Spend a few minutes in the app store and you will find dozens of programs promising to turn your smartphone into a "Structured Light Sensor" capable of detecting ghosts. Ghost Seeker: SLS Camera claims to offer "cutting-edge ghost hunting with our AI-powered visual detection system." Spirit Box: Ghost SLS calls itself "one of the most advanced ghost hunting tools, designed to detect spirits, entities, and anomalies that are invisible to the human eye." GhostTube SLS Camera, with more than one hundred thousand downloads on Google Play alone, tells users it "detects humanoid bodies in your environment similar to the Kinect SLS cameras used by real paranormal investigators."
Here is the problem. Most smartphones have no structured light hardware. The "Kinect SLS cameras used by real paranormal investigators" were originally video game controllers. And the stick figures that appear on screen — the dancing skeletons that paranormal television presents as disembodied spirits — are produced by a machine-learning algorithm that was explicitly designed to guess at human body positions from partial data. False positives are not a malfunction. They are the entire point of the system.
This is a story about borrowed authority, algorithmic hallucination, and what happens when a technology built for bowling on your Xbox gets rebranded as supernatural surveillance system.
The Hardware That Is Not There
To understand why these apps cannot do what they claim, you have to understand what a structured light sensor actually is.
Structured light 3D scanning is a legitimate technology used in industrial quality control, cultural heritage preservation, and medical imaging. The process works by projecting a known pattern of light — usually infrared dots — onto a scene and measuring how that pattern deforms when it strikes physical objects. A camera reads the deformation, software calculates depth, and the result is a three-dimensional map of the environment. Wikipedia's entry on the technology notes applications including "precision inspection and dimensional analysis" and "the documentation and restoration of archaeological artifacts." It does not mention ghost hunting.
The device that started the paranormal SLS trend was the Xbox 360 Kinect, launched by Microsoft in November 2010. As Kenny Biddle, Chief Investigator of Paranormal Claims for the Committee for Skeptical Inquiry (CSI), described it in a 2017 Skeptical Inquirer investigation, the Kinect was "a sleek black bar" housing an RGB camera, a monochrome infrared depth sensor, an infrared light emitter, and four microphones. It was designed for one purpose: to let Xbox players control games with body movements instead of handheld controllers.
The Kinect's depth sensor projected a speckle pattern of infrared light into a room. The sensor read how that pattern deformed when it struck physical objects, then used triangulation to calculate distances and build a 3D map of the environment. But the stick figures that paranormal investigators treat as raw sensor output were not produced by the depth sensor alone. They were the product of a separate machine-learning system.
The Algorithm That Hallucinates Humans
Jamie Shotton, a researcher at Microsoft Research Cambridge, developed the Kinect's body-tracking algorithm. The system was trained on what Biddle describes as "literally millions of images and seconds of motion-capture video" of everyday people — different heights, weights, ages, clothing styles, and poses — gathered from homes around the world. Programmers labeled every body part in every frame. The data was fed into a machine-learning model that learned to map depth sensor readings to human joint positions.
The key word here is "guesses." Biddle emphasizes this repeatedly:
Based on the millions of reference images and video it has, the 'brain' (the software) guesses which parts of your body are the arms, legs, head, etc. That's right, I said 'guesses.'
The software then applies a wire 'skeleton' to you (the player), based on reference images and video, probabilities assigned to various areas of the body, formal kinematics... and guessing. Yes, the software can make assumptions on where it thinks a hidden body part (say, an arm that is swung behind the player) is or the path of travel an appendage might take.
The algorithm was trained to find humans. It has no concept of "not a human." When the depth data from a scene partially matches the reference library — a couch cushion, a trash bag, a curtain, a stucco wall — the system does not return an error. It extrapolates. It fills in the missing limbs. It draws a stick figure. Thirty times per second, in real time, the algorithm guesses, adjusts, and guesses again. The result is a dancing, wiggling skeleton attached to an inanimate object.
This is not a conspiracy or a cover-up. It is how the system was designed to work. Microsoft wanted the Kinect to track human players even when body parts were obscured, angled awkwardly, or moving unpredictably. The solution was probabilistic inference — educated guessing based on patterns. The side effect was that anything with vaguely humanoid depth contours could trigger a stick figure.
How "Ghosts" Are Manufactured
Kenny Biddle demonstrated this during an on-site investigation. In the basement of an old mansion, he and a ghost hunting group set up a Kinect facing a room with shelves, an old sink, and packed trash bags.
