Why Your VoIP Calls Sound Bad — And How to Fix It

VoIP call quality problems are among the most complained-about issues in business technology — and among the most misdiagnosed. When calls sound choppy, echo, cut out, or drop entirely, the instinct is to blame the phone system or the provider. Often the problem is elsewhere entirely.

This article covers the most common causes of poor VoIP call quality, how to diagnose which one you are dealing with, and what the fixes actually look like. If your business uses VoIP and the experience is not what it should be, the answer is almost certainly in here.

How VoIP Works and Why Quality Is Fragile

Traditional phone calls travel over dedicated circuits — a physical connection reserved for that call. The quality is consistent because the bandwidth is guaranteed.

VoIP calls travel over your internet connection as data packets — the same connection your staff use for email, file downloads, video conferences, and cloud applications. These packets must arrive at the other end in the right order, at consistent intervals, and without significant delay.

When the network cannot deliver those guarantees, the call quality suffers in specific and recognisable ways. Understanding which symptom you are experiencing points directly to the cause.

The Four Quality Metrics That Matter

Latency is the time it takes for audio to travel from one end of the call to the other. Below 150 milliseconds one-way, most people cannot notice the delay. Above 200 milliseconds, conversations become awkward — both parties start talking at the same time because the audio delay disrupts natural conversational rhythm. Above 300 milliseconds, calls become genuinely difficult to conduct.

Jitter is variation in latency. A connection where latency is consistently 80 milliseconds is manageable. A connection where latency varies between 20 milliseconds and 300 milliseconds produces choppy, robotic-sounding audio — because the audio packets are arriving at irregular intervals and the phone system cannot reconstruct them smoothly.

Packet loss is the percentage of audio packets that do not arrive at all. Even 1% packet loss is audible on a VoIP call — you hear brief gaps in the audio. At 3-5%, calls become significantly degraded. Above 5%, calls are frequently unusable.

Bandwidth is the least important of the four metrics for call quality. A single VoIP call uses approximately 80-100 kilobits per second in each direction — a tiny fraction of any business internet connection. Unless you have dozens of simultaneous calls, bandwidth is rarely the cause of quality problems.

This is why upgrading your internet connection often does not fix VoIP quality problems. More bandwidth does not reduce latency, jitter, or packet loss — the three metrics that actually determine call quality.

The Most Common Causes

*Bufferbloat — the hidden culprit*

Bufferbloat is the single most common cause of VoIP quality problems on business connections — and the least understood. When your network equipment has large buffers and no active queue management, large data transfers (file downloads, video streaming, backup uploads) fill those buffers and create significant queuing delays for all other traffic, including VoIP packets.

The result: calls that sound fine when the network is quiet but deteriorate significantly during business hours when staff are actively using the connection. The internet speed test still shows good results — because the speed test fills the buffer and measures throughput, not queue latency. But every VoIP packet is waiting in line behind the file downloads.

The fix is active queue management — specifically algorithms like CAKE or fq_codel deployed on the router. These algorithms keep buffer queues short and prioritise latency-sensitive traffic, eliminating bufferbloat without requiring more bandwidth.

*Insufficient QoS (Quality of Service)*

Even without bufferbloat, VoIP traffic competes with other traffic on your network. Without QoS rules that prioritise voice packets, a large file upload from one staff member can degrade call quality for everyone else on the connection.

QoS marks VoIP traffic as high priority at the router level, ensuring that voice packets are processed before bulk data transfers. On a properly configured MikroTik router, QoS rules can guarantee that VoIP traffic always gets the bandwidth and latency it needs regardless of what else is happening on the connection.

*WiFi problems*

VoIP calls made over WiFi are inherently more vulnerable to quality problems than calls made over wired connections. WiFi is a shared medium — all devices on the same access point share the available bandwidth and compete for airtime. In environments with many connected devices, WiFi congestion causes jitter and packet loss that directly affects call quality.

