Blog / VoIP & Video

What Jitter Does to Your Voice and Video Calls

January 28, 2026 6 min read VoIP · Video Calls · Jitter

You've been on that call. The audio cuts out mid-sentence, comes back robotic for a moment, then someone freezes on screen. Everyone says "can you hear me?" Nobody knows whose connection is bad. Most of the time, the answer isn't low bandwidth or bad ping — it's jitter. Here's exactly why it happens and what to do about it.

Quick Answer

Voice and video calls use real-time protocols (UDP) that cannot retransmit lost or late packets. When jitter causes packets to arrive unevenly, the application must either play silence, repeat audio frames, or freeze video. The codec's jitter buffer absorbs small amounts of jitter — but when jitter exceeds the buffer size, quality degrades instantly. Under 15ms jitter is the target for clean calls.

The Symptoms — and What's Actually Causing Them

Every call quality problem has a specific technical cause. Understanding which symptom maps to which failure mode helps you diagnose and fix the right thing.

🤖
Robotic / Metallic Voice
Jitter buffer overflow
Packets arrive in bursts after the jitter buffer window. The codec fills missing frames with synthesized audio — the characteristic "robot voice" artifact. Most noticeable during sustained high jitter.
✂️
Choppy / Cut-Out Audio
Packet loss + late packets
Packets that arrive too late to be played in sequence are discarded — treated as lost. The codec has no audio data for those frames, producing the clipped, cut-out effect. Frequent short gaps rather than silence.
⏸️
Frozen Video / Frame Drops
Video keyframe loss
Video codecs (VP9, H.264) rely on keyframes to reconstruct the image. When a keyframe packet is lost or arrives late due to jitter, the decoder can't update the frame — the video freezes until the next keyframe arrives.
🔇
Full Audio Dropout
Jitter spike exceeds buffer
A sudden jitter spike — common on Wi-Fi when a nearby device starts transmitting — can empty the jitter buffer entirely, producing 200–500ms of complete silence before the buffer refills.
🔁
Echo / Repeated Audio
Out-of-order packets
Packets that arrive out of sequence after their correct playback point get replayed when the jitter buffer flushes. This creates a brief echo effect or stutter where a word appears to repeat.
📈
Increasing Delay During Call
Adaptive buffer growth
Adaptive jitter buffers grow in response to sustained jitter. As the buffer grows to absorb late packets, the overall call latency increases — you notice a growing conversational delay even as audio quality improves.

The Jitter Buffer: Your Call's Last Line of Defense

Every VoIP and video call application uses a jitter buffer — a small memory queue that holds incoming audio packets for a short time before playing them out. The idea is simple: instead of playing packets the instant they arrive (which would produce choppy output if they arrive unevenly), the buffer collects a few packets first and plays them back at a smooth, fixed rate.

// How the Jitter Buffer Works
Low jitter
1
2
3
4
5
6
7
8
Buffer absorbs variation ✓
Med jitter
1
3
2
4
6
5
7
8
Reordering absorbed ✓
High jitter
1
?
?
×
2
3
4
5
Buffer overflow → artifacts ✗

The jitter buffer has a fixed or adaptive size, measured in milliseconds. A 50ms jitter buffer can absorb up to 50ms of delivery variation without any perceptible degradation. But there's a tradeoff: the larger the buffer, the higher the call latency. A 200ms jitter buffer produces a 200ms delay in the call — a noticeable conversational pause.

Most modern VoIP applications use adaptive jitter buffers that automatically grow and shrink based on observed network conditions. When your jitter increases, the buffer grows to compensate — absorbing the variation but adding latency. You may notice calls becoming slightly more "delayed" when your connection is under stress, even as the audio quality holds up. This is the buffer doing its job.

When jitter exceeds even the adaptive buffer's capacity — which happens during sustained Wi-Fi interference or ISP congestion events — the buffer overflows and artifacts appear.

Jitter Thresholds for Call Quality

JitterAudio QualityVideo QualityUser Experience
< 5ms Perfect Perfect Indistinguishable from being in the same room. No artifacts under any conditions.
5–15ms Excellent Excellent All applications function flawlessly. Buffer handles the variation with no perceptible impact on quality or latency.
15–30ms Good Good Occasional very brief audio glitches on sensitive codecs (G.711). Most modern apps (Zoom, Teams) handle this transparently. No video impact.
30–50ms Acceptable Acceptable Noticeably longer buffer delay adds conversational latency. Occasional choppy audio. Video may freeze briefly during spikes.
50–100ms Poor Poor Frequent audio artifacts. Robotic voice common. Video freezes regularly. Calls are frustrating but technically possible.
> 100ms Unusable Unusable Calls effectively non-functional. Audio is largely unintelligible. Video is a slideshow. Even consumer-grade apps can't compensate.

How Zoom, Teams, and Google Meet Handle Jitter Differently

Not all video call applications respond to jitter the same way. They differ in codec choice, jitter buffer implementation, and how aggressively they degrade quality to maintain stability.

