An OTDR trace is a graph that maps the health of a fiber optic cable from one end to the other. The horizontal axis shows distance (in meters or kilometers), and the vertical axis shows signal power in decibels (dB). Reading the trace means understanding what each bump, spike, and slope on that graph tells you about connectors, splices, bends, and breaks along the fiber. Once you know the visual patterns, you can pinpoint problems to within a few meters.
What the Basic Trace Looks Like
When you fire an OTDR, it sends a pulse of light down the fiber and measures what bounces back. The result is a line that slopes gradually downward from left to right. That downward slope is the fiber slowly absorbing light as the pulse travels farther. The steeper the slope, the more signal the fiber is losing per kilometer.
The slope is calibrated in dB/km and represents the fiber’s attenuation coefficient. For standard singlemode fiber, you’d expect roughly 0.35 dB/km at 1310 nm or about 0.22 dB/km at 1550 nm. If the slope is noticeably steeper than the spec for your fiber type, something is degrading the cable, whether that’s age, damage, or poor installation. Any sudden change in the slope’s angle points to a localized problem worth investigating.
At the far end of the fiber (or at a break), you’ll see the trace drop sharply to the noise floor. This is the point where the OTDR can no longer detect a return signal. Everything past that drop is just instrument noise.
Reflective Events: Spikes on the Trace
Connectors and mechanical splices show up as upward spikes on the trace. These spikes are caused by Fresnel reflection, which happens whenever light hits a glass-to-air boundary or a change in the material between two fiber ends. The spike rises above the backscatter line and then drops back down.
A connector produces a tall, prominent spike with a measurable loss. You’ll see the backscatter line resume at a slightly lower level after the spike, and that vertical gap is the connector’s insertion loss. A mechanical splice also shows a spike, but it’s typically smaller because the index-matching gel inside the splice reduces the reflection. If you see a large spike where you expected a small one, the connector or splice may be dirty, damaged, or poorly aligned.
The height of the spike tells you the reflectance, while the drop in the backscatter level after the spike tells you the loss. Both numbers matter. High reflectance can degrade signal quality in high-speed networks even if the insertion loss looks acceptable.
Non-Reflective Events: Drops Without Spikes
Fusion splices and tight bends show up as a sudden downward step in the trace with no upward spike at all. There’s no air gap at a fusion splice, so there’s nothing to reflect light back. You simply see the backscatter line drop to a lower level and continue sloping down from there.
A high-quality fusion splice may show as little as 0.02 to 0.05 dB of loss, which can be hard to spot on a zoomed-out trace. If you see a non-reflective drop larger than about 0.1 dB, it’s worth checking whether the splice was done properly or if the fiber is being stressed at that point. Macrobends (the fiber bent too sharply around a corner or crushed in a cable tray) also appear as non-reflective losses. These are especially common at patch panels and cable entry points.
How to Detect Macrobends
One of the most useful tricks in OTDR analysis is comparing traces at two different wavelengths. Singlemode fiber carries light in a slightly wider pattern at 1550 nm than at 1310 nm, which makes the longer wavelength far more sensitive to bends. If you test the same fiber at both wavelengths and a particular event shows roughly the same loss at each, it’s a normal splice or connector. If the loss at 1550 nm is more than 0.2 dB higher than at 1310 nm, you’re looking at a macrobend.
This comparison is the standard method for catching bends that might otherwise look like a mediocre splice. Many modern OTDRs can overlay dual-wavelength traces to make this comparison easy.
Understanding Dead Zones
Every reflective event temporarily blinds the OTDR’s detector. During this recovery period, the instrument can’t accurately detect or measure anything else. This creates two types of dead zones you need to account for.
The event dead zone is the minimum distance after a reflection before the OTDR can detect another event. If two connectors are closer together than the event dead zone, the second one will be invisible on the trace. The attenuation dead zone is longer: it’s the minimum distance needed before the OTDR can accurately measure the loss of a following event. Even if you can see a second event within the attenuation dead zone, you can’t trust the loss reading.