Additional WiFi-specific problems include: roaming events (the phone switching between access points mid-call, causing a brief audio interruption), interference from neighbouring networks on the same channel, and access points running older WiFi standards that handle voice traffic less efficiently.

For staff who make frequent or important calls, a wired connection or a properly planned enterprise WiFi deployment is significantly more reliable than a consumer WiFi setup.

*Inadequate internet connection*

While bandwidth is rarely the primary cause, some internet connections are simply not suitable for VoIP. Home broadband connections sold for residential use are typically asymmetric — much more download bandwidth than upload — and may have high latency to international exchange points. A dedicated business internet connection (DIA) with symmetric bandwidth and guaranteed service levels is a more appropriate foundation for business VoIP.

*Network congestion at peak hours*

Business internet connections sold on a shared or contended basis share capacity with other customers on the same provider infrastructure. During peak hours — typically 9am-12pm and 2pm-5pm on business days — congestion on the provider's network can increase latency and packet loss, degrading VoIP quality across the board.

If your calls are consistently worse at specific times of day, provider-side congestion is a likely cause. A dedicated internet connection with no contention eliminates this variable.

*SIP ALG interference*

Many consumer and small-business routers include a feature called SIP ALG (Application Layer Gateway), intended to help VoIP traffic pass through NAT (Network Address Translation). In practice, SIP ALG often causes more problems than it solves — corrupting SIP packets, causing one-way audio, and dropping calls randomly.

If you are experiencing one-way audio (you can hear the other person but they cannot hear you, or vice versa), or calls that connect but immediately drop, SIP ALG is a likely cause. The fix is to disable it on the router.

How to Diagnose Your Problem

*Step 1: Test from a wired connection*

If your VoIP phones or softphones are on WiFi, test the same call from a wired connection. If quality improves significantly, the problem is WiFi-related. If quality is the same, the problem is elsewhere.

*Step 2: Test during off-peak hours*

Make test calls at 7am and at 11am. If quality is noticeably better during off-peak hours, the problem is congestion — either on your local network during business hours or on the provider's network.

*Step 3: Run a VoIP quality test*

Tools like VoIP Spear test your connection specifically for VoIP suitability — measuring latency, jitter, and packet loss under simulated call conditions and reporting a MOS (Mean Opinion Score). A score below 3.5 indicates noticeable quality problems. Run this test during business hours when the network is under real load.

*Step 4: Check for bufferbloat*

Run a bufferbloat test at dslreports.com/speedtest (or a similar tool that tests for bufferbloat specifically). A grade of C or below indicates significant bufferbloat that will affect VoIP quality. An A grade indicates the router is managing queues well.

*Step 5: Check SIP ALG*

Log into your router configuration and search for SIP ALG. If it is enabled, disable it and test again.

What IJA Deploys for VoIP Customers

Every IJA VoIP deployment includes network quality assessment before any phone system is configured. We test latency, jitter, and packet loss on the existing connection and identify any issues before they become call quality complaints.

On MikroTik routers — our standard deployment platform — we configure CAKE active queue management to eliminate bufferbloat, QoS rules to prioritise SIP and RTP traffic, and SIP ALG is disabled by default.

For WiFi deployments, we ensure VoIP-capable access points with proper band steering and roaming configuration, minimising mid-call roaming events.

The 3CX phone system we deploy includes built-in call quality monitoring — every call is scored, and quality problems are visible in the system dashboard before they become user complaints.

The Right Foundation for VoIP

VoIP call quality is a network problem as much as a phone system problem. A premium phone system on a poorly configured network will sound worse than a basic system on a properly managed connection.

The investment in a proper network foundation — active queue management, QoS, enterprise WiFi, and a dedicated internet connection — pays for itself in staff productivity and customer experience. Calls that work reliably, sound clear, and do not drop are not a luxury. They are a basic operational requirement.

IJA Technologies deploys and manages 3CX VoIP systems on properly configured network infrastructure. If your VoIP calls are not performing as they should, talk to us.