LoFreq-SETI Prepares a Search for Technology in Low-Frequency Radio

Dark Fires editorial artwork of Earth from orbit. Illustration, not an observation from the LoFreq-SETI search.

A new SETI project is preparing to examine a relatively neglected part of the radio spectrum. LoFreq-SETI, described in an October 2, 2026 preprint by Craig Taylor, Lillian Ekness and Chenoa Tremblay, adapts observations from the Long Wavelength Array for searches for possible technological signals below 100 megahertz.

The work reports a processing demonstration rather than a discovery. Eight events appeared in the comparison between stations, but none survived visual inspection as a plausible technosignature. The authors plan to extend the approach to observations of exoplanet systems.

An array built for long radio waves

The University of New Mexico’s LWA overview describes three stations in central New Mexico, operating between 10 and 88 megahertz. Instead of a single enormous dish, the project uses collections of antennas. Its proposed Swarm expansion would distribute stations across universities and colleges in the United States, combining widely separated observations through a technique called very long baseline interferometry.

That separation matters because an array’s ability to distinguish directions depends on how its measurements are combined. The observatory’s wider Swarm concept is therefore an instrumental development with uses extending beyond the search for extraterrestrial technology.

Turning recordings into testable candidates

LoFreq-SETI converts station recordings into a format accepted by the Breakthrough Listen Interesting Signal Search toolkit, or BLISS, then compares signals across observing locations. Its initial test used a limited selection of archival observations aimed at fast radio bursts. Those targets supplied a practical dataset for checking the machinery before a dedicated search.

The BLISS project’s public documentation describes software designed to find narrowband signals whose frequency changes over time. It uses graphics-processor acceleration for tasks including flagging, estimating noise, combining measurements and identifying hits. Its worked example uses recordings of Voyager 1, giving developers a known human spacecraft signal with which to demonstrate the workflow.

That example makes the terminology concrete: a search program can successfully detect technology without discovering anything extraterrestrial. A flagged hit is the beginning of an investigation into the transmitter and observing conditions.

A useful lesson from an earlier false alarm

The published investigation of the signal known as blc1 provides a cautionary precedent. It appeared in 2019 observations toward Proxima Centauri and initially had several interesting properties. Detailed analysis connected it to local, time-varying interference and to related signals at other frequencies.

The lesson was methodological: an apparently promising frequency drift and observing pattern can emerge from ordinary electronics. Searching for related features across the data can be more informative than concentrating on the most striking trace alone.

For LoFreq-SETI, the immediate achievement is a documented route from raw measurements to candidates that can be scrutinized. The next step is broader observing and validation. Its limited demonstration cannot establish how common technological transmitters are, but it makes a specific future search easier to carry out and evaluate.