One of the bags had a stick figure on it. We sat and watched it for a little over ten minutes doing nothing more than wiggling its 'arms' and 'legs' until we moved the trash bag. Poof — it disappeared. We got it to come back by moving the trash bag around until the stick figure popped back on screen. We also turned the Kinect to face the stucco wall two feet in front of it. Guess what? We got a stick figure there too.
Biddle analyzed over two hundred "ghost" videos from paranormal investigators and identified consistent patterns. The stick figures were usually locked to specific objects: couches, chairs, jackets hung over chairs, potted plants, water heaters, camera bags, vacuum cleaners, tripods, pillows, pedestal fans, curtains. They rarely moved across a scene like a walking person. They stayed in one place and twitched, because they were attached to stationary objects whose depth contours happened to trigger the algorithm's pattern-matching threshold.
Jon Wood, a science performer who builds his own Kinect rigs for educational demonstrations, documented the same phenomenon:
A wet, grassy floor, in the darkness, and the Kinect thinks it can see a human form lying on the ground, where there is no physical form. Is it a ghost? No, it's the result of poor environmental conditions, well outside the limits of the intended use. The software is merely guessing, desperately searching for what it expects: a human form.
The Kinect was designed to be placed two to six feet from the floor, on a stable surface, facing a clear play area six to twelve feet wide, with no furniture in the way. Paranormal investigators point it at cluttered rooms, hold it while walking, and aim it at uneven surfaces in near-total darkness. Every one of these conditions increases the false positive rate. And when additional infrared lights — standard equipment for paranormal night-vision filming — are introduced, the depth sensor's measurements become even less reliable.
Biddle's conclusion is direct:
The cause for the 'stick figure ghosts' is multi-layered. The first layer is a limitation in the software, caused by a machine being forced to make guesses. Another layer involves people not paying attention to the simple rule that the 'play area' must be clear of obstacles except for the players. And the last layer falls heavily on the shoulders of ghost hunters, who too easily accept strange things they can't understand as paranormal without good reason or a decent effort at researching the topic.
The problem is not just user error. The device was built for a living room, not a haunted basement, and its reputation as a ghost detector has no scientific pedigree at all.
From Xbox to Ghost Hunter: A History of Entertainment, Not Science
The Kinect's paranormal reputation does not originate from peer-reviewed research or controlled experiments. It was popularized by a horror movie.
In October 2012, Paranormal Activity 4 featured the Kinect's infrared dot pattern as a cinematic device. Characters used it to see ghostly figures in the dark. As Biddle notes, "It seems to have followed in the 'footsteps' of the Ouija Board by gaining a reputation based on a movie." The film gave the ghost hunting community the idea that the Kinect could "see" ghosts. The idea was based on fiction, not evidence. But the Kinect used infrared light, and paranormalists already believed ghosts could be detected with infrared. It was a natural addition to their toolkit. (See our article Spirits and Thermodynamics or The Paranormal Thermodynamic Hypothesis: Is it Really Scientific? for more on this.)
By 2014, Bill Chappell — a paranormal gadget inventor whose company, Digital Dowsing, supplied equipment to the television show Ghost Adventures — was selling rehoused Kinect units for approximately $1,250 apiece, complete with a tablet and a 3D-printed case. The device appeared on Ghost Adventures and other paranormal television shows, and "the rest is history," as Wood puts it.
Newsweek profiled Chappell in 2020, noting that "the Ghost Adventures Crew made a deal with Chappell's company, Digital Dowsing, to field test equipment that isn't yet on the market. Every episode of Season 11 featured a different tool." The result was a feedback loop: television popularized the device, device sales funded more television appearances, and viewers concluded that a gadget used on a ghost hunting show must be legitimate.
Smoke and Mirrors in the App Store
This brings us to the current generation of SLS phone apps — and the place where marketing crosses into misrepresentation.
GhostTube, the most prominent of these apps, publishes a "How does GhostTube SLS work?" page that contains two radically different explanations depending on which phone you own.
For users with LiDAR-equipped iPhones (Pro models from 2020 onward):
Using the latest LiDAR technology (Light Detection and Ranging), GhostTube SLS projects a grid of Infrared light just like the traditional Kinect SLS camera and uses the infrared grid to detect depth and objects in the room.
For everyone else — the majority of all smartphone users, including Android owners and non-Pro iPhone users:
If your device lacks a LiDAR sensor, it is still possible to detect people shapes even though we have no infrared grid. This is achieved using complex machine-learning algorithms for detecting human poses and shapes.
