reward. Higher means better under this particular proxy combination, not automatically more natural or more enjoyable to a listener.
50% target emotion/style, 25% content enjoyment, 25% genuineness. The target is vn_S_CONV_reg. VoiceCLAP and absolute speaker cosine are diagnostics here, not extra weighted rewards.
Fixed target normalization: z = (raw score − 2.533) / 1.305; target_CDF = Φ(z). This is a percentile-like training-distribution transform, not an emotion probability. CE and genuineness are clipped to [0, 1] after division. For burst tasks, the same target transform is used for the two arc endpoints; blend has its own frozen normalization.
WER example: WER 5% multiplies the score by 0.95; WER 20% by 0.80; WER ≥ 100% by zero. Reported raw WER remains uncapped. The displayed fitness is the mean of per-take products, not the product of the displayed means.
Separate qualification gates: at least 90% valid, terminated takes; per-take WER ≤ 15% to count as valid; candidate mean WER ≤ 10%; paired reference cosine must not fall by more than 0.05 over all 50 reference comparisons. Corrected burst conditions additionally need every requested family detected in at least 60% of takes. Passing a gate is not a weighted score bonus.
Start here: best observed recipe, not a universal optimum
The current starting point uses EARS Steering v2 across the available canonical search conditions, with unchanged qualification gates taking priority. Historical candidates are used only where no canonical search exists. Held-out results below refer to the earlier selected wording and are not automatically inherited by a newly re-ranked winner.
How the strategies compare
Click any column header to sort. Do not compare scalar fitness across different missions or between legacy and event-aware scoring contracts.
Baseline, leading recipes and runners-up retain every language/reference/seed combination. These are not hand-picked best takes. Players use individual MP3 URLs, no base64 downloads, and preload nothing. Initial view is paginated for fast loading.
English fixed reference ·12seconds
A clean fixed codec-reconstructed reference, a different utterance from every search scene. The same recording is used by both models.
German fixed reference ·12seconds
A clean fixed codec-reconstructed reference, a different utterance from every search scene. The same recording is used by both models.
Recipe details and exact prompt contracts
g0-p06 · CFG 3
full-v2-gen0CANONICAL
Current ranking: EARS Steering v2
Emotion/style only. Content Enjoyment: 0%. Genuineness: 0%. These remain visible diagnostics, not reward terms. WER and speaker identity are separate qualification gates, not terms in this score.
Observed head means below refer to g0-p06 at CFG 3; every recipe also shows its own values below. These are recipe means, not one hand-picked audio.
Φ is a Gaussian CDF using the frozen EARS training statistics. This makes mixed emotion/style heads comparable; it is not an emotion probability. It compresses extreme scores, so raw means remain visible too. A pure-emotion condition has only its named emotion head, at 100%. A mixed anger/ranting scene uses 50% Anger + 50% Ranting. Each recipe score is the mean over all its takes, including failed takes.
Unchanged gates: ≥90% valid terminated takes; per-take WER ≤15% to count as valid; candidate mean WER ≤10%; paired reference cosine decline no worse than −0.05. Existing family-aware burst conditions still need all requested families detected in ≥60% of takes to qualify as reliable burst recipes. A high emotion/style score alone does not establish a successful scream or a correctly timed transition.
Historical quality-fitness diagnostic — not the current steering ranking
Exact score for this condition
This is the prompt-search fitness, not the training RL reward. Higher means better under this particular proxy combination, not automatically more natural or more enjoyable to a listener.
50% target emotion/style, 25% content enjoyment, 25% genuineness. The target is vn_S_CONV_reg. VoiceCLAP and absolute speaker cosine are diagnostics here, not extra weighted rewards.
Fixed target normalization: z = (raw score − 2.533) / 1.305; target_CDF = Φ(z). This is a percentile-like training-distribution transform, not an emotion probability. CE and genuineness are clipped to [0, 1] after division. For burst tasks, the same target transform is used for the two arc endpoints; blend has its own frozen normalization.
WER example: WER 5% multiplies the score by 0.95; WER 20% by 0.80; WER ≥ 100% by zero. Reported raw WER remains uncapped. The displayed fitness is the mean of per-take products, not the product of the displayed means.
Separate qualification gates: at least 90% valid, terminated takes; per-take WER ≤ 15% to count as valid; candidate mean WER ≤ 10%; paired reference cosine must not fall by more than 0.05 over all 50 reference comparisons. Corrected burst conditions additionally need every requested family detected in at least 60% of takes. Passing a gate is not a weighted score bonus.
Passing study gates. N=100 across both languages, both reference arms and all five seeds. Full-coverage steering-v2 evolution; transfer, fixed-wording CFG and sealed splits remain distinct.
0.882
6.386
4.435
0.41%
0.782
100.00%
-0.008
Prompting strategy: containment contrast.
Reusable direction recipe
Stored GENERAL content
A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.
Three local directions
1. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.
2. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.
3. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.
Exact tested prompt — fixed EN example, seed1701, no reference
GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.
SCRIPT:
(A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [4.2 seconds duration] I thought you would be here before the lights went out.
