Book a demo
Feedback
Overview
Analysis
Audio Preference Evaluation
Evaluating preference between two speech models
Loading...
Loading...
Question 1. Listen to the speech/audio samples and rate which one you prefer.
Single select | Scored | 5 radio buttons
Overall
Model 2 (Model B)
-0.13
Model 1 (Model A)
By Tag
Model 2 (Model B)
Model 1 (Model A)
Answers
2. A is strongly preferred
1. A is preferred
0. About the same
-1. B is preferred
-2. B is strongly preferred
188 (9.35%)
374 (18.61%)
720 (35.82%)
439 (21.84%)
289 (14.38%)
2. A is strongly preferred
1. A is preferred
0. About the same
-1. B is preferred
-2. B is strongly preferred
Tags
All tags
Model 1
Model 2
188 (9.35%)
188 (9.35%)
Deep analysis
Distribution
Mean
Files
2. A is strongly preferred
1. A is preferred
0. About the same
-1. B is preferred
-2. B is strongly preferred
MAX 2
1
0
-1
MIN -2
MAX 2
1
0
-1
MIN -2
Tags: All tags
Filename
Model
Tags
Statistics
x̅
CI
σ
x̅
00:00
--:--
1\00000101_259892.wav
00:00
--:--
2\00000101_259892.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
-0.10
CI
0.92
σ
x̅
0.41
00:00
--:--
1\00000102_260555.wav
00:00
--:--
2\00000102_260555.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
-0.30
CI
0.68
σ
x̅
0.30
00:00
--:--
1\00000103_256771.wav
00:00
--:--
2\00000103_256771.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
0.30
CI
0.96
σ
x̅
0.42
00:00
--:--
1\00000104_260511.wav
00:00
--:--
2\00000104_260511.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
0.90
CI
0.92
σ
x̅
0.41
00:00
--:--
1\00000105_258305.wav
00:00
--:--
2\00000105_258305.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
-0.30
CI
0.68
σ
x̅
0.30
00:00
--:--
1\00000105_259604.wav
00:00
--:--
2\00000105_259604.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
-0.30
CI
0.48
σ
x̅
0.21
00:00
--:--
1\00000105_259649.wav
00:00
--:--
2\00000105_259649.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
0.00
CI
0.95
σ
x̅
0.42
00:00
--:--
1\00000105_259989.wav
00:00
--:--
2\00000105_259989.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
-0.30
CI
0.83
σ
x̅
0.37
1
2
3
4
5
...
26
Question 2. Which audio would you prefer for Naturalness?
Single select | Scored | 5 radio buttons
Overall
Model 2 (Model B)
-0.22
Model 1 (Model A)
By Tag
Model 2 (Model B)
Model 1 (Model A)
Answers
2. A is strongly preferred
1. A is preferred
0. About the same
-1. B is preferred
-2. B is strongly preferred
146 (7.26%)
303 (15.07%)
807 (40.15%)
480 (23.88%)
274 (13.63%)
2. A is strongly preferred
1. A is preferred
0. About the same
-1. B is preferred
-2. B is strongly preferred
Tags
All tags
Model 1
Model 2
146 (7.26%)
146 (7.26%)
Deep analysis
Distribution
Mean
Files
2. A is strongly preferred
1. A is preferred
0. About the same
-1. B is preferred
-2. B is strongly preferred
MAX 2
1
0
-1
MIN -2
MAX 2
1
0
-1
MIN -2
Tags: All tags
Filename
Model
Tags
Statistics
x̅
CI
σ
x̅
00:00
--:--
1\00000101_259892.wav
00:00
--:--
2\00000101_259892.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
-0.20
CI
0.94
σ
x̅
0.42
00:00
--:--
1\00000102_260555.wav
00:00
--:--
2\00000102_260555.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
-0.20
CI
0.56
σ
x̅
0.25
00:00
--:--
1\00000103_256771.wav
00:00
--:--
2\00000103_256771.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
-0.10
CI
0.71
σ
x̅
0.31
00:00
--:--
1\00000104_260511.wav
00:00
--:--
2\00000104_260511.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
0.80
CI
0.88
σ
x̅
0.39
00:00
--:--
1\00000105_258305.wav
00:00
--:--
2\00000105_258305.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
-0.10
CI
0.86
σ
x̅
0.38
00:00
--:--
1\00000105_259604.wav
00:00
--:--
2\00000105_259604.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
0.00
CI
0.67
σ
x̅
0.30
00:00
--:--
1\00000105_259649.wav
00:00
--:--
2\00000105_259649.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
0.50
CI
0.97
σ
x̅
0.43
00:00
--:--
1\00000105_259989.wav
00:00
--:--
2\00000105_259989.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
-0.60
CI
0.77
σ
x̅
0.34
1
2
3
4
5
...
