Understanding Fax Reliability: What to Expect and Why Fax Behaves Differently

Overview

Faxing remains in use today primarily for legal and regulatory reasons. While modern fax services provide convenient web portals, applications, and APIs, the underlying technology that delivers a fax has changed very little over time.

This article explains:

• How we maximize fax deliverability
• How faxing actually works
• Why fax transmissions can sometimes fail
• Why retries are normal and expected
• Why modern fax services cannot guarantee perfect delivery
• Why fax continues to exist despite its limitations

Understanding these points helps set realistic expectations and reduces frustration when fax behavior does not resemble modern digital services like email or file transfer.

How we maximize fax deliverability

Fax was designed for the traditional phone network, where a call might cross many switches but each one carried the signal over a dedicated circuit with constant timing and no loss. Fax modems depend on that. VoIP networks carry the same call as packets, which can arrive late, out of order, or not at all, and each carrier along the way may handle fax-over-IP (T.38) differently or not support it. So a fax that works perfectly to one destination can fail to another for reasons outside anyone's direct control. Instead of relying on a single setting and hoping, our fax platform adapts to every route on every call.

Two kinds of fax calls, two strategies

Our platform handles faxes in two fundamentally different ways, and each needs its own protection.

• Faxes we send or receive ourselves. Our server is the fax machine. It speaks the fax protocol directly, so it sees exactly what happens on every call: negotiated speed, error correction, whether the far end accepted fax-over-IP, and per-page results. When something goes wrong, we know precisely what failed and can retry with smarter settings.
• Faxes we relay between two other parties (for example, a customer's phone system to a carrier). Our server sits in the middle, often translating live between audio-based fax on one side and fax-over-IP on the other. We never see the fax itself, only whether the call succeeded. The call also crosses two network segments instead of one, which doubles its exposure to network problems.

In short, faxes we handle directly fail informatively, so we react precisely. Relayed faxes fail opaquely, so we protect them up front. Each route's settings for the two modes are tracked and learned independently, so a fix for one never breaks the other.

Per-route compatibility rules. Every call is checked against a policy engine that decides how to transmit: whether to use T.38 (fax-over-IP), whether to use error correction, and what transmission speed to allow. Rules can target a destination, a sender, or a specific sender-to-destination pair, so one problem route never degrades service for anyone else. When rules conflict, the safer, more compatible option wins.

Automatic learning from failures. When a fax fails for a technical reason (not busy or no answer), the system adjusts its approach for that route, stepping down from fax-over-IP to audio fax, then disabling error correction, then lowering speed, until it finds what works. For relayed calls, where we can't see inside the fax, learning is based on call-level signals and kept separate from the direct-fax rules.

Self-healing. Learned workarounds don't stick forever. After several consecutive successes, or after 30 days, they're cleared and the route is retested, so faxes don't stay on slower settings after a carrier fixes their side.

Smart probing on unknown routes. For routes with no history, consecutive calls alternate between fax-over-IP and audio fax to discover which works, without committing permanently either way. This runs separately for direct and relayed calls.

Smarter retries. When a fax we're sending fails and shows signs of trouble, such as repeated renegotiation or the far end refusing fax-over-IP, each retry switches to more compatible settings. Permanent failures like an invalid number are recognized and not retried pointlessly.

Live protocol conversion. When one side of a relayed call speaks fax-over-IP and the other speaks audio fax, our servers convert between them in real time. Fax-over-IP is only offered toward carrier-grade upstreams; customer phone systems stay on standard audio, where it's most reliable.

Verifying what actually happened. For relayed calls, the system inspects each side of the call afterward to confirm what was actually negotiated, rather than assuming our requested settings took effect. When a call fails over between multiple carriers, each failure is attributed to the correct carrier. That accuracy feeds the learning above.

Network-level protection. Jitter buffers smooth out the packet delay and reordering that corrupt fax audio. This matters most on relayed calls, where buffers are normally switched off during the connection; we keep them running on both sides. Calls use uncompressed G.711 audio only, with timing that doesn't depend on packet arrival.

