Speech AI Preferences Test

Preference test between speech synthesis models

Overall

model1_f0_new (Model A)

3.05

model2_f0_old (Model B)

By Tag

model1_f0_new (Model A)

model2_f0_old (Model B)

Answers

1. A is strongly preferred

2. A is preferred

3. About the same

4. B is preferred

5. B is strongly preferred

86 (6.28%)

266 (19.42%)

607 (44.31%)

315 (22.99%)

96 (7.01%)

Tags

All tags

model1_f0_new

model2_f0_old

86 (6.28%)

Deep analysis

1. A is strongly preferred

2. A is preferred

3. About the same

4. B is preferred

5. B is strongly preferred

		Tags	Statistics
00:00 --:-- 1_f0_new\Loud-normal_Female_audio_32.wav 00:00 --:-- 2_f0_old\Loud-normal_Female_audio_32.wav	model1_f0_new model2_f0_old	model1 model2	3.40 0.37 0.16
00:00 --:-- 1_f0_new\Loud-normal_Female_audio_34.wav 00:00 --:-- 2_f0_old\Loud-normal_Female_audio_34.wav	model1_f0_new model2_f0_old	model1 model2	3.00 0.48 0.21
00:00 --:-- 1_f0_new\Loud-normal_female_audio_35.wav 00:00 --:-- 2_f0_old\Loud-normal_female_audio_35.wav	model1_f0_new model2_f0_old	model1 model2	2.80 0.45 0.20
00:00 --:-- 1_f0_new\Loud-normal_male_audio_33.wav 00:00 --:-- 2_f0_old\Loud-normal_male_audio_33.wav	model1_f0_new model2_f0_old	model1 model2	3.00 0.48 0.21
00:00 --:-- 1_f0_new\Loud-quiet_male_audio_33.wav 00:00 --:-- 2_f0_old\Loud-quiet_male_audio_33.wav	model1_f0_new model2_f0_old	model1 model2	3.80 0.56 0.25
00:00 --:-- 1_f0_new\Loud_Normal_Female_Audio-08.wav 00:00 --:-- 2_f0_old\Loud_Normal_Female_Audio-08.wav	model1_f0_new model2_f0_old	model1 model2	3.00 0.34 0.15
00:00 --:-- 1_f0_new\Loud_Normal_Female_Audio-09.wav 00:00 --:-- 2_f0_old\Loud_Normal_Female_Audio-09.wav	model1_f0_new model2_f0_old	model1 model2	2.70 0.59 0.26
00:00 --:-- 1_f0_new\Loud_Normal_Female_Audio-10.wav 00:00 --:-- 2_f0_old\Loud_Normal_Female_Audio-10.wav	model1_f0_new model2_f0_old	model1 model2	3.00 0.67 0.30

Overall

model1_f0_new (Model A)

3.10

model2_f0_old (Model B)

By Tag

model1_f0_new (Model A)

model2_f0_old (Model B)

Answers

1. A is strongly preferred

2. A is preferred

3. About the same

4. B is preferred

5. B is strongly preferred

47 (3.43%)

227 (16.57%)

720 (52.55%)

292 (21.31%)

84 (6.13%)

Tags

All tags

model1_f0_new

model2_f0_old

47 (3.43%)

Deep analysis

1. A is strongly preferred

2. A is preferred

3. About the same

4. B is preferred

5. B is strongly preferred

		Tags	Statistics
00:00 --:-- 1_f0_new\Loud-normal_Female_audio_32.wav 00:00 --:-- 2_f0_old\Loud-normal_Female_audio_32.wav	model1_f0_new model2_f0_old	model1 model2	3.40 0.69 0.31
00:00 --:-- 1_f0_new\Loud-normal_Female_audio_34.wav 00:00 --:-- 2_f0_old\Loud-normal_Female_audio_34.wav	model1_f0_new model2_f0_old	model1 model2	3.40 0.50 0.22
00:00 --:-- 1_f0_new\Loud-normal_female_audio_35.wav 00:00 --:-- 2_f0_old\Loud-normal_female_audio_35.wav	model1_f0_new model2_f0_old	model1 model2	3.00 0.34 0.15
00:00 --:-- 1_f0_new\Loud-normal_male_audio_33.wav 00:00 --:-- 2_f0_old\Loud-normal_male_audio_33.wav	model1_f0_new model2_f0_old	model1 model2	3.00 0.34 0.15
00:00 --:-- 1_f0_new\Loud-quiet_male_audio_33.wav 00:00 --:-- 2_f0_old\Loud-quiet_male_audio_33.wav	model1_f0_new model2_f0_old	model1 model2	3.40 0.50 0.22
00:00 --:-- 1_f0_new\Loud_Normal_Female_Audio-08.wav 00:00 --:-- 2_f0_old\Loud_Normal_Female_Audio-08.wav	model1_f0_new model2_f0_old	model1 model2	3.00 0.34 0.15
00:00 --:-- 1_f0_new\Loud_Normal_Female_Audio-09.wav 00:00 --:-- 2_f0_old\Loud_Normal_Female_Audio-09.wav	model1_f0_new model2_f0_old	model1 model2	3.00 0.83 0.37
00:00 --:-- 1_f0_new\Loud_Normal_Female_Audio-10.wav 00:00 --:-- 2_f0_old\Loud_Normal_Female_Audio-10.wav	model1_f0_new model2_f0_old	model1 model2	3.00 0.48 0.21

