Sevana PVQA – Passive Voice Quality Analyzer is the tool that will help you with this task.
It is a unique powerful tool to non-intrusively (without comparing with reference audio file) evaluate voice quality by waveform analysis. PVQA MOS is perhaps the most accurate compared to other tools that predict MOS score by single file waveform analysis. Tool performance and ability to work in real-time give companies an opportunity to develop their own voice quality testing solutions not being dependent on particular hardware and software. PVQA is available as a cross platform application or library for all main desktop and mobile operating systems. Keep an eye on call quality with

How well do you know your network?
Do you know that in a packet switched network you may not only rely on ITU-T G.107 E-Model MOS and R-Factor? There are lots of other issues that affect QoE of your subscribers besides typical metrics such as jitter, latency, packet loss, or RTT delay.
Learning about issues “inside” the call gives you a unique chance to detect the issues not necessarily connected with network KPIs.
Ask us how knowing about Silence, Audio gaps, Clipping, Echo and Noise will help you to increase callers satisfaction and prevent their complains.


AQuA perfectly suits IVR systems testing when one has a reference audio (original quality IVR prompt) to compare with the recorded audio. But how to validate IVR prompt quality if there is no reference to compare with? However, there is a possibility to validate IVR prompt quality even without comparison by using Sevana PVQA. It is just the same way as one analyzes voice quality of call recordings, because PVQA can forecast a quality score by waveform analysis of a single audio file. Contact us to learn more!

IVR prompt verification and IVR prompt test automation are quite common tasks, however, there is probably no particular tool to make it really easy. And this is because you might have not heard about Sevana AQuA.
General purpose of AQuA is to analyze quality of the test audio by comparing it with the reference of original quality. However, unlike other similar tools one can use AQuA for IVR prompt verification, because comparing different audio files will naturally result in low quality score. Using AQuA for IVR prompt verification you verify the prompt and test connection quality within the same test call.

Contact us to learn more on using AQuA for IVR prompt verification.

Do you want to compare reference and test audio to evaluate the quality in your own language, but don’t have a tool?
AQuA is an audio quality analyzer that gives you even more flexibility:
✔️Language Independence – AQuA can evaluate quality of audio in any language.
✔️Long and short calls support – analyzes audio shorter than 4 seconds and longer than 30 seconds.
✔️Works with Stereo Audio – test signals may be mono or stereo, narrow, wideband and super wideband.
✔️Allows parameters profiling – configurable perceptual model that best suits your test environment.

Are you into conversational analysis?
Do you have millions of call recordings that you like to range based on audio quality?
Sevana PVQA is the tool that can analyze hours of recorded audio in a couple of minutes.
Request a call with our team to learn more about call recordings quality analysis and other related technologies for audio quality assessment.

Sevana QualTest is a mobile testprobe GSM VoIP application that checks current network conditions and estimates voice quality in mobile and VoIP networks. The platform is designed for end-to-end and single-end call testing, as well as for gathering and analyzing call audio quality metrics. Measure network metrics for VoIP calls and use waveform analysis to correlate audio problems with network conditions.

QualTest mobile test probe is an application for Android-powered devices that can work in both VoIP and cellular networks. A powerful tool for end-to-end and single-ended call testing. Fully functional frontend allows easy set up of tests specifying calling and called parties, reference audio and devices. Reporting shows overall quality MOS scores throughout different time periods, successful and failed test calls and speech-to-text engine provides information on call audio contents. The system works perfectly with unrooted and rooted mobile phones and analysis can be performed as on mobile device itself as well as on the devices it is connected to (e.g. Raspberry Pi in case of regular unrooted phones).

Being a multipurpose tool for voice quality analysis PVQA Server can be the core of a quality analysis and monitoring system to detect and investigate QoE and QoS related issues. PVQA Server combines the force of audio waveform analysis using PVQA technology with continuous network monitoring:
🔎Real-time call quality monitoring
🔎Proactive maintenance based on non-intrusive waveform analysis
🔎Monitoring network and media quality simultaneously
🔎Matching network metrics (RTT, jitter, latency, packet loss) with payload metrics (noise, clipping, dead air, echo etc)

Voice quality is the term used to describe the clarity of your voice. It’s more than just the sound of your voice—it also includes things like background noise and interference, which can have an adverse effect on how well your voice is heard. This makes passive voice quality testing and analysis and important part of modern communications.

Sevana AQuA and PVQA libraries collect all the data related to impairments that may affect voice quality in real-time providing immediate alerting and storage of all metrics related to call audio degradation. Audio impairments one can map on network parameters (jitter, packet loss, RTT delay) that will form patterns related to certain routes, destinations, originations. Every call analyzed by the system creates data that one can study using bigdata analysis tools to predict network behavior and call quality.

Silent Call – one leg of the voice call has no speech
Echo – this is typically due to a blockage or mismatch, which results in the signals bouncing back from where they came. Additionally, the presence of echo effect in packet switched networks can be traced back to the functionality of the line. Where standard lines function with a delay of 10 milliseconds, packet switched networks can have up to 400 milliseconds of delay. As a result, the echo effect is much more noticeable.
Amplitude clipping – is typically a result of a misconfigured voice gateway on the voice path.
Dynamic clipping – notifies about possible clipping happening in another network as the audio waveform looks like amplitude clipping, however, it is not present at the moment.
VAD clipping – this impairment detects incorrect work of Voice Activity Detector (VAD). Detector finds edges of active and inactive fragments of the signal considering VAD worked too late (in the beginning of the speech) or too early (in the end of the speech).
Click – is a single energy spike in the spectrum of the audio, which may happen for different reasons.
DeadAir-01 – is a zero-signal level inside speech, a gap in speech that may be caused by different reasons.
DeadAir-00 – is a constant signal level inside speech and may have positive and negative value.
SNR – this means that noise level present in the call affects human perception and disturbs the caller.

Within impairments analysis one can not only detect impairments that degraded voice quality in real-time, but in case of post call analysis pinpoint actual time period when certain impairment took place (as it is specified on the picture, Poor intervals are marked with the same color the impairment detectors that triggered the problem, e.g. Echo)

Further more, big data analysis of reports generated by PVQA help to identify impairment patterns specific for certain problems in the networks, e.g.

Click + DeadAir-01 – high probability that packet loss affected the call audio
Click + VAD clipping – possible packet loss

Contact us to have full demo on PVQA analysis or PVQA Server functionality.