Page adaptation. When we send a fax to a machine that only supports lower resolution or narrower pages, the document is resized to fit instead of the call failing.

Engineer override. For unusual cases, our engineers can apply targeted, time-limited fixes to individual routes, preview exactly how any call will be handled before a change takes effect, and switch any automatic behavior on or off globally.

Bottom line: every fax starts on the best available path, the system learns from every failure and recovers on its own when conditions improve, relayed faxes are protected proactively because they can't be diagnosed mid-call, and engineers can step in for anything automation can't solve. When a fax still fails after all of this, the cause is almost always at the receiving end or within a downstream carrier's network.

How Faxing Works (In Simple Terms)

Although faxing may begin with uploading a document through a website or application, the document is not delivered digitally end-to-end.

Instead:

1. The document is converted into fax signaling data
2. That data is transmitted using fax protocols
3. The receiving fax machine or gateway reconstructs the document from those signals

Fax protocols were originally designed to transmit documents over traditional telephone lines using audio-based signaling. These protocols assume:

• Precise timing
• Stable signal quality
• Minimal interruptions

They were created at a time when communication systems were known to be complex and imperfect, and occasional failure was expected.

What a Fax Actually Looks Like “On the Wire”

Why Faxing Behaves Differently Than Modern Digital Services

Modern services like email and file transfer are designed to:

• Automatically retransmit lost data
• Correct errors without user involvement
• Hide network complexity from the user

Faxing does not work this way.

Fax relies on continuous signaling with limited error recovery. If conditions are not ideal at any point during transmission, the fax may:

• Fail entirely
• Require a retry
• Partially transmit
• Succeed inconsistently

This behavior is inherent to fax technology and is not unique to any specific provider.

Faxing Over Modern Networks

Today, nearly all fax transmissions—including those sent through our service—are delivered using Fax over IP (FoIP) over SIP-based networks. This is true not only for modern cloud fax services, but also for many traditional carriers and local telephone providers.

In practice, fax transmissions often traverse a combination of:

• Long-distance fiber networks
• Carrier interconnection points
• Microwave and wireless backhaul
• Cellular transport segments
• VoIP and SIP-based systems

Even when a fax appears to be sent to or from a “traditional fax line,” it is very common for that line to terminate on a SIP adapter (ATA) rather than a true, end-to-end analog copper circuit. In many regions, fully analog copper fax lines no longer exist in the way they did decades ago—they have been replaced with IP-based infrastructure behind the scenes.

What This Means for Reliability

Fax protocols were originally designed for continuous, analog circuits with predictable timing. Modern packet-based networks operate very differently.

As a result, fax signaling remains sensitive to factors such as:

• Network latency
• Packet loss
• Timing variation (jitter)
• Codec handling and negotiation
• Transitions between different carriers or network segments

Even small variations at any point in the transmission path can affect fax delivery.

Our Role as a Fax-over-IP Provider

As a Fax-over-IP provider, we work closely with:

• Network monitoring and quality tools
• Carrier partners
• Routing and compatibility controls

These measures allow us to:

• Detect and mitigate common fax issues
• Optimize routing and protocol behavior
• Improve compatibility with a wide range of fax equipment

This gives us more visibility and control than many providers in the fax ecosystem.

However, it is important to understand that:

• No provider has full control over the entire end-to-end fax path
• Conditions on external networks, destination equipment, or local carrier infrastructure can still affect delivery
• Even local telephone providers often rely on SIP-based adapters rather than true analog copper lines

Because of this, occasional fax retries may still be necessary—even in well-managed environments.

Setting the Right Expectations

Fax reliability today is generally high, but not absolute. Variability can occur due to factors outside the control of any single provider.

This behavior does not indicate a service issue. It reflects the reality of delivering a legacy, analog-based protocol across modern telecommunications networks.

Understanding this helps ensure fax is used appropriately and with realistic expectations in today’s environment.

What Modern Fax Services Improve (and What They Cannot)

Modern fax platforms offer many improvements compared to traditional fax machines, including:

• Automatic retries
• Improved protocol handling
• Better compatibility with diverse fax equipment
• Smarter gateway logic
• Web-based visibility and tracking

These measures significantly improve overall success rates.