26
Question 3. Which audio would you prefer for Intelligibility?
Single select | Scored | 5 radio buttons
Overall
Model 2 (Model B)
-0.25
Model 1 (Model A)
By Tag
Model 2 (Model B)
Model 1 (Model A)
Answers
2. A is strongly preferred
1. A is preferred
0. About the same
-1. B is preferred
-2. B is strongly preferred
139 (6.92%)
303 (15.07%)
764 (38.01%)
519 (25.82%)
285 (14.18%)
2. A is strongly preferred
1. A is preferred
0. About the same
-1. B is preferred
-2. B is strongly preferred
Tags
All tags
Model 1
Model 2
139 (6.92%)
139 (6.92%)
Deep analysis
Distribution
Mean
Files
2. A is strongly preferred
1. A is preferred
0. About the same
-1. B is preferred
-2. B is strongly preferred
MAX 2
1
0
-1
MIN -2
MAX 2
1
0
-1
MIN -2
Tags: All tags
Filename
Model
Tags
Statistics
x̅
CI
σ
x̅
00:00
--:--
1\00000101_259892.wav
00:00
--:--
2\00000101_259892.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
-0.30
CI
0.68
σ
x̅
0.30
00:00
--:--
1\00000102_260555.wav
00:00
--:--
2\00000102_260555.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
-0.40
CI
0.37
σ
x̅
0.16
00:00
--:--
1\00000103_256771.wav
00:00
--:--
2\00000103_256771.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
-0.10
CI
0.92
σ
x̅
0.41
00:00
--:--
1\00000104_260511.wav
00:00
--:--
2\00000104_260511.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
0.60
CI
0.84
σ
x̅
0.37
00:00
--:--
1\00000105_258305.wav
00:00
--:--
2\00000105_258305.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
0.00
CI
0.75
σ
x̅
0.33
00:00
--:--
1\00000105_259604.wav
00:00
--:--
2\00000105_259604.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
-0.20
CI
0.66
σ
x̅
0.29
00:00
--:--
1\00000105_259649.wav
00:00
--:--
2\00000105_259649.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
-0.10
CI
0.71
σ
x̅
0.31
00:00
--:--
1\00000105_259989.wav
00:00
--:--
2\00000105_259989.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
-0.60
CI
0.77
σ
x̅
0.34
1
2
3
4
5
...
26
Question 4. Which audio would you prefer for American Accent?