(A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Wait... I need a moment to understand what just happened.
(A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.5 seconds duration] Stay with me until I can say this properly.
Literal words in the separate text field:
I thought you would be here before the lights went out. Wait... I need a moment to understand what just happened. Stay with me until I can say this properly.
Exact neutral CFG branch
GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.
SCRIPT:
(Natural clear speech.) [4.2 seconds duration] I thought you would be here before the lights went out.
(Natural clear speech.) [3.8 seconds duration] Wait... I need a moment to understand what just happened.
(Natural clear speech.) [3.5 seconds duration] Stay with me until I can say this properly.
Reference versus no-reference breakdown
reference
50
0.882
6.381
4.398
0.48%
0.589
none
50
0.882
6.391
4.473
0.33%
n/a
g0-p05 · CFG 2.5
full-v2-gen0CANONICAL
Current ranking: EARS Steering v2
Emotion/style only. Content Enjoyment: 0%. Genuineness: 0%. These remain visible diagnostics, not reward terms. WER and speaker identity are separate qualification gates, not terms in this score.
Observed head means below refer to g0-p05 at CFG 2.5; every recipe also shows its own values below. These are recipe means, not one hand-picked audio.
Φ is a Gaussian CDF using the frozen EARS training statistics. This makes mixed emotion/style heads comparable; it is not an emotion probability. It compresses extreme scores, so raw means remain visible too. A pure-emotion condition has only its named emotion head, at 100%. A mixed anger/ranting scene uses 50% Anger + 50% Ranting. Each recipe score is the mean over all its takes, including failed takes.
Unchanged gates: ≥90% valid terminated takes; per-take WER ≤15% to count as valid; candidate mean WER ≤10%; paired reference cosine decline no worse than −0.05. Existing family-aware burst conditions still need all requested families detected in ≥60% of takes to qualify as reliable burst recipes. A high emotion/style score alone does not establish a successful scream or a correctly timed transition.
Historical quality-fitness diagnostic — not the current steering ranking
Exact score for this condition
This is the prompt-search fitness, not the training RL reward. Higher means better under this particular proxy combination, not automatically more natural or more enjoyable to a listener.
50% target emotion/style, 25% content enjoyment, 25% genuineness. The target is vn_S_CONV_reg. VoiceCLAP and absolute speaker cosine are diagnostics here, not extra weighted rewards.
Fixed target normalization: z = (raw score − 2.533) / 1.305; target_CDF = Φ(z). This is a percentile-like training-distribution transform, not an emotion probability. CE and genuineness are clipped to [0, 1] after division. For burst tasks, the same target transform is used for the two arc endpoints; blend has its own frozen normalization.
WER example: WER 5% multiplies the score by 0.95; WER 20% by 0.80; WER ≥ 100% by zero. Reported raw WER remains uncapped. The displayed fitness is the mean of per-take products, not the product of the displayed means.
Separate qualification gates: at least 90% valid, terminated takes; per-take WER ≤ 15% to count as valid; candidate mean WER ≤ 10%; paired reference cosine must not fall by more than 0.05 over all 50 reference comparisons. Corrected burst conditions additionally need every requested family detected in at least 60% of takes. Passing a gate is not a weighted score bonus.
Passing study gates. N=100 across both languages, both reference arms and all five seeds. Full-coverage steering-v2 evolution; transfer, fixed-wording CFG and sealed splits remain distinct.
0.880
6.367
4.439
0.31%
0.782
100.00%
0.037
Prompting strategy: attack, sustain and release.
Reusable direction recipe
Stored GENERAL content
A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. Build an audible attack, sustain the intended quality, then release it without losing the final words.
Three local directions
1. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. Build an audible attack, sustain the intended quality, then release it without losing the final words.
2. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. Build an audible attack, sustain the intended quality, then release it without losing the final words.
3. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. Build an audible attack, sustain the intended quality, then release it without losing the final words.
Exact tested prompt — fixed EN example, seed1701, no reference
GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. Build an audible attack, sustain the intended quality, then release it without losing the final words.
SCRIPT:
(A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. Build an audible attack, sustain the intended quality, then release it without losing the final words.) [4.2 seconds duration] I thought you would be here before the lights went out.
(A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. Build an audible attack, sustain the intended quality, then release it without losing the final words.) [3.8 seconds duration] Wait... I need a moment to understand what just happened.
(A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. Build an audible attack, sustain the intended quality, then release it without losing the final words.) [3.5 seconds duration] Stay with me until I can say this properly.
Literal words in the separate text field:
I thought you would be here before the lights went out. Wait... I need a moment to understand what just happened. Stay with me until I can say this properly.
Exact neutral CFG branch
GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.
SCRIPT:
(Natural clear speech.) [4.2 seconds duration] I thought you would be here before the lights went out.
(Natural clear speech.) [3.8 seconds duration] Wait... I need a moment to understand what just happened.
(Natural clear speech.) [3.5 seconds duration] Stay with me until I can say this properly.