AppAudio CodecJitter BufferDegradation StrategyJitter Tolerance
Zoom Opus Adaptive (up to 400ms) Drops video resolution first, then frame rate, then video entirely before touching audio High
Microsoft Teams Opus / Siren Adaptive (up to 200ms) Similar to Zoom — preserves audio at cost of video quality. Uses forward error correction on audio stream High
Google Meet Opus Adaptive Aggressive quality degradation — reduces bitrate early to keep the stream stable rather than waiting for artifacts Very High
WhatsApp / FaceTime Opus / AAC-ELD Fixed + adaptive Prioritizes low latency over resilience. Degrades faster but also recovers faster Medium
Traditional VoIP (SIP/G.711) G.711 / G.729 Usually fixed (20–60ms) No degradation strategy — either works or produces artifacts. G.711 is extremely intolerant of jitter above buffer size Low
Discord Opus Adaptive Very aggressive jitter compensation. Optimized for variable home network conditions Very High
Why Opus Changed Everything

The Opus codec (used by Zoom, Teams, Discord, WhatsApp) includes built-in packet loss concealment (PLC) and forward error correction (FEC). FEC sends redundant data so the receiver can reconstruct lost packets without retransmission. This makes Opus-based apps significantly more resilient to jitter than older codecs like G.711, which is why modern consumer apps feel more stable than enterprise VoIP systems on the same network.

Is Jitter Actually Your Problem?

Before fixing anything, confirm jitter is the issue. The symptoms of high jitter overlap with symptoms of other problems (low bandwidth, high ping, packet loss) — and the fixes are different for each.

Run a jitter test on jitter.is during or immediately after a bad call. A value above 20ms confirms jitter is contributing to the problem.

Fixing Call Quality: In Order of Impact

01

Switch to Ethernet

Wi-Fi is the dominant cause of call jitter. A wired connection eliminates radio interference and contention, typically reducing jitter from 15–40ms down to under 2ms. This single change fixes most call quality problems instantly. If you can't run a cable, a MoCA adapter over coaxial wiring delivers near-Ethernet performance.

02

Pause background bandwidth usage

Cloud backups, software updates, and 4K streaming on other devices all compete with your call. Upload saturation is particularly damaging — most home connections have asymmetric bandwidth, and a full upload pipe introduces queuing jitter even on your outbound audio stream. Pause background activity before important calls.

03

Enable QoS and prioritize voice traffic

On routers that support QoS (Quality of Service), enable traffic prioritization and set voice/video conferencing applications to highest priority. Better routers let you prioritize by DSCP markings — Zoom, Teams, and Meet all mark their audio streams with EF (Expedited Forwarding) DSCP markings. Enabling DSCP-based QoS ensures those packets always jump to the front of the queue.

04

Use the app's built-in network settings

Most video call apps have settings that can compensate for poor connections. In Zoom: Settings → Video → disable HD video during poor connection. In Teams: Settings → Devices → reduce video bandwidth. Reducing video quality frees bandwidth headroom for the audio stream, which is far more important for call intelligibility.

05

Move closer to your router (or change band)

If Ethernet isn't an option, signal strength directly affects Wi-Fi jitter. Connecting on 5GHz instead of 2.4GHz, moving within direct line of sight of the router, or using a Wi-Fi 6E device on the 6GHz band all meaningfully reduce wireless jitter. Avoid calls from rooms where Wi-Fi signal is weak.

06

Restart your router before important calls

Router memory fragmentation and buffer bloat accumulate over time. A cold restart clears routing tables, resets buffers, and often resolves jitter that has gradually crept up over weeks of continuous uptime. Make it a habit before calls that matter.

For Business / Professional VoIP

If you're running a SIP-based business phone system, the jitter thresholds are tighter than consumer apps because G.711 has no error correction. Target under 10ms jitter and under 1% packet loss. Dedicated VoIP routers (with hardware QoS) and MPLS or SD-WAN connections are worth the investment if call quality is business-critical. Consumer-grade internet with high jitter will produce unacceptable quality on SIP despite looking fine on speed tests.

Measure your jitter before the next important call — takes 30 seconds, no signup.

// TEST MY JITTER

Summary

Jitter disrupts voice and video calls by causing audio and video packets to arrive unevenly. The application's jitter buffer absorbs small variations, but when jitter exceeds the buffer capacity, artifacts appear: robotic voice, choppy audio, frozen video. Under 15ms jitter is the target for clean calls; above 50ms, most applications struggle to compensate.

Modern apps like Zoom, Teams, and Discord are significantly more resilient than older VoIP systems thanks to the Opus codec's packet loss concealment and forward error correction. But they're not magic — persistent jitter above 30ms will degrade even the best codec.

The fix is almost always the same: get off Wi-Fi, pause background traffic, and enable QoS. In most cases, switching from Wi-Fi to Ethernet alone resolves the problem completely before any other changes are needed.