Microsoft Lync Server 2010 made significant enhancements to the Enterprise Voice workload. This includes improving voice quality in Microsoft Lync Server and its clients. This article briefly discusses these improvements.

Author: Taimoor Husain       

Publication date: April 2011

Product version: Microsoft Lync Server 2010, Microsoft Lync 2010

Voice quality in Microsoft Lync Server 2010 is improved due to changes in the following key areas. These changes range from technological advances to user experience and interaction.  

  • Endpoint enhancements
  • Call Admission Control
  • Network and QoS implications
  • Media Bypass
  • Conferencing
  • Codec enhancements
  • Monitoring Voice

Endpoint Enhancements

Microsoft Lync 2010 made several enhancements to improve the overall voice quality experience. Lync 2010 includes an Audio Test Service, as shown in Figure 1 below. The Audio Test Service allows the user to place a test call prior to making a call or joining a conference. This feature gives users feedback on the quality of the call, given current network conditions.

Figure 1. Audio Test Service in Lync 2010

While in a call, Lync also provides a visual indicator of network conditions, as shown in Figure 2 below. This indicator is similar to the signal bars on cell phones that associate voice quality degradation with network conditions. Actionable feedback to end users provides guidance on call quality and mitigation steps.

Figure 2. Visual indicator of network conditions

Clicking on the link in the alert provides a dialog box, as shown in Figure 3 below.

Figure 3. Feedback for poor network connections in Lync 2010

Table 1 below describes the events that generate visual feedback and the measurements/thresholds that trigger these events.

Table 1. Events, descriptions, and measurements/thresholds

Caused By

Event

Description

Measurements - Thresholds

Network

Network Send Quality

Packet loss and jitter on receive stream is severe and introducing distortion

Jitter: Good <20ms, Bad >30ms
Packet Loss: Good <3%, Bad >7%

Network Receive Quality

Concealed packet ratio on send stream is severe and introducing distortion

Concealed Packet Ratio: Good <2%, Bad >3%

Low Bandwidth

Available bandwidth is insufficient for acceptable voice/video experience

Dynamic based on codec

High Latency

Network latency is severe and preventing interactive communication

RTT: Good <300ms, Bad >500ms

Machine

Low CPU cycles

Insufficient CPU for processing current modalities and applications, causing audio distortion

Flag when audio encoding/decoding engine is not getting sufficient CPU cycles

Remote user

Low SNR

Poor capture quality from remote user; distortion from noise or user being too far from microphone 

Flag if participant in the conference has a noisy environment

Echo

Remote user's device or setup is causing echo beyond the ability of the system to compensate

Flag if remote user (or participant in a conference) has a device setup that is injecting echo into the call

Device

Echo

Device or setup is causing echo beyond the ability of the system to compensate

* Timestamp noise
* Dynamic and Adaptive NLP attenuation
* Post-AEC echo percentage
* Microphone clipping due to far-end signal

Howling

Audio feedback loop detected (caused by multiple endpoints sharing audio path)

Check for howling/screeching from other endpoints in the room

Capture Device Not Functioning

Microphone currently used is not functioning correctly, causing one-way audio issues

Check capture buffer status

Render Device Not

Functioning

Speaker currently used is not functioning correctly, causing one-way audio issues

Check render buffer status

Render Glitches

Severe glitches in audio rendering, causing distortion; can be caused by driver issues, deferred procedure call (DPC) storm (drivers), high CPU usage

Look for glitches after adaptive render buffer

Low SNR

Poor capture quality; distortion from noise or user being too far from microphone

Low SNR
High absolute noise level after AGC

Microphone Clipping

User's speech level is too high for the system to handle and is causing distortion

Microphone clipping during near end-only portions

Near End to Echo Ratio

User's speech is too low compared to the echo being captured, limits ability to interrupt a user

Near end-to-echo ratio

Speaker volume to high or too far from the  microphone

Half Duplex Mode

To prevent echo, system enters half duplex mode (dynamic switching between render and capture streams), which limits ability to interrupt a user

Flag the event when device is in "voice switch" mode

Multiple Audio Endpoints

Multiple audio endpoints detected in the same session, system compensates by reducing render volume

Detect conference join tone in the room

 

 

 

 

Lync 2010 significantly improves device handling. Users can now switch between devices during a call. Lync automatically selects the most capable device from the devices available on the machine, as shown in Figure 4 below.