Dead zone distances vary by instrument and pulse width setting, but they’re always listed on the OTDR’s spec sheet. This is exactly why launch cables exist.
Why You Need a Launch Cable
The OTDR’s own output connector creates a massive reflective event right at the start of the trace. That reflection, combined with internal crosstalk in the instrument, means the first stretch of fiber is inside the dead zone and unreadable. If you connect the OTDR directly to the cable under test, you’ll never see the condition of the first connector.
A launch cable (sometimes called a pulse suppressor) is a known-good length of fiber that you connect between the OTDR and the cable under test. It gives the instrument time to settle after the initial pulse so the trace is clean by the time light reaches your first real connector. Most multimode launch cables are at least 150 meters long, while singlemode launch cables are often over 1 kilometer. Adding a receive cable at the far end lets you see the last connector as well, since the trace will otherwise just drop off a cliff at the fiber’s endpoint.
Reading a Trace Step by Step
Start at the left side of the trace. The first large spike is the OTDR’s own connector (or the junction between the launch cable and your cable under test, if you’re using one). Note the distance marker here as your starting reference point. Everything to the right of this spike is the fiber you care about.
Move along the trace and identify each event in sequence. For every spike or drop, note three things: the distance from the OTDR, whether it’s reflective or non-reflective, and the loss in dB. Most OTDRs have cursors you can place on either side of an event to read these values automatically. Compare each loss value to what you’d expect. A typical connector should show 0.3 to 0.5 dB of loss. A fusion splice should be well under 0.1 dB. Anything significantly higher deserves attention.
Between events, check the slope. It should be smooth and consistent. A sudden steepening of the slope between two known points suggests a problem in that cable section, possibly a stress point or a degraded segment. If the slope looks noisy or jagged near the end of the trace, you may be approaching the OTDR’s dynamic range limit. You want at least 2 to 3 dB of margin between the fiber’s backscatter level and the noise floor to trust your readings at the far end of a long run.
Finally, look at how the trace ends. A clean connector at the far end (with a receive cable) shows a normal reflective spike followed by the launch/receive cable’s slope. A break in the fiber shows a sharp reflective spike followed by an immediate drop to the noise floor. An open, unconnected fiber end also produces a strong reflection. If the trace simply fades into noise without a clear endpoint, the fiber may be longer than your OTDR’s range at the current pulse width setting.
Pulse Width and Range Tradeoffs
The pulse width setting controls how much light the OTDR sends into the fiber with each test pulse. A wider pulse carries more energy, which means the signal can travel farther before it falls into the noise. This gives you more range but worse resolution: wide pulses create larger dead zones, and closely spaced events may blur together.
A narrow pulse gives you sharper detail and shorter dead zones, making it easier to distinguish events that are close together. The tradeoff is reduced range, since the weaker pulse runs out of detectable energy sooner. For short cable runs in a building, use a narrow pulse. For long outside-plant runs spanning many kilometers, you’ll need a wider pulse to see the far end. Many technicians test the same fiber twice, once with a narrow pulse for detail near the OTDR and once with a wider pulse to reach the full length.
Spotting Ghost Reflections
Ghosts are false events that appear on the trace at locations where no real connector, splice, or fault exists. They’re caused by light bouncing back and forth between two highly reflective connectors. The pulse reflects off one connector, travels back to a closer one, reflects again, and returns to the OTDR a second time. The instrument interprets this double trip as a real event at roughly twice the distance of the original reflector.
You can identify ghosts by a few telltale signs. They usually appear at a distance that’s a neat multiple of the spacing between two real connectors. They show reflection but zero loss, since there’s no actual junction consuming light. And they disappear if you clean or replace the connectors causing the original high reflectance, or if you test from the opposite end of the fiber. If an event seems suspicious, testing from the far end is always a good way to confirm whether it’s real. A true event will appear at the same physical location regardless of which direction you test from; a ghost will shift position.