What Jitter Actually Measures
Jitter is the variation in round-trip time between consecutive packets. If one packet takes 30 ms and the next takes 85 ms and the one after that takes 28 ms, your jitter is high — even though the average ping might only look like 47 ms.
This matters because real-time applications — voice calls, video conferencing, gaming — depend on packets arriving at a steady, predictable pace. A stable 60 ms ping is far easier for applications to compensate for than a ping that swings between 15 ms and 200 ms. The audio codec or game engine knows what to expect with consistent timing; erratic timing breaks the experience.
Jitter Benchmarks by Use Case
| Jitter | Rating | Impact |
|---|---|---|
| Under 5 ms | Excellent | Imperceptible in all applications |
| 5–15 ms | Good | Gaming, VoIP, and video calls all run smoothly |
| 15–30 ms | Acceptable | Minor occasional artifacts; most users do not notice |
| 30–50 ms | Noticeable | Video calls stutter; voice may sound choppy; gaming feels inconsistent |
| 50–100 ms | Poor | Calls are difficult; competitive gaming is compromised |
| Over 100 ms | Very Poor | VoIP unusable; gaming unplayable at a competitive level |
Why Jitter Hurts VoIP More Than High Ping
A voice call with 100 ms consistent ping sounds fine — both sides simply experience a slight delay, like talking on a walkie-talkie. Most people do not even notice it in normal conversation. But a voice call with 30 ms average ping and 80 ms jitter sounds terrible: audio drops out, words get cut off, the other person sounds like they are on a broken radio.
The reason is the jitter buffer. VoIP applications buffer incoming audio packets and play them back at a steady rate. When packets arrive in a tight window of variation, the buffer is small and delay stays low. When packets scatter unpredictably — some arriving 10 ms early, some 100 ms late — the buffer has to grow to accommodate the spread. If it cannot grow fast enough, packets get dropped. That is the choppy audio you hear.
The Jitter Buffer: How Apps Cope
Most real-time applications implement a jitter buffer — a small adaptive delay designed to smooth out arrival timing. Zoom, Teams, Discord, and game clients all use some form of this. The buffer waits for packets to arrive and reorders them before processing.
The problem is the tradeoff: a larger buffer means the app can handle more jitter, but it also means more added latency. When jitter is low, apps keep buffers small and overall delay stays low. When jitter is high, apps grow the buffer to compensate — and suddenly a "20 ms ping" call actually sounds like 80–150 ms of delay once buffering is accounted for. Jitter adds hidden latency that your speed test cannot directly show you.
What Causes High Jitter
- Wi-Fi interference. The most common cause by far. Wireless signals contend with neighbors' networks, microwave ovens, baby monitors, and the physics of multipath reflection. A device two rooms from the router on 2.4 GHz may show excellent average ping but terrible jitter. Ethernet eliminates this category of problem entirely.
- Bufferbloat. When a router's queues fill up during a download or upload, latency-sensitive packets (gaming, VoIP) sit behind bulk data packets waiting to be sent. The wait time is unpredictable — some packets slip through quickly, others wait in queue for tens or hundreds of milliseconds. This is bufferbloat, and it is the reason gaming feels awful during a Steam download even on a fast connection.
- Shared cable congestion. Cable internet shares bandwidth on a neighborhood segment. During peak hours, the segment becomes congested and packet arrival timing gets inconsistent. Fiber dedicated lines do not share the segment the same way and generally show lower jitter under load.
- Background application bursts. A cloud backup, Windows Update, or browser sync that sends data in irregular bursts can saturate the upload link briefly, spiking latency for anything else trying to send packets at the same time.
How to Test Your Jitter
Running a speed test at SpeedTestHQ reports jitter alongside download, upload, and ping. For a more detailed view, run continuous pings from your terminal: ping -t 8.8.8.8 on Windows, or ping 8.8.8.8 on Mac and Linux. Watch the round-trip times over 30–60 seconds. A stable connection shows variation under 5 ms. A problematic one shows swings of 50 ms or more, or occasional jumps to 200–500 ms.
For a proper bufferbloat test, run the ping while simultaneously doing a large download or upload. If jitter spikes dramatically under load but is fine at idle, bufferbloat is the culprit — a router problem, not an ISP problem.
How to Fix High Jitter
- Switch to Ethernet. Plugging in a cable instead of using Wi-Fi is the single highest-impact fix for most users. Wired connections reliably drop jitter to under 2 ms for local hops.
- Enable SQM or QoS on your router. Active Queue Management (SQM with CAKE or FQ-CoDel) prevents bufferbloat by controlling queue depth and scheduling latency-sensitive packets ahead of bulk data. Many modern routers support this, and it can reduce jitter from 50+ ms to under 5 ms under load.
- Reduce background traffic during calls or gaming. Pause cloud backups, software updates, and video uploads while on an important call or in a competitive match.
- Upgrade your router if it is old. Older consumer routers with limited CPU cannot process packets quickly enough under load, adding processing jitter on top of network jitter.
- Check for ISP-side issues. If jitter is high on Ethernet with no load, check whether it is consistent across different times of day. Evening spikes point to congestion; consistent high jitter at any hour suggests a signal or hardware problem worth reporting to your ISP.
Jitter vs Bufferbloat: Related but Different
Bufferbloat is a specific cause of jitter — not the same thing. Bufferbloat happens when router queues fill up and hold packets too long. Jitter is the symptom: packets arrive at unpredictable intervals. You can have jitter without bufferbloat (Wi-Fi interference is a common cause that has nothing to do with router queues). But bufferbloat always produces jitter, and fixing bufferbloat almost always dramatically reduces jitter under load.
Frequently Asked Questions
Is jitter the same as ping?
No. Ping is the average round-trip time. Jitter measures how much that time varies between consecutive packets. A connection with 40 ms ping and 2 ms jitter feels stable. One with 40 ms ping and 50 ms jitter produces constant lag spikes — even though the average looks identical.
What causes high jitter?
The most common causes: Wi-Fi interference (especially 2.4 GHz congestion), bufferbloat (router queues filling during downloads), shared cable segment congestion during peak hours, and background apps generating burst traffic. Test on Ethernet first — if jitter drops sharply, Wi-Fi is the culprit.
Can a VPN reduce jitter?
Sometimes. If your ISP routes traffic through inconsistent or congested paths, a VPN may find a more stable route. More often, a VPN adds consistent overhead without addressing the actual jitter source. Test with and without VPN and compare jitter readings — do not assume either way.
What jitter is acceptable for video calls?
Under 15 ms is comfortable for Zoom, Teams, or Google Meet. Applications use jitter buffers that absorb moderate variation, but above 30–40 ms the buffer cannot compensate fast enough and you get choppy audio or frozen video. Most residential connections on Ethernet fall well under 15 ms.
Does upgrading internet speed fix jitter?
Rarely. Jitter is almost never caused by insufficient bandwidth — it is caused by packet timing inconsistency. Switching from Wi-Fi to Ethernet, enabling SQM on your router, and avoiding heavy background downloads during calls are far more effective than buying a faster plan.
How do I test my jitter?
Run a speed test at SpeedTestHQ — it reports jitter alongside other metrics. For a more thorough look, run continuous pings from your terminal and watch for variation in the round-trip times. A stable connection shows variation under 5 ms. A problematic one swings 50 ms or more.