Single select | Scored | 5 radio buttons
Overall
Model 2 (Model B)
-0.22
Model 1 (Model A)
By Tag
Model 2 (Model B)
Model 1 (Model A)
Answers
2. A is strongly preferred
1. A is preferred
0. About the same
-1. B is preferred
-2. B is strongly preferred
130 (6.47%)
242 (12.04%)
990 (49.25%)
358 (17.81%)
290 (14.43%)
2. A is strongly preferred
1. A is preferred
0. About the same
-1. B is preferred
-2. B is strongly preferred
Tags
All tags
Model 1
Model 2
130 (6.47%)
130 (6.47%)
Deep analysis
Distribution
Mean
Files
2. A is strongly preferred
1. A is preferred
0. About the same
-1. B is preferred
-2. B is strongly preferred
MAX 2
1
0
-1
MIN -2
MAX 2
1
0
-1
MIN -2
Tags: All tags
Filename
Model
Tags
Statistics
x̅
CI
σ
x̅
00:00
--:--
1\00000101_259892.wav
00:00
--:--
2\00000101_259892.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
0.30
CI
0.76
σ
x̅
0.33
00:00
--:--
1\00000102_260555.wav
00:00
--:--
2\00000102_260555.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
-0.40
CI
0.50
σ
x̅
0.22
00:00
--:--
1\00000103_256771.wav
00:00
--:--
2\00000103_256771.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
-0.30
CI
0.83
σ
x̅
0.37
00:00
--:--
1\00000104_260511.wav
00:00
--:--
2\00000104_260511.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
0.90
CI
0.79
σ
x̅
0.35
00:00
--:--
1\00000105_258305.wav
00:00
--:--
2\00000105_258305.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
-0.10
CI
0.86
σ
x̅
0.38
00:00
--:--
1\00000105_259604.wav
00:00
--:--
2\00000105_259604.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
0.10
CI
0.63
σ
x̅
0.28
00:00
--:--
1\00000105_259649.wav
00:00
--:--
2\00000105_259649.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
-0.10
CI
0.86
σ
x̅
0.38
00:00
--:--
1\00000105_259989.wav
00:00
--:--
2\00000105_259989.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
-0.50
CI
0.70
σ
x̅
0.31
1
2
3
4
5
...
26
Question 5. Which audio would you prefer for Sound Quality?
Single select | Scored | 5 radio buttons
Overall
Model 2 (Model B)
-0.17
Model 1 (Model A)
By Tag
Model 2 (Model B)
Model 1 (Model A)
Answers
2. A is strongly preferred
1. A is preferred
0. About the same
-1. B is preferred
-2. B is strongly preferred
153 (7.61%)
274 (13.63%)
839 (41.74%)
565 (28.11%)
179 (8.91%)
2. A is strongly preferred
1. A is preferred
0. About the same
-1. B is preferred
-2. B is strongly preferred
Tags
All tags
Model 1
Model 2
153 (7.61%)
153 (7.61%)
Deep analysis
Distribution
Mean
Files
2. A is strongly preferred
1. A is preferred
0. About the same
-1. B is preferred
-2. B is strongly preferred
MAX 2
1
0
-1
MIN -2
MAX 2
1
0
-1
MIN -2
Tags: All tags
Filename
Model
Tags
Statistics
x̅
CI
σ
x̅
00:00
--:--
1\00000101_259892.wav
00:00
--:--
2\00000101_259892.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
-0.20
CI
0.81
σ
x̅
0.36
00:00
--:--
1\00000102_260555.wav
00:00
--:--
2\00000102_260555.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
-0.50
CI
0.38
σ
x̅
0.17
00:00
--:--
1\00000103_256771.wav
00:00
--:--
2\00000103_256771.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
0.00
CI
0.89
σ
x̅
0.39
00:00
--:--
1\00000104_260511.wav
00:00
--:--
2\00000104_260511.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
0.80
CI
0.66
σ
x̅
0.29
00:00
--:--
1\00000105_258305.wav
00:00
--:--
2\00000105_258305.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
-0.10
CI
0.71
σ
x̅
0.31
00:00
--:--
1\00000105_259604.wav
00:00
--:--
2\00000105_259604.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
-0.20
CI
0.56
σ
x̅
0.25
00:00
--:--
1\00000105_259649.wav
00:00
--:--
2\00000105_259649.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
0.00
CI
0.75
σ
x̅
0.33
00:00
--:--
1\00000105_259989.wav
00:00
--:--
2\00000105_259989.wav
Model 1
Model 2
model1
synthesized
model2
synthesized
x̅
-0.30
CI
0.83
σ
x̅
0.37
1
2
3
4
5
...
26