Figure 4. Changing the audio device while in a call using Lync 2010

Call Admission Control

Lync Server 2010 now enforces Call Admission Control (CAC) to manage network utilization for audio and video.

Administrators can configure bandwidth constraints to define network links to prevent degradation of service due to oversaturating the network link. Lync Server 2010 uses policy servers to control the number of audio and video calls that can be established across network links. When a network link reaches saturation or the threshold, as defined by the administrator, the policy server reroutes the call over the PSTN or an alternate route.

Call Admission Control allows administrators and network engineers to control the amount of bandwidth utilized by audio and video traffic to ensure call quality for calls in progress.

Network and QoS Implications

Lync Server 2010 supports quality of service (QoS) by using Differentiated Services Code Point (DSCP). Audio traffic can be marked at the applications layer (both client and server) with a DSCP value that is differentiated at the network layer. It is recommended practice to mark the traffic with the DSCP value "ef" (expedited forwarding), as per IETF recommendations.

Audio traffic can be further differentiated by tuning the port range used for audio traffic on both clients and servers. Choose a port range that is exclusive from all other applications on the network.

After voice traffic is marked, the network treats this traffic differently to ensure QoS through LAN prioritization, priority queuing over the WAN, and so on.

Media Bypass

Lync Server 2010 introduces a feature called Media Bypass. Media Bypass allows Lync Server to select the optimum media path between capable endpoints. Lync 2010 can send media directly to a third party endpoint or call control server and "bypass" the Mediation Server. By negotiating media capabilities directly with the endpoint, the need for media to traverse through the Mediation Server (perhaps over a sub optimal network path) and be transcoded before it reaches its destination is eliminated.

Conferencing

Lync Server 2010 enhances the conferencing experience with G.722 codec support for Conference calls. G.722 is a wideband fixed bit rate codec that provides a high level of audio quality to the Conferencing Server (MCU). Endpoints joining conferences will stream audio using G.722 by default.

Codec Enhancements

Lync Server 2010 uses RTAudio to provide a superior audio quality experience. RTAudio is an adaptive codec that dynamically adjusts itself based on exiting network conditions. This can range from changing the bit rate, to reducing bandwidth consumption on saturated links, to introducing forward error correction and redundancy in lossy networks, to compensating for jitter.

The RTAudio stack is constantly evolving, and in Lync 2010, there is an improved AEC filter bank and Dynamic Non-Linear Processing to further improve audio quality by improving echo cancellation and  reducing noise.

Monitoring Improvements

Lync Monitoring Server provides additional data and reporting capabilities to monitor call quality and call detail records (CDR). This new data includes performance reports between sites and average network conditions. It also includes improved session diagnostics and better audio quality metrics to determine Mean Opinion Scores (MOS).

The monitoring dashboard, shown below in Figure 5 and reports provide a snapshot of the voice deployment. These reports are accessible through SQL Server Reporting Services (SSRS). Additional insight on utilization of Enterprise Voice helps administrators monitor the voice network proactively and diagnose issues.

Figure 5. Monitoring Server Dashboard for Lync Server 2010

Summary

Lync Server 2010 includes a number of improvements that affect voice quality. These improvements range from end user experience, to administration, architecture, and design. Together, these enhancements offer a substantial improvement to the overall Enterprise Voice experience.

Lync Server Resources

We Want to Hear from You

Keywords: voice quality, audio, codec