Overall

model1_f0_new (Model A)

3.05

model2_f0_old (Model B)

By Tag

model1_f0_new (Model A)

model2_f0_old (Model B)

Answers

1. A is strongly preferred

2. A is preferred

3. About the same

4. B is preferred

5. B is strongly preferred

86 (6.28%)

266 (19.42%)

607 (44.31%)

315 (22.99%)

96 (7.01%)

Tags

All tags

model1_f0_new

model2_f0_old

86 (6.28%)

Deep analysis

1. A is strongly preferred

2. A is preferred

3. About the same

4. B is preferred

5. B is strongly preferred

		Tags	Statistics
00:00 --:-- 1_f0_new\Loud-normal_Female_audio_32.wav 00:00 --:-- 2_f0_old\Loud-normal_Female_audio_32.wav	model1_f0_new model2_f0_old	model1 model2	3.40 0.37 0.16
00:00 --:-- 1_f0_new\Loud-normal_Female_audio_34.wav 00:00 --:-- 2_f0_old\Loud-normal_Female_audio_34.wav	model1_f0_new model2_f0_old	model1 model2	3.00 0.48 0.21
00:00 --:-- 1_f0_new\Loud-normal_female_audio_35.wav 00:00 --:-- 2_f0_old\Loud-normal_female_audio_35.wav	model1_f0_new model2_f0_old	model1 model2	2.80 0.45 0.20
00:00 --:-- 1_f0_new\Loud-normal_male_audio_33.wav 00:00 --:-- 2_f0_old\Loud-normal_male_audio_33.wav	model1_f0_new model2_f0_old	model1 model2	3.00 0.48 0.21
00:00 --:-- 1_f0_new\Loud-quiet_male_audio_33.wav 00:00 --:-- 2_f0_old\Loud-quiet_male_audio_33.wav	model1_f0_new model2_f0_old	model1 model2	3.80 0.56 0.25
00:00 --:-- 1_f0_new\Loud_Normal_Female_Audio-08.wav 00:00 --:-- 2_f0_old\Loud_Normal_Female_Audio-08.wav	model1_f0_new model2_f0_old	model1 model2	3.00 0.34 0.15
00:00 --:-- 1_f0_new\Loud_Normal_Female_Audio-09.wav 00:00 --:-- 2_f0_old\Loud_Normal_Female_Audio-09.wav	model1_f0_new model2_f0_old	model1 model2	2.70 0.59 0.26
00:00 --:-- 1_f0_new\Loud_Normal_Female_Audio-10.wav 00:00 --:-- 2_f0_old\Loud_Normal_Female_Audio-10.wav	model1_f0_new model2_f0_old	model1 model2	3.00 0.67 0.30

Overall

model1_f0_new (Model A)

3.10

model2_f0_old (Model B)

By Tag

model1_f0_new (Model A)

model2_f0_old (Model B)

Answers

1. A is strongly preferred

2. A is preferred

3. About the same

4. B is preferred

5. B is strongly preferred

47 (3.43%)

227 (16.57%)

720 (52.55%)

292 (21.31%)

84 (6.13%)

Tags

All tags

model1_f0_new

model2_f0_old

47 (3.43%)

Deep analysis

1. A is strongly preferred

2. A is preferred

3. About the same

4. B is preferred

5. B is strongly preferred

		Tags	Statistics
00:00 --:-- 1_f0_new\Loud-normal_Female_audio_32.wav 00:00 --:-- 2_f0_old\Loud-normal_Female_audio_32.wav	model1_f0_new model2_f0_old	model1 model2	3.40 0.69 0.31
00:00 --:-- 1_f0_new\Loud-normal_Female_audio_34.wav 00:00 --:-- 2_f0_old\Loud-normal_Female_audio_34.wav	model1_f0_new model2_f0_old	model1 model2	3.40 0.50 0.22
00:00 --:-- 1_f0_new\Loud-normal_female_audio_35.wav 00:00 --:-- 2_f0_old\Loud-normal_female_audio_35.wav	model1_f0_new model2_f0_old	model1 model2	3.00 0.34 0.15
00:00 --:-- 1_f0_new\Loud-normal_male_audio_33.wav 00:00 --:-- 2_f0_old\Loud-normal_male_audio_33.wav	model1_f0_new model2_f0_old	model1 model2	3.00 0.34 0.15
00:00 --:-- 1_f0_new\Loud-quiet_male_audio_33.wav 00:00 --:-- 2_f0_old\Loud-quiet_male_audio_33.wav	model1_f0_new model2_f0_old	model1 model2	3.40 0.50 0.22
00:00 --:-- 1_f0_new\Loud_Normal_Female_Audio-08.wav 00:00 --:-- 2_f0_old\Loud_Normal_Female_Audio-08.wav	model1_f0_new model2_f0_old	model1 model2	3.00 0.34 0.15
00:00 --:-- 1_f0_new\Loud_Normal_Female_Audio-09.wav 00:00 --:-- 2_f0_old\Loud_Normal_Female_Audio-09.wav	model1_f0_new model2_f0_old	model1 model2	3.00 0.83 0.37
00:00 --:-- 1_f0_new\Loud_Normal_Female_Audio-10.wav 00:00 --:-- 2_f0_old\Loud_Normal_Female_Audio-10.wav	model1_f0_new model2_f0_old	model1 model2	3.00 0.48 0.21