Reference versus no-reference breakdown
reference
50
0.875
6.354
4.370
0.13%
0.634
none
50
0.885
6.380
4.507
0.48%
n/a
g0-p09 · CFG 2.5
full-v2-gen0CANONICAL
Current ranking: EARS Steering v2
Emotion/style only. Content Enjoyment: 0%. Genuineness: 0%. These remain visible diagnostics, not reward terms. WER and speaker identity are separate qualification gates, not terms in this score.
Observed head means below refer to g0-p09 at CFG 2.5; every recipe also shows its own values below. These are recipe means, not one hand-picked audio.
Φ is a Gaussian CDF using the frozen EARS training statistics. This makes mixed emotion/style heads comparable; it is not an emotion probability. It compresses extreme scores, so raw means remain visible too. A pure-emotion condition has only its named emotion head, at 100%. A mixed anger/ranting scene uses 50% Anger + 50% Ranting. Each recipe score is the mean over all its takes, including failed takes.
Unchanged gates: ≥90% valid terminated takes; per-take WER ≤15% to count as valid; candidate mean WER ≤10%; paired reference cosine decline no worse than −0.05. Existing family-aware burst conditions still need all requested families detected in ≥60% of takes to qualify as reliable burst recipes. A high emotion/style score alone does not establish a successful scream or a correctly timed transition.
Historical quality-fitness diagnostic — not the current steering ranking
Exact score for this condition
This is the prompt-search fitness, not the training RL reward. Higher means better under this particular proxy combination, not automatically more natural or more enjoyable to a listener.
50% target emotion/style, 25% content enjoyment, 25% genuineness. The target is vn_S_CONV_reg. VoiceCLAP and absolute speaker cosine are diagnostics here, not extra weighted rewards.
Fixed target normalization: z = (raw score − 2.533) / 1.305; target_CDF = Φ(z). This is a percentile-like training-distribution transform, not an emotion probability. CE and genuineness are clipped to [0, 1] after division. For burst tasks, the same target transform is used for the two arc endpoints; blend has its own frozen normalization.
WER example: WER 5% multiplies the score by 0.95; WER 20% by 0.80; WER ≥ 100% by zero. Reported raw WER remains uncapped. The displayed fitness is the mean of per-take products, not the product of the displayed means.
Separate qualification gates: at least 90% valid, terminated takes; per-take WER ≤ 15% to count as valid; candidate mean WER ≤ 10%; paired reference cosine must not fall by more than 0.05 over all 50 reference comparisons. Corrected burst conditions additionally need every requested family detected in at least 60% of takes. Passing a gate is not a weighted score bonus.
Passing study gates. N=100 across both languages, both reference arms and all five seeds. Full-coverage steering-v2 evolution; transfer, fixed-wording CFG and sealed splits remain distinct.
0.879
6.365
4.460
0.16%
0.783
100.00%
-0.000
Prompting strategy: inline-only delivery cues.
Reusable direction recipe
Stored GENERAL content — NOT forwarded in inline-onlyP09
A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. Let the local direction carry the intended delivery beside each sentence.
Three local directions
1. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. Let the local direction carry the intended delivery beside each sentence.
2. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. Let the local direction carry the intended delivery beside each sentence.
3. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. Let the local direction carry the intended delivery beside each sentence.
Exact tested prompt — fixed EN example, seed1701, no reference
GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.
SCRIPT:
(A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. Let the local direction carry the intended delivery beside each sentence.) [4.2 seconds duration] I thought you would be here before the lights went out.
(A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. Let the local direction carry the intended delivery beside each sentence.) [3.8 seconds duration] Wait... I need a moment to understand what just happened.
(A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. Let the local direction carry the intended delivery beside each sentence.) [3.5 seconds duration] Stay with me until I can say this properly.
Literal words in the separate text field:
I thought you would be here before the lights went out. Wait... I need a moment to understand what just happened. Stay with me until I can say this properly.
Exact neutral CFG branch
GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.
SCRIPT:
(Natural clear speech.) [4.2 seconds duration] I thought you would be here before the lights went out.
(Natural clear speech.) [3.8 seconds duration] Wait... I need a moment to understand what just happened.
(Natural clear speech.) [3.5 seconds duration] Stay with me until I can say this properly.
Reference versus no-reference breakdown
reference
50
0.875
6.360
4.412
0.13%
0.597
none
50
0.884
6.371
4.507
0.19%
n/a
g0-p08 · CFG 2
full-v2-gen0CANONICAL
Current ranking: EARS Steering v2
Emotion/style only. Content Enjoyment: 0%. Genuineness: 0%. These remain visible diagnostics, not reward terms. WER and speaker identity are separate qualification gates, not terms in this score.
Observed head means below refer to g0-p08 at CFG 2; every recipe also shows its own values below. These are recipe means, not one hand-picked audio.
Φ is a Gaussian CDF using the frozen EARS training statistics. This makes mixed emotion/style heads comparable; it is not an emotion probability. It compresses extreme scores, so raw means remain visible too. A pure-emotion condition has only its named emotion head, at 100%. A mixed anger/ranting scene uses 50% Anger + 50% Ranting. Each recipe score is the mean over all its takes, including failed takes.
