Current Location ：
Jitter, Packet Loss, and the Jitter Buffer: Voice Quality Under Stress
A caller says the line is choppy. Nothing in the phone system reports an error. Bandwidth graphs look fine. The problem is usually not how much data arrives but when it arrives, and the mechanism that compensates for that timing is one of the most misunderstood settings in voice engineering. This article explains jitter and loss in plain terms, describes what the jitter buffer does with them, and sets out why increasing the buffer can turn a choppy call into a delayed one without making it better.
Three Different Problems Wearing One Name
Callers describe everything as bad line. The underlying causes divide into three categories that need different fixes. Latency is total delay, and beyond roughly 150 milliseconds in one direction it starts to interrupt conversation regardless of clarity. Jitter is variation in delay, packet arriving early then late, which sounds like uneven or robotic speech. Loss is packets that never arrive, which sounds like clipped syllables or gaps. Measuring one and assuming the others are fine is the most common diagnostic error in this area.
| Condition | What the caller hears | First thing to check |
|---|---|---|
| High latency | Talk-over, long pauses, awkward turn-taking | Path length, satellite or mobile legs, transcoding |
| High jitter | Uneven speech, robotic or warbling audio | Queueing on the WAN link, congestion bursts |
| Packet loss | Clipped words, dropouts, missing syllables | Link errors, duplex mismatch, Wi-Fi retries |
| Buffer too small | Constant chop under load | Configured buffer versus measured jitter |
| Buffer too large | Noticeable delay, echo complaints | Buffer depth relative to actual jitter |
What the Jitter Buffer Is Doing
Packets do not arrive at even intervals. The jitter buffer holds them briefly so that playback can proceed at a constant rate, absorbing the variation. Its depth is a trade: deep enough to cover the worst burst you expect, shallow enough to keep delay acceptable. Most platforms adapt the depth automatically, growing when jitter rises and shrinking when the line settles. When someone fixes a choppy call by forcing a fixed large buffer, they trade chop for delay, and delay is often the more annoying of the two.
- Measure jitter over a full working day, not a five-minute test call
- Check for bursty congestion, which causes more jitter than steady high utilisation
- Confirm quality of service is applied on the congested link, not just configured
- Treat Wi-Fi as a source of jitter until proven otherwise, particularly with roaming
- Avoid forcing fixed buffer depths unless the platform documentation makes a specific case for it
Loss Is Not Always Network Loss
Packets can be lost on the wire, or discarded in the router when queues overflow, or dropped by the endpoint when it is overloaded. Loss that appears only during busy periods points to queueing. Loss that appears only on specific handsets points to the device or its power and link. Loss that appears on one direction only is usually a path asymmetry problem rather than a general network fault. Establishing which of these you have shortens the investigation considerably.
Start From the Path, Not the Platform
Before adjusting anything on the phone system, capture what the endpoint reports about the call. Modern handsets and softphones expose jitter, loss, and round-trip delay per call. If those figures are clean, the problem is elsewhere, possibly acoustic or in the handset itself. If they are poor, the platform settings are the wrong place to look.
Redstone Approach
Redstone voice deployments include quality of service planning on the links that carry media, and our support engagements start from endpoint-reported metrics rather than from assumptions about the cause. Where jitter buffer or codec settings are implicated, we change one variable at a time and compare measurements, so that the improvement can be attributed rather than guessed at.
Redstone Systems, Inc. founded in Delaware, USA in December 2002, has been the ODM vendor for many well-known communications companies, serving the Southeast Asian market. In 2020, Redstone Systems will begin to return to the North America market with its self-developed brand.
Redstone has a complete product line of intelligent voice gateways, providing IP-PBXs, analog VoIP gateways (FXS/FXO), digital VoIP gateways (E1/T1), border appliances, and session boundary controllers (SBCs).
With advanced technology in digital signal processor (DSP), speech coding and speech processing, as well as efficient operational tools such as cloud remote management, auto provisioning, Redstone gateways are widely used in markets of enterprise communications, cloud communications, call centers, operators’ IMS/SIP trunks, bringing users friendly, efficient and reliable communication experience.
For more information about Redstone, You can follow us on Facebook, Linkedin, and Youtube to be the first to get the latest news.
