AQuA
AQuA is a proprietary technology of Estonian business entity Sevana Ou. It implements a proprietary perceptual model based on research of several recognized scientists that determined frequency bands specific for human perception of audio signals. AQuA main return value is percentage of audio quality similarity between reference and test sound signals. AQuA can work with audio from 8kHz, 16 bit, Mono up to 192kHz, 32 bit, Stereo and has lots of parameters to adjust system perceptual model for client specific test scenarios to achieve best performance and accuracy.
Latest AQuA version provides the following sound quality metrics:
- MOS score according to P.800 – reliable objective MOS scores from 1 to 5
- Percentage of sound quality similarity between reference and test audio
- Duration distortion
- Delay or advancing of audio signal activity
- Sound signal activity mistiming
- Signal spectrum damages
- Signal spectrum vibration
- Unidirectional signal spectrum analysis: amplification or attenuation
- Most affected signal spectrum: low, medium, high bands
- Average pitch
- Pitch deviation
- Pitch frequencies distribution distortion
- Quantization steps distribution distortion
- Amplitude clipping level
- RMS
- Sound signal: Minimum Energy, Maximum Energy, Average Energy
- Sound signal: Min. Sample, Max. Sample, Average Sample
AQuA waveform analysis technology has been recognized in voice quality assessment across various networks, both TDM and packet switched. AQuA algorithm does not depend on the language, can work with audio files of arbitrary (from 10 seconds to several hours) duration.
VoIPmonitor
VoIPmonitor is an open source network packet sniffer with commercial frontend for SIP RTP and RTCP VoIP protocols running on Linux. VoIPmonitor is designed to analyze quality of VoIP call based on network parameters – delay variation and packet loss according to ITU-T G.107 E-model which predicts quality on MOS scale. Calls with all relevant statistics are saved to MySQL or ODBC database. Optionally each call can be saved to pcap file with either only SIP protocol or SIP/RTP/RTCP/T.38/udptl protocols. VoIPmonitor can also decode speech and play it over or save it to disk as WAV. Supported codecs are G.711 alaw/ulaw and commercial plug-ins support G.729a G.723 iLBC Speex GSM SILK iSAC OPUS.
Integration benefits
AQuA wrapper allows substitute PESQ when using VoIPmonitor to extend possibilities for intrusive voice quality analysis with AQuA technology:
- Use of flexible AQuA algorithms for perceptual voice quality analysis (both narrowband and wideband)
- Filtering analyzed calls based on IP
- Filtering analyzed calls based on called party number
Integration
Integrating Sevana AQuA into VoIPMonitor is a sequence of the following actions:
- Request AQuA wrapper for VoIPmonitor at Sevana web site: http://sevana.biz/downloads/
- Unpack all files under /usr/local/sbin folder
- Make sure aqua-v (aqua-wb) rights are set to executable
- Adjust settings in aqua_wrapper.conf
- Set parameters at voipmonitor.conf
- mos_lqo_bin=aqua_wrapper.sh
- mos_lqo=yes
- mos_lqo_ref=<file>
- saveaudio=wav
- In the database one should add records with IP and mask or phone number in the following tables: filter_ip or/and tilter_telnum. Field mos_lqo has the following values:
- 0 – do not invoke mos_lqo_bin
- 1 – mos_lqo_bin processes RTP streams with specified originating IP
- 2 – mos_lqo_bin processes RTP streams with specified destination IP
- 3 – process incoming and outgoing RTP from specified IPs
Example:
INSERT INTO `filter_ip` (`id`, `ip`, `mask`, `direction`, `rtp`, `sip`, `register`, `graph`, `wav`, `skip`, `script`, `mos_lqo`, `hide_message`, `note`, `remove_at`) VALUES (1, 2887453184, 24, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, 3, NULL, NULL, NULL);
AQuA (Audio Quality Analyzer) integration with VoIPmonitor
No comments:
Post a Comment