However, they cannot eliminate all limitations of fax technology. Fax delivery still depends on tone-based signaling and end-to-end compatibility. Even small issues—such as signal distortion, timing mismatches, or physical line conditions at the destination—can cause failures.

Modern services add resilience, not perfection.

Real-World Engineering Evidence

The limitations of faxing are well understood by telecommunications engineers and operators and are consistently observed in real production environments.

A detailed discussion among professionals using FreeSWITCH—a widely used open-source telephony platform—provides practical, measured insight into fax reliability over modern networks.

In this real-world analysis, an experienced operator reviewed results from:

• 1,000+ individual fax transmissions
• 4,000+ total pages
• 500+ unique destination fax numbers

Observed Reliability Metrics

Key findings from the data:

• The average number of attempts per successful fax was 1.33, excluding busy signals→ This means a meaningful percentage of faxes required one or more retries to complete successfully.
• First-attempt success was common but not guaranteed, even in a well-engineered environment with retries and fallback logic.

Success Rates by Transmission Method

Different fax transmission methods showed significantly different success rates:

• T.38 with Error Correction Mode (ECM)– Used for ~75% of attempts– ~82% success rate
• ECM without T.38– ~46% success rate
• ECM with T.38 disabled– ~60% success rate
• Other fallback combinations– Typically 56–60% success, depending on conditions

These results show that fax reliability varies not only by destination, but also by protocol negotiation and network behavior.

Common Failure Causes

When fax transmissions failed, the most frequent causes were:

• ~40% — Call dropped prematurely
• ~25% — No response after sending a page
• ~11% — Invalid response received after sending a page
• Remaining cases involved timeouts, protocol mismatches, or unexpected signaling behavior

Notably, many failures occurred after a fax had already partially transmitted, reinforcing how sensitive faxing is to timing and signal stability.

What This Demonstrates

This real-world data confirms that:

• Faxing can be made more reliable, but not perfect
• Retries are a normal and expected part of successful fax delivery
• Even well-designed systems experience variability
• Failures are often caused by conditions outside the sender’s or provider’s control

These observations closely match fax behavior seen across the telecommunications industry and help explain why occasional retries remain necessary even with modern fax services.

Why Faxing Still Exists

Fax persists primarily because:

• Certain laws and regulations explicitly recognize fax as a valid transmission method
• Regulatory frameworks change slowly
• Many compliance-driven industries rely on long-standing workflows

In regions such as Europe and the UK, legal frameworks were updated to recognize secure digital delivery and electronic signatures, allowing fax usage to decline naturally.

In North America, many regulations still reference fax explicitly, which keeps it in active use despite more modern alternatives.

Why Expectations Have Changed

When fax technology was first adopted, users generally understood that:

• Long-distance communication involved many physical systems
• Transmission paths were complex
• Occasional failure was normal

Modern digital services now hide this complexity, creating an expectation of near-perfect reliability. Faxing does not benefit from the same design assumptions, which can make its behavior feel unexpected or outdated.

What to Expect When Sending a Fax

A normal fax experience may include:

• Occasional failures
• One or more retries
• Varying delivery times
• Different outcomes depending on the destination

These behaviors do not necessarily indicate a service issue. They reflect the inherent characteristics of fax communication.

Best Practices for Reliable Faxing

To improve success rates:

• Allow time for automatic retries
• Confirm receipt for critical documents
• Avoid peak congestion periods when possible
• Use alternative delivery methods when legally permitted

Our systems are designed to maximize compatibility and reliability, but some factors remain beyond any provider’s control.

Summary

Faxing is a legacy technology operating in a modern telecommunications environment. While modern services improve usability and reliability, they cannot remove all inherent limitations of fax protocols.

Occasional failures and retries are expected and normal. Fax behaves differently than modern digital communication because it was designed in a different era, under different assumptions.

Understanding this helps set realistic expectations and ensures fax is used appropriately within today’s workflows.

If you have questions or need assistance with a specific fax transmission, our support team is always available to help.