Unchanged gates: ≥90% valid terminated takes; per-take WER ≤15% to count as valid; candidate mean WER ≤10%; paired reference cosine decline no worse than −0.05. Existing family-aware burst conditions still need all requested families detected in ≥60% of takes to qualify as reliable burst recipes. A high emotion/style score alone does not establish a successful scream or a correctly timed transition.
Historical quality-fitness diagnostic — not the current steering ranking
Exact score for this condition
This is the prompt-search fitness, not the training RL reward. Higher means better under this particular proxy combination, not automatically more natural or more enjoyable to a listener.
50% target emotion/style, 25% content enjoyment, 25% genuineness. The target is vn_S_CONV_reg. VoiceCLAP and absolute speaker cosine are diagnostics here, not extra weighted rewards.
Fixed target normalization: z = (raw score − 2.533) / 1.305; target_CDF = Φ(z). This is a percentile-like training-distribution transform, not an emotion probability. CE and genuineness are clipped to [0, 1] after division. For burst tasks, the same target transform is used for the two arc endpoints; blend has its own frozen normalization.
WER example: WER 5% multiplies the score by 0.95; WER 20% by 0.80; WER ≥ 100% by zero. Reported raw WER remains uncapped. The displayed fitness is the mean of per-take products, not the product of the displayed means.
Separate qualification gates: at least 90% valid, terminated takes; per-take WER ≤ 15% to count as valid; candidate mean WER ≤ 10%; paired reference cosine must not fall by more than 0.05 over all 50 reference comparisons. Corrected burst conditions additionally need every requested family detected in at least 60% of takes. Passing a gate is not a weighted score bonus.
Passing study gates. N=100 across both languages, both reference arms and all five seeds. Full-coverage steering-v2 evolution; transfer, fixed-wording CFG and sealed splits remain distinct.
0.879
6.357
4.389
0.37%
0.778
100.00%
-0.015
Prompting strategy: GENERAL-only delivery cues.
Reusable direction recipe
Stored GENERAL content
A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. Carry the full intended delivery through the global direction.
Three local directions — NOT forwarded in GENERAL-onlyP08; neutral speech cues remain
1. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. Carry the full intended delivery through the global direction.
2. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. Carry the full intended delivery through the global direction.
3. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. Carry the full intended delivery through the global direction.
Exact tested prompt — fixed EN example, seed1701, no reference
GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. Carry the full intended delivery through the global direction.
SCRIPT:
(Natural clear speech.) [4.2 seconds duration] I thought you would be here before the lights went out.
(Natural clear speech.) [3.8 seconds duration] Wait... I need a moment to understand what just happened.
(Natural clear speech.) [3.5 seconds duration] Stay with me until I can say this properly.
Literal words in the separate text field:
I thought you would be here before the lights went out. Wait... I need a moment to understand what just happened. Stay with me until I can say this properly.
Exact neutral CFG branch
GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.
SCRIPT:
(Natural clear speech.) [4.2 seconds duration] I thought you would be here before the lights went out.
(Natural clear speech.) [3.8 seconds duration] Wait... I need a moment to understand what just happened.
(Natural clear speech.) [3.5 seconds duration] Stay with me until I can say this properly.
Reference versus no-reference breakdown
reference
50
0.881
6.337
4.332
0.27%
0.582
none
50
0.876
6.377
4.447
0.47%
n/a
g0-p07 · CFG 1.5
full-v2-gen0CANONICAL
Current ranking: EARS Steering v2
Emotion/style only. Content Enjoyment: 0%. Genuineness: 0%. These remain visible diagnostics, not reward terms. WER and speaker identity are separate qualification gates, not terms in this score.
Observed head means below refer to g0-p07 at CFG 1.5; every recipe also shows its own values below. These are recipe means, not one hand-picked audio.
Φ is a Gaussian CDF using the frozen EARS training statistics. This makes mixed emotion/style heads comparable; it is not an emotion probability. It compresses extreme scores, so raw means remain visible too. A pure-emotion condition has only its named emotion head, at 100%. A mixed anger/ranting scene uses 50% Anger + 50% Ranting. Each recipe score is the mean over all its takes, including failed takes.
Unchanged gates: ≥90% valid terminated takes; per-take WER ≤15% to count as valid; candidate mean WER ≤10%; paired reference cosine decline no worse than −0.05. Existing family-aware burst conditions still need all requested families detected in ≥60% of takes to qualify as reliable burst recipes. A high emotion/style score alone does not establish a successful scream or a correctly timed transition.
Historical quality-fitness diagnostic — not the current steering ranking
Exact score for this condition
This is the prompt-search fitness, not the training RL reward. Higher means better under this particular proxy combination, not automatically more natural or more enjoyable to a listener.
50% target emotion/style, 25% content enjoyment, 25% genuineness. The target is vn_S_CONV_reg. VoiceCLAP and absolute speaker cosine are diagnostics here, not extra weighted rewards.
