GeoShake vs Raspberry Shake: Cost, Accuracy & Use Case Comparison

If you're looking for a home earthquake sensor, you've probably come across two names: GeoShake and Raspberry Shake. They both detect earthquakes. They both involve community networks. But they're fundamentally different devices designed for different purposes — and they sit at very different price points.

This guide breaks down exactly how they compare so you can pick the right one for your needs — whether you're a homeowner who wants earthquake alerts, a maker building a DIY seismic station, or a citizen scientist contributing to seismological research.

TL;DR — Quick Comparison

What Are These Devices?

GeoShake T1

GeoShake T1 is an open-source earthquake early warning sensor. Its core mission is speed: detect a seismic event, quantify its intensity using Peak Ground Acceleration (PGA), and alert the network in milliseconds. Built around an ESP32-S3 microcontroller with four LSM6DSO MEMS accelerometers sampling at 104 Hz, it transmits PGA data via MQTT — meaning the alert travels at the speed of light, not the speed of seismic waves.

It's designed for anyone who wants reliable, real-time earthquake detection at home without paying research-grade prices. No Raspberry Pi required, no complex configuration, no subscription fees.

Raspberry Shake

Raspberry Shake is a family of personal seismographs built on the Raspberry Pi platform. It comes in multiple models — from the basic 1D (single geophone) to the 4D (geophone + 3-axis accelerometer). Raspberry Shake records full seismic waveforms in standard miniSEED format, making the data directly usable by professional seismologists and research institutions.

It's designed for people who want to study earthquakes in detail — not just detect them, but analyze the waveforms, measure frequencies, and contribute research-grade data to the global seismological community.

Head-to-Head Specs

Cost Breakdown

The price difference is meaningful — but the story is more nuanced than "cheap vs expensive." Both devices are in premium territory; GeoShake just sits at the accessible end of that range.

Neither device charges ongoing subscription fees. Both use WiFi for data transmission. The key difference in total cost is purely in the hardware — GeoShake is roughly 60% cheaper than the entry-level Raspberry Shake 1D and about 80% cheaper than the Raspberry Shake 3D — comparing turnkey to turnkey — while still delivering a premium 4-sensor MEMS array, 104 Hz sampling, and MQTT alerting out of the box.

A note for academic and institutional buyers: GeoShake offers dedicated pricing for schools, universities, research labs, and bulk community deployments. Reach out via our contact form for a quote tailored to your project.

Detection Accuracy & Sensitivity

Left — 4× MEMS accelerometer array
GeoShake T1 · proof masses on silicon springs

Right — geophone
Raspberry Shake · coil, magnet and spring

This is where the comparison gets nuanced, because these devices measure different things in different ways.

GeoShake: Speed Over Detail

GeoShake doesn't try to record the full earthquake waveform. Instead, it uses the STA/LTA algorithm to detect sudden changes in ground acceleration. When it detects an event, it sends the Peak Ground Acceleration (PGA) value — a single number that tells you how intense the shaking is, measured in units of gravity (g).

The 104 Hz sample rate is fast enough to capture the frequency range of damaging seismic waves (typically 0.1–20 Hz).

Why four sensors instead of one

A single MEMS accelerometer is a noisy instrument. Its output carries the ground motion you want plus the chip's own electronic noise — and at the amplitudes a modest earthquake produces, that noise is the only thing standing between detection and silence.

The T1 carries four LSM6DSO accelerometers on one board and averages them before the detector sees anything. This works because of an asymmetry: each chip's electronic noise is independent, while a real ground motion moves all four identically. Averaging N sensors therefore suppresses the noise by √N while leaving the signal untouched — a factor of two for four sensors.

That is not a marketing figure, it is measurable. The LSM6DSO datasheet gives a noise density of roughly 110 µg/√Hz, which across the T1's measurement band predicts about 0.0078 m/s² of noise per axis for a single chip — and 0.0039 m/s² after quad averaging. Seven days of field data from three T1 stations, roughly 1.7 million one-second windows, measure 0.00385 m/s². Prediction and measurement agree to within about one percent.

In practical terms, halving the noise floor is worth roughly 0.3 magnitude units of extra reach at any given distance. An earthquake that would sit just below the trigger threshold on a single-sensor design clears it on the T1.

The economics are the point. Three extra MEMS chips cost a few dollars. Buying the same factor-of-two improvement from one quieter sensor costs vastly more, and buying it from a geophone means changing instrument class altogether — along with the price, the size and the fragility that come with a moving coil on a spring. Sensor fusion is simply the cheapest signal-to-noise on the market, which is exactly what a network that needs to be dense can afford.

And its limit, stated plainly. √N has sharply diminishing returns: going to eight sensors would buy only another factor of 1.41. Quad averaging does not turn a MEMS array into a geophone, and it is not meant to. It buys the margin that makes a €199 instrument genuinely useful for the earthquakes people feel — without pretending to compete for the microquakes a geophone was built to hear.

Raspberry Shake: Detail Over Speed

Raspberry Shake records continuous waveform data in miniSEED format — the industry standard for seismological data. This means every wiggle, every frequency component, every nuance of the seismic signal is captured and stored.

The geophone models (1D, 4D) are sensitive enough to detect distant earthquakes thousands of kilometers away — something MEMS accelerometers typically can't do. For local events, the accelerometer models provide strong-motion data similar to what GeoShake captures.

Think of it this way: GeoShake is a smoke detector — it tells you there's a fire NOW. Raspberry Shake is a surveillance camera — it records everything for later analysis. Both valuable, but for different reasons.

Who Should Buy What?

Network & Community

Both devices feed into community networks, but with different architectures:

GeoShake Network

The GeoShake network is optimized for speed. Sensors transmit PGA data via MQTT with millisecond latency. When multiple sensors in an area report elevated PGA values, the network can triangulate the event and push alerts to all users — potentially before the destructive S-waves arrive.

The network's strength grows with density. Because GeoShake sensors cost a fraction of research-grade instruments, it's feasible to create hyper-dense networks — sensors every few blocks — which dramatically improves detection speed and location accuracy.

Raspberry Shake Network

The Raspberry Shake network (ShakeNet) is a global network of personal seismographs. It's primarily designed for data sharing — stations appear on a map, and their waveform data can be accessed by anyone. Many professional seismologists use ShakeNet data to supplement their own research.

ShakeNet's strength is in data quality — each station provides continuous, research-grade waveform data that can be analyzed using standard seismological tools like ObsPy, SeisComP, and SWARM.

Can You Use Both?

Absolutely. In fact, using both gives you the best of both worlds:

Many serious citizen seismologists run both devices side by side. GeoShake handles the time-critical alerting; Raspberry Shake handles the deep data analysis.

The Verdict

There's no single "best" earthquake sensor — it depends on your goals:

For most people the choice is not really about sensitivity. It is about what you want the instrument to do on the day of an earthquake. A Raspberry Shake records that day beautifully and lets you study it long afterwards. A GeoShake T1 is built to tell you something in the seconds before the shaking arrives — and then to answer the question everyone actually asks once it stops: how hard did it shake here? At €199 with VAT and worldwide shipping included, it is priced so that the answer can come from your own street rather than a station 40 km away. That is the whole argument. Not a better sensor — enough of them.

Frequently Asked Questions