|
|
 |
AQuA Voice Testing Results
AQuA - Audio Quality Analyzer page
In the tables below you can find voice quality testing results obtained with the help of test files presented by www.signalogic.com and available for download from their web site.
Table 1
| PESQ/Sevana, % |
Sevana PESQ Score |
Sevana Quality Score, % |
(ITU) PESQ Score |
English Language 1 |
Sample |
| 100% |
4.5 |
100 |
4.5 |
male.wav |
Original
Fs = 8 kHz
(128000 bps) |
| 100% |
4.5 |
100 |
4.5 |
female.wav |
| 96,15% |
2,78 |
51,11 |
2,673 |
male600.wav |
MELPe-Plus17
600 bps
(30 ms) |
| 80,17% |
2,86 |
56,29 |
2,293 |
female600.wav |
Table 2
| PESQ/Sevana, % |
Sevana MOS Score, % |
Sevana PESQ Score |
Sevana Quality Score, % |
(ITU) PESQ Score |
English Language 2 |
Sample |
| 100% |
5 |
4.5 |
100 |
4.5 |
ENG_M.wav |
Original
Fs = 8 kHz
(128000 bps) |
| 100% |
5 |
4.5 |
100 |
4.5 |
ENG_F.wav |
| 101,15% |
1,55 |
2,61 |
39,45 |
2,666 |
eng_m1.wav |
MELP
2400 bps
(22.5 msec) |
| 84,02% |
3,66 |
2,91 |
70,35 |
2,445 |
eng_f1.wav |
| 110,42% |
1,52 |
2,59 |
38,47 |
2,86 |
eng_m2.wav |
MELPe
2400 bps
(22.5 msec) |
| 89,69% |
3,0 |
2,88 |
62,36 |
2,583 |
eng_f2.wav |
| 103,56% |
1,61 |
2,33 |
24,88 |
2,413 |
eng_m3.wav |
MELPe
1200 bps
(67.5 msec) |
| 83,56% |
2,08 |
2,78 |
50,68 |
2,323 |
eng_f3.wav |
| 101,14% |
3,2 |
2,89 |
64,77 |
2,923 |
eng_m4.wav |
MELPe-Plus
2400 bps
(22.5 msec) |
| 93,89% |
2,96 |
2,88 |
61,86 |
2,704 |
eng_f4.wav |
| 101,65% |
3,54 |
2,91 |
68,79 |
2,958 |
eng_m5.wav |
MELPe-Plus
2700 bps
(20 msec) |
| 94,93% |
2,98 |
2,88 |
62,09 |
2,734 |
eng_f5.wav |
| 112,23% |
3,56 |
2,91 |
69,08 |
3,266 |
eng_m6.wav |
MELPe-Plus
4000 bps
(20 msec) |
| 107,62% |
2,54 |
2,86 |
56,89 |
3,078 |
eng_f6.wav |
| 125,26% |
2,45 |
2,85 |
55,7 |
3,57 |
eng_m7.wav |
MELPe-Plus17
600 bps
(30 msec) |
| 116,19% |
2,26 |
2,81 |
53,23 |
3,265 |
eng_f7.wav |
Fig. 1. Voice quality comparison for English language
Table 3
| PESQ/Sevana, % |
Sevana MOS Score, % |
Sevana PESQ Score |
Sevana Quality Score, % |
(ITU) PESQ Score |
Chinese Language |
Sample |
| 100% |
5 |
4.5 |
100 |
4.5 |
ch_m.wav |
Original
Fs = 8 kHz
(128000 bps) |
| 100% |
5 |
4.5 |
100 |
4.5 |
ch_f.wav |
| 109,92% |
1,48 |
2,53 |
34,22 |
2,781 |
ch_m1.wav |
MELP
2400 bps
(22.5 msec) |
| 88,69% |
3,29 |
2,9 |
65,75 |
2,572 |
ch_f1.wav |
| 123,69% |
1,48 |
2,49 |
31,86 |
3,08 |
ch_m2.wav |
MELPe
2400 bps
(22.5 msec) |
| 97,84% |
2,32 |
2,83 |
54,82 |
2,769 |
ch_f2.wav |
| 126,76% |
1,72 |
2,16 |
20,83 |
2,738 |
ch_m3.wav |
MELPe
1200 bps
(67.5 msec) |
| 86,01% |
3,01 |
2,88 |
62,5 |
2,477 |
ch_f3.wav |
| 107,96% |
3,16 |
2,89 |
64,24 |
3,12 |
ch_m4.wav |
MELPe-Plus
2400 bps
(22.5 msec) |
| 95,10% |
2,98 |
2,88 |
62,17 |
2,739 |
ch_f4.wav |
| 108,10% |
3,13/TD>
| 2,89 |
63,89 |
3,124 |
ch_m5.wav |
MELPe-Plus
2700 bps
(20 msec) |
| 97,53% |
3,01 |
2,88 |
62,48 |
2,809 |
ch_f5.wav |
| 119,52% |
3,74 |
2,92 |
71,35 |
3,49 |
ch_m6.wav |
MELPe-Plus
4000 bps
(20 msec) |
| 109,86% |
3,01 |
2,88 |
62,52 |
3,164 |
ch_f6.wav |
| 130,42% |
2,5 |
2,86 |
56,4 |
3,73 |
ch_m7.wav |
MELPe-Plus17
600 bps
(30 msec) |
| 127,01% |
2,42 |
2,84 |
55,38 |
3,607 |
ch_f7.wav |
Fig. 2. Voice quality comparison for Chinese language
|