Fixed target normalization: z = (raw score − 2.533) / 1.305; target_CDF = Φ(z). This is a percentile-like training-distribution transform, not an emotion probability. CE and genuineness are clipped to [0, 1] after division. For burst tasks, the same target transform is used for the two arc endpoints; blend has its own frozen normalization.
WER example: WER 5% multiplies the score by 0.95; WER 20% by 0.80; WER ≥ 100% by zero. Reported raw WER remains uncapped. The displayed fitness is the mean of per-take products, not the product of the displayed means.
Separate qualification gates: at least 90% valid, terminated takes; per-take WER ≤ 15% to count as valid; candidate mean WER ≤ 10%; paired reference cosine must not fall by more than 0.05 over all 50 reference comparisons. Corrected burst conditions additionally need every requested family detected in at least 60% of takes. Passing a gate is not a weighted score bonus.
Passing study gates. N=100 across both languages, both reference arms and all five seeds. Full-coverage steering-v2 evolution; transfer, fixed-wording CFG and sealed splits remain distinct.
0.877
6.383
4.398
0.44%
0.778
100.00%
0.015
Prompting strategy: reacting to an unheard partner.
Reusable direction recipe
Stored GENERAL content
A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. React to a silent partner standing beside you; generate only one adult voice.
Three local directions
1. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. React to a silent partner standing beside you; generate only one adult voice.
2. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. React to a silent partner standing beside you; generate only one adult voice.
3. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. React to a silent partner standing beside you; generate only one adult voice.
Exact tested prompt — fixed EN example, seed1701, no reference
GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. React to a silent partner standing beside you; generate only one adult voice.
SCRIPT:
(A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. React to a silent partner standing beside you; generate only one adult voice.) [4.2 seconds duration] I thought you would be here before the lights went out.
(A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. React to a silent partner standing beside you; generate only one adult voice.) [3.8 seconds duration] Wait... I need a moment to understand what just happened.
(A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. React to a silent partner standing beside you; generate only one adult voice.) [3.5 seconds duration] Stay with me until I can say this properly.
Literal words in the separate text field:
I thought you would be here before the lights went out. Wait... I need a moment to understand what just happened. Stay with me until I can say this properly.
Exact neutral CFG branch
GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.
SCRIPT:
(Natural clear speech.) [4.2 seconds duration] I thought you would be here before the lights went out.
(Natural clear speech.) [3.8 seconds duration] Wait... I need a moment to understand what just happened.
(Natural clear speech.) [3.5 seconds duration] Stay with me until I can say this properly.
Reference versus no-reference breakdown
reference
50
0.871
6.364
4.338
0.61%
0.613
none
50
0.884
6.403
4.458
0.26%
n/a
g0-p02 · CFG 1.25
full-v2-gen0CANONICAL
Current ranking: EARS Steering v2
Emotion/style only. Content Enjoyment: 0%. Genuineness: 0%. These remain visible diagnostics, not reward terms. WER and speaker identity are separate qualification gates, not terms in this score.
Observed head means below refer to g0-p02 at CFG 1.25; every recipe also shows its own values below. These are recipe means, not one hand-picked audio.
Φ is a Gaussian CDF using the frozen EARS training statistics. This makes mixed emotion/style heads comparable; it is not an emotion probability. It compresses extreme scores, so raw means remain visible too. A pure-emotion condition has only its named emotion head, at 100%. A mixed anger/ranting scene uses 50% Anger + 50% Ranting. Each recipe score is the mean over all its takes, including failed takes.
Unchanged gates: ≥90% valid terminated takes; per-take WER ≤15% to count as valid; candidate mean WER ≤10%; paired reference cosine decline no worse than −0.05. Existing family-aware burst conditions still need all requested families detected in ≥60% of takes to qualify as reliable burst recipes. A high emotion/style score alone does not establish a successful scream or a correctly timed transition.
Historical quality-fitness diagnostic — not the current steering ranking
Exact score for this condition
This is the prompt-search fitness, not the training RL reward. Higher means better under this particular proxy combination, not automatically more natural or more enjoyable to a listener.
50% target emotion/style, 25% content enjoyment, 25% genuineness. The target is vn_S_CONV_reg. VoiceCLAP and absolute speaker cosine are diagnostics here, not extra weighted rewards.
Fixed target normalization: z = (raw score − 2.533) / 1.305; target_CDF = Φ(z). This is a percentile-like training-distribution transform, not an emotion probability. CE and genuineness are clipped to [0, 1] after division. For burst tasks, the same target transform is used for the two arc endpoints; blend has its own frozen normalization.
WER example: WER 5% multiplies the score by 0.95; WER 20% by 0.80; WER ≥ 100% by zero. Reported raw WER remains uncapped. The displayed fitness is the mean of per-take products, not the product of the displayed means.
Separate qualification gates: at least 90% valid, terminated takes; per-take WER ≤ 15% to count as valid; candidate mean WER ≤ 10%; paired reference cosine must not fall by more than 0.05 over all 50 reference comparisons. Corrected burst conditions additionally need every requested family detected in at least 60% of takes. Passing a gate is not a weighted score bonus.
