Access Official Technical PDF

Share Summary:LinkedInEmail

Noise-like pulses are an unusual operating regime in a passively mode-locked fiber laser. Their overall envelope and average spectral properties can remain comparatively stable, even though the pulse is made up of many ultrashort components whose positions, durations and amplitudes fluctuate. Understanding both scales matters when evaluating fiber lasers for applications such as optical sensing, supercontinuum generation and micromachining.

Experimental characterization of two figure-eight fiber lasers illustrates how cavity dispersion and birefringence affect these pulses. The lasers, approximately 300 m and 225 m long, operated in normal and anomalous dispersion regimes, respectively. Their results also show why averaged measurements alone can miss important details of pulse behavior.

What makes noise-like pulses different?

A conventional ultrashort laser pulse is generally described by a coherent waveform with a defined duration and shape. A noise-like pulse, by contrast, is a longer temporal packet containing a dense collection of much shorter subpulses. The packet may extend from picoseconds toward nanosecond or longer timescales, while its internal components can be shorter than a picosecond.

At the local scale, these subpulses vary from one another and can change from one cavity round trip to the next. At the global scale, however, the packet can show a relatively stable envelope, energy and average spectrum. Its optical spectrum is often broad and smooth, while its autocorrelation trace typically combines a narrow central peak with a much wider pedestal. The central feature reflects the short internal timescale; the pedestal indicates the duration of the overall packet.

This combination is one reason noise-like pulses are sometimes described as a state of partial mode locking. Their internal structure is not equivalent to a single, uniformly shaped pulse, even when standard instruments display a smooth averaged profile.

Noise-like pulses in two figure-eight fiber lasers

The experimental study compared two passively mode-locked figure-eight fiber lasers: one about 300 m long in normal dispersion and another about 225 m long in anomalous dispersion. In both systems, adjustments to cavity birefringence produced changes in the observed pulse behavior. Birefringence describes how a fiber’s optical properties differ with polarization, and changing it can alter the conditions experienced by light circulating in the cavity.

Normal dispersion: varied envelopes and internal structure

In the approximately 300 m laser, adjustments to birefringence produced noise-like pulses with different temporal profiles and durations. For some settings, the measured optical spectrum reached a bandwidth of about 130 nm at a 10 dB level. This is a reported experimental result for that setup, not a general specification for figure-eight lasers.

To examine the packet beyond its averaged envelope, the researchers used two optical components as measurement filters. A birefringent Sagnac interferometer selected spectral components, while a nonlinear optical loop mirror (NOLM) acted as an intensity filter. Comparing the filtered signals indicated that different spectral components were distributed throughout the temporal envelope. In contrast, the amplitude and density of the short subpulses varied substantially across that envelope.

Some observations suggested an intermediate organizational scale: structure between the full nanosecond-scale packet and its subpicosecond constituents. This is important for characterization because a two-scale description may not capture every feature present in the pulse.

Anomalous dispersion: fragments that separate and return

In the approximately 225 m laser, birefringence adjustments could cause smaller packets to emerge from the main noise-like pulse. Two representative behaviors were observed. In one, released fragments moved away from the main packet and decayed after traveling through part of the cavity. In the other, sub-packets continued around the cavity without decaying and eventually merged back into the main waveform.

The fragments and the main packet were reported to consist of units with sub-nanosecond durations and roughly similar energies. These observations show that a noise-like pulse can exhibit dynamics beyond the changing positions and amplitudes of its shortest internal subpulses: larger fragments can also form and move through the cavity.

Why measurement bandwidth and averaging matter

Standard instruments commonly average measurements over time or across repeated pulses. That averaging helps reveal the overall temporal envelope and smooth spectral profile, but it can conceal rapid fluctuations and fine internal structure. The observed signal also depends on the detection bandwidth: a measurement system cannot resolve temporal features that are faster than its response permits.

For engineers and researchers characterizing a fiber laser, this means instrument settings and measurement methods must be considered when interpreting results. Useful complementary methods include:

  • Optical spectrum measurements to assess the average spectral shape and bandwidth.
  • Autocorrelation to distinguish the short central feature from the broader packet-scale pedestal.
  • Fast temporal detection to observe changes and fragment motion across cavity round trips, within the limits of the detector and oscilloscope.
  • Spectral and intensity filtering to investigate how components are distributed within the pulse envelope.

No single measurement provides a complete picture. In particular, an averaged trace should not be treated as evidence that the pulse’s internal structure is itself smooth or stationary.

Practical implications

Noise-like pulses offer a distinctive combination of broad, smooth average spectra and complex temporal structure. The two laser experiments show that dispersion regime and cavity birefringence are closely connected to the forms and dynamics observed, while specialized filtering can reveal organization hidden by conventional averaging.

For teams assessing fiber laser sources for sensing, nonlinear frequency conversion, supercontinuum generation or micromachining, the practical takeaway is to match characterization methods to the timescales of interest. Measure both the packet-level behavior and its faster internal dynamics, and interpret results in light of the system’s detection bandwidth. That approach provides a more useful basis for comparing operating regimes and understanding what a noise-like pulse is doing inside the cavity.

Access Official Technical PDF

About The Author