Passing study gates. N=100 across both languages, both reference arms and all five seeds. Full-coverage steering-v2 evolution; transfer, fixed-wording CFG and sealed splits remain distinct.
0.875
6.374
4.418
0.23%
0.779
100.00%
0.007
Prompting strategy: audible mechanics.
Reusable direction recipe
Stored GENERAL content
A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. Use physical breath pressure, pitch motion, consonant attack and release rather than naming the feeling.
Three local directions
1. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. Use physical breath pressure, pitch motion, consonant attack and release rather than naming the feeling.
2. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. Use physical breath pressure, pitch motion, consonant attack and release rather than naming the feeling.
3. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. Use physical breath pressure, pitch motion, consonant attack and release rather than naming the feeling.
Exact tested prompt — fixed EN example, seed1701, no reference
GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. Use physical breath pressure, pitch motion, consonant attack and release rather than naming the feeling.
SCRIPT:
(A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. Use physical breath pressure, pitch motion, consonant attack and release rather than naming the feeling.) [4.2 seconds duration] I thought you would be here before the lights went out.
(A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. Use physical breath pressure, pitch motion, consonant attack and release rather than naming the feeling.) [3.8 seconds duration] Wait... I need a moment to understand what just happened.
(A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. Use physical breath pressure, pitch motion, consonant attack and release rather than naming the feeling.) [3.5 seconds duration] Stay with me until I can say this properly.
Literal words in the separate text field:
I thought you would be here before the lights went out. Wait... I need a moment to understand what just happened. Stay with me until I can say this properly.
Exact neutral CFG branch
GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.
SCRIPT:
(Natural clear speech.) [4.2 seconds duration] I thought you would be here before the lights went out.
(Natural clear speech.) [3.8 seconds duration] Wait... I need a moment to understand what just happened.
(Natural clear speech.) [3.5 seconds duration] Stay with me until I can say this properly.
Reference versus no-reference breakdown
reference
50
0.870
6.353
4.372
0.20%
0.604
none
50
0.880
6.395
4.463
0.27%
n/a
g0-p04 · CFG 2
full-v2-gen0CANONICAL
Current ranking: EARS Steering v2
Emotion/style only. Content Enjoyment: 0%. Genuineness: 0%. These remain visible diagnostics, not reward terms. WER and speaker identity are separate qualification gates, not terms in this score.
Observed head means below refer to g0-p04 at CFG 2; every recipe also shows its own values below. These are recipe means, not one hand-picked audio.
Φ is a Gaussian CDF using the frozen EARS training statistics. This makes mixed emotion/style heads comparable; it is not an emotion probability. It compresses extreme scores, so raw means remain visible too. A pure-emotion condition has only its named emotion head, at 100%. A mixed anger/ranting scene uses 50% Anger + 50% Ranting. Each recipe score is the mean over all its takes, including failed takes.
Unchanged gates: ≥90% valid terminated takes; per-take WER ≤15% to count as valid; candidate mean WER ≤10%; paired reference cosine decline no worse than −0.05. Existing family-aware burst conditions still need all requested families detected in ≥60% of takes to qualify as reliable burst recipes. A high emotion/style score alone does not establish a successful scream or a correctly timed transition.
Historical quality-fitness diagnostic — not the current steering ranking
Exact score for this condition
This is the prompt-search fitness, not the training RL reward. Higher means better under this particular proxy combination, not automatically more natural or more enjoyable to a listener.
50% target emotion/style, 25% content enjoyment, 25% genuineness. The target is vn_S_CONV_reg. VoiceCLAP and absolute speaker cosine are diagnostics here, not extra weighted rewards.
Fixed target normalization: z = (raw score − 2.533) / 1.305; target_CDF = Φ(z). This is a percentile-like training-distribution transform, not an emotion probability. CE and genuineness are clipped to [0, 1] after division. For burst tasks, the same target transform is used for the two arc endpoints; blend has its own frozen normalization.
WER example: WER 5% multiplies the score by 0.95; WER 20% by 0.80; WER ≥ 100% by zero. Reported raw WER remains uncapped. The displayed fitness is the mean of per-take products, not the product of the displayed means.
Separate qualification gates: at least 90% valid, terminated takes; per-take WER ≤ 15% to count as valid; candidate mean WER ≤ 10%; paired reference cosine must not fall by more than 0.05 over all 50 reference comparisons. Corrected burst conditions additionally need every requested family detected in at least 60% of takes. Passing a gate is not a weighted score bonus.
Passing study gates. N=100 across both languages, both reference arms and all five seeds. Full-coverage steering-v2 evolution; transfer, fixed-wording CFG and sealed splits remain distinct.
0.875
6.375
4.407
0.38%
0.778
100.00%
0.030
Prompting strategy: organic conversation.
Reusable direction recipe
Stored GENERAL content
A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. Think while speaking, with organic micro pauses and spontaneous emphasis, not a polished reading.
Three local directions
1. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. Think while speaking, with organic micro pauses and spontaneous emphasis, not a polished reading.
2. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. Think while speaking, with organic micro pauses and spontaneous emphasis, not a polished reading.
3. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. Think while speaking, with organic micro pauses and spontaneous emphasis, not a polished reading.
Exact tested prompt — fixed EN example, seed1701, no reference
GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. Think while speaking, with organic micro pauses and spontaneous emphasis, not a polished reading.
SCRIPT:
(A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. Think while speaking, with organic micro pauses and spontaneous emphasis, not a polished reading.) [4.2 seconds duration] I thought you would be here before the lights went out.
(A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. Think while speaking, with organic micro pauses and spontaneous emphasis, not a polished reading.) [3.8 seconds duration] Wait... I need a moment to understand what just happened.
(A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. Think while speaking, with organic micro pauses and spontaneous emphasis, not a polished reading.) [3.5 seconds duration] Stay with me until I can say this properly.
Literal words in the separate text field:
I thought you would be here before the lights went out. Wait... I need a moment to understand what just happened. Stay with me until I can say this properly.
Exact neutral CFG branch
GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.
SCRIPT:
(Natural clear speech.) [4.2 seconds duration] I thought you would be here before the lights went out.
(Natural clear speech.) [3.8 seconds duration] Wait... I need a moment to understand what just happened.
(Natural clear speech.) [3.5 seconds duration] Stay with me until I can say this properly.
Reference versus no-reference breakdown
reference
50
0.871
6.359
4.357
0.48%
0.627
none
50
0.879
6.392
4.456
0.27%
n/a
g0-p01 · CFG 1
full-v2-gen0CANONICAL
Current ranking: EARS Steering v2
Emotion/style only. Content Enjoyment: 0%. Genuineness: 0%. These remain visible diagnostics, not reward terms. WER and speaker identity are separate qualification gates, not terms in this score.
Observed head means below refer to g0-p01 at CFG 1; every recipe also shows its own values below. These are recipe means, not one hand-picked audio.
Φ is a Gaussian CDF using the frozen EARS training statistics. This makes mixed emotion/style heads comparable; it is not an emotion probability. It compresses extreme scores, so raw means remain visible too. A pure-emotion condition has only its named emotion head, at 100%. A mixed anger/ranting scene uses 50% Anger + 50% Ranting. Each recipe score is the mean over all its takes, including failed takes.
Unchanged gates: ≥90% valid terminated takes; per-take WER ≤15% to count as valid; candidate mean WER ≤10%; paired reference cosine decline no worse than −0.05. Existing family-aware burst conditions still need all requested families detected in ≥60% of takes to qualify as reliable burst recipes. A high emotion/style score alone does not establish a successful scream or a correctly timed transition.
Historical quality-fitness diagnostic — not the current steering ranking
Exact score for this condition
This is the prompt-search fitness, not the training RL reward. Higher means better under this particular proxy combination, not automatically more natural or more enjoyable to a listener.
50% target emotion/style, 25% content enjoyment, 25% genuineness. The target is vn_S_CONV_reg. VoiceCLAP and absolute speaker cosine are diagnostics here, not extra weighted rewards.
Fixed target normalization: z = (raw score − 2.533) / 1.305; target_CDF = Φ(z). This is a percentile-like training-distribution transform, not an emotion probability. CE and genuineness are clipped to [0, 1] after division. For burst tasks, the same target transform is used for the two arc endpoints; blend has its own frozen normalization.
WER example: WER 5% multiplies the score by 0.95; WER 20% by 0.80; WER ≥ 100% by zero. Reported raw WER remains uncapped. The displayed fitness is the mean of per-take products, not the product of the displayed means.
Separate qualification gates: at least 90% valid, terminated takes; per-take WER ≤ 15% to count as valid; candidate mean WER ≤ 10%; paired reference cosine must not fall by more than 0.05 over all 50 reference comparisons. Corrected burst conditions additionally need every requested family detected in at least 60% of takes. Passing a gate is not a weighted score bonus.
Passing study gates. N=100 across both languages, both reference arms and all five seeds. Full-coverage steering-v2 evolution; transfer, fixed-wording CFG and sealed splits remain distinct.
0.872
6.383
4.422
0.54%
0.776
100.00%
0.000
Prompting strategy: concise trained baseline.
Reusable direction recipe
Stored GENERAL content
A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. Make the intended feeling unmistakably audible while preserving every word.
Three local directions
1. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. Make the intended feeling unmistakably audible while preserving every word.
2. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. Make the intended feeling unmistakably audible while preserving every word.
3. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. Make the intended feeling unmistakably audible while preserving every word.
Exact tested prompt — fixed EN example, seed1701, no reference
GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. Make the intended feeling unmistakably audible while preserving every word.
SCRIPT:
(A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. Make the intended feeling unmistakably audible while preserving every word.) [4.2 seconds duration] I thought you would be here before the lights went out.
(A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. Make the intended feeling unmistakably audible while preserving every word.) [3.8 seconds duration] Wait... I need a moment to understand what just happened.
(A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. Make the intended feeling unmistakably audible while preserving every word.) [3.5 seconds duration] Stay with me until I can say this properly.
Literal words in the separate text field:
I thought you would be here before the lights went out. Wait... I need a moment to understand what just happened. Stay with me until I can say this properly.
Exact neutral CFG branch
GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.
SCRIPT:
(Natural clear speech.) [4.2 seconds duration] I thought you would be here before the lights went out.
(Natural clear speech.) [3.8 seconds duration] Wait... I need a moment to understand what just happened.
(Natural clear speech.) [3.5 seconds duration] Stay with me until I can say this properly.
Reference versus no-reference breakdown
reference
50
0.873
6.368
4.372
0.40%
0.597
none
50
0.872
6.397
4.471
0.68%
n/a
g0-p03 · CFG 1.5
full-v2-gen0CANONICAL
Current ranking: EARS Steering v2
Emotion/style only. Content Enjoyment: 0%. Genuineness: 0%. These remain visible diagnostics, not reward terms. WER and speaker identity are separate qualification gates, not terms in this score.
Observed head means below refer to g0-p03 at CFG 1.5; every recipe also shows its own values below. These are recipe means, not one hand-picked audio.
Φ is a Gaussian CDF using the frozen EARS training statistics. This makes mixed emotion/style heads comparable; it is not an emotion probability. It compresses extreme scores, so raw means remain visible too. A pure-emotion condition has only its named emotion head, at 100%. A mixed anger/ranting scene uses 50% Anger + 50% Ranting. Each recipe score is the mean over all its takes, including failed takes.
Unchanged gates: ≥90% valid terminated takes; per-take WER ≤15% to count as valid; candidate mean WER ≤10%; paired reference cosine decline no worse than −0.05. Existing family-aware burst conditions still need all requested families detected in ≥60% of takes to qualify as reliable burst recipes. A high emotion/style score alone does not establish a successful scream or a correctly timed transition.
Historical quality-fitness diagnostic — not the current steering ranking
Exact score for this condition
This is the prompt-search fitness, not the training RL reward. Higher means better under this particular proxy combination, not automatically more natural or more enjoyable to a listener.
50% target emotion/style, 25% content enjoyment, 25% genuineness. The target is vn_S_CONV_reg. VoiceCLAP and absolute speaker cosine are diagnostics here, not extra weighted rewards.
Fixed target normalization: z = (raw score − 2.533) / 1.305; target_CDF = Φ(z). This is a percentile-like training-distribution transform, not an emotion probability. CE and genuineness are clipped to [0, 1] after division. For burst tasks, the same target transform is used for the two arc endpoints; blend has its own frozen normalization.
WER example: WER 5% multiplies the score by 0.95; WER 20% by 0.80; WER ≥ 100% by zero. Reported raw WER remains uncapped. The displayed fitness is the mean of per-take products, not the product of the displayed means.
Separate qualification gates: at least 90% valid, terminated takes; per-take WER ≤ 15% to count as valid; candidate mean WER ≤ 10%; paired reference cosine must not fall by more than 0.05 over all 50 reference comparisons. Corrected burst conditions additionally need every requested family detected in at least 60% of takes. Passing a gate is not a weighted score bonus.
Passing study gates. N=100 across both languages, both reference arms and all five seeds. Full-coverage steering-v2 evolution; transfer, fixed-wording CFG and sealed splits remain distinct.
0.868
6.359
4.407
0.51%
0.773
99.00%
0.016
Prompting strategy: situation and intention.
Reusable direction recipe
Stored GENERAL content
A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. You urgently need one unheard listener to understand this; speak to that person, not to an audience.
Three local directions
1. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. You urgently need one unheard listener to understand this; speak to that person, not to an audience.
2. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. You urgently need one unheard listener to understand this; speak to that person, not to an audience.
3. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. You urgently need one unheard listener to understand this; speak to that person, not to an audience.
Exact tested prompt — fixed EN example, seed1701, no reference
GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. You urgently need one unheard listener to understand this; speak to that person, not to an audience.
SCRIPT:
(A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Establish it immediately. You urgently need one unheard listener to understand this; speak to that person, not to an audience.) [4.2 seconds duration] I thought you would be here before the lights went out.
(A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Sustain the delivery. You urgently need one unheard listener to understand this; speak to that person, not to an audience.) [3.8 seconds duration] Wait... I need a moment to understand what just happened.
(A spontaneous conversation with one unheard partner; responsive emphasis and natural thought pauses. Preserve the intended style through the end. You urgently need one unheard listener to understand this; speak to that person, not to an audience.) [3.5 seconds duration] Stay with me until I can say this properly.
Literal words in the separate text field:
I thought you would be here before the lights went out. Wait... I need a moment to understand what just happened. Stay with me until I can say this properly.
Exact neutral CFG branch
GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.
SCRIPT:
(Natural clear speech.) [4.2 seconds duration] I thought you would be here before the lights went out.
(Natural clear speech.) [3.8 seconds duration] Wait... I need a moment to understand what just happened.
(Natural clear speech.) [3.5 seconds duration] Stay with me until I can say this properly.