Translation: on most phones, the app is looking at ordinary visible-light video through the standard camera and running a pose-estimation AI model on the footage. It is not projecting infrared light. It is not measuring time of flight. It is not using structured light. It is a camera filter that draws stick figures where its model thinks it sees a person.
This is the equivalent of selling a thermometer as a "Doppler radar storm tracker" because both give you information about the weather.
The distinction is buried in a blog post. The app store descriptions say nothing of the kind. Ghost Seeker promises "AI-powered visual detection" without mentioning that most phones lack the hardware for depth detection. Spirit Box calls itself a "Structured Light Sensor" without noting that this label only applies to a small subset of devices. The average user downloading these apps on a standard Android phone or non-Pro iPhone is buying a visible-light camera overlay marketed as advanced infrared sensing.
There is a secondary technical inaccuracy here. Even on LiDAR-equipped iPhones, the technology being used is not structured light. It is time-of-flight LiDAR. Structured light (Kinect v1) projects a known pattern and measures deformation. Time-of-flight (iPhone LiDAR) measures the round-trip time of individual light pulses. They produce similar outputs, but they are different technologies. Calling a LiDAR phone a "Structured Light Sensor" is technically wrong.
GhostTube's own documentation tacitly admits the app's limitations. Its "Tips for investigating" page states:
Remember that just like the traditional SLS camera, not every trigger is necessarily paranormal in nature. GhostTube SLS and traditional SLS cameras are designed to look for people shapes, so most figures captured can be explained.
A company selling a ghost-detection app acknowledges that "most figures captured can be explained" by non-paranormal causes. It also notes that the traditional Kinect "can operate in complete darkness," while the phone app cannot, because the visible-light camera lenses "are not capable of seeing in the dark" — the exact condition under which most paranormal investigations take place.
So if the apps cannot do what they claim, and the original hardware was never validated for ghost detection, what would it actually take to make a credible case?
What Would It Take to Be Credible?
A credible claim that a device detects disembodied spirits would require, at minimum:
-
- A defined, testable mechanism by which spirits interact with infrared light or depth sensors
- Controlled experiments showing the device produces stick figures in "haunted" locations at rates significantly above those in non-haunted locations
- Blind testing where operators do not know which locations are allegedly haunted
- Publication of failure rates and false positive frequencies
- Independent replication by researchers without financial interest in the paranormal equipment market
None of this exists. What exists are app store listings, television entertainment, and blog posts admitting that "most figures captured can be explained."
Biddle's assessment from 2017 remains the most concise summary:
False positives can happen anywhere with these pieces of equipment, and honestly, false positives are what drive the para-gadget industry.
The stick figure on your phone screen is not a spirit. It is an algorithm trained to find humans, doing exactly what it was built to do — guessing at patterns, filling in blanks, and producing a compelling visual in conditions it was never designed to handle. The only thing paranormal about it is how thoroughly the explanation has been ignored.
Works Cited
- **Biddle, Kenny.** "The Xbox Kinect and Paranormal Investigation." *Skeptical Inquirer*, July 7, 2017. https://skepticalinquirer.org/exclusive/the-xbox-kinect-and-paranormal-investigation/
- **"How does GhostTube SLS work?"** GhostTube blog. https://ghosttube.com/blogs/ghosttube/ghosttube-sls (Accessed June 27, 2026).
- **"Kinect."** Wikipedia. https://en.wikipedia.org/wiki/Kinect (Accessed June 27, 2026).
- **Wynne, Kelly.** "Meet 'Ghost Adventures' Friend Bill Chappell the Paranormal Inventor." *Newsweek*, June 25, 2020. https://www.newsweek.com/meet-ghost-adventures-friend-bill-chappell-paranormal-inventor-1513493
- **Nikolopoulos, P., et al.** "SL Sensor: An open-source, real-time and robot operating system-based structured light sensor for high accuracy construction robotic applications." *Automation in Construction*, Vol. 141, 2022. https://www.sciencedirect.com/science/article/pii/S0926580522002977
- **Shotton, Jamie, et al.** "Real-Time Human Pose Recognition in Parts from a Single Depth Image." *CVPR 2011*. Microsoft Research. https://www.microsoft.com/en-us/research/publication/real-time-human-pose-recognition-in-parts-from-a-single-depth-image/
- **"Spirit Box: Ghost SLS."** Google Play Store listing, May 2026. https://play.google.com/store/apps/details?id=com.spectralseekers.ghostcam (Accessed June 27, 2026).
- **"Ghost Seeker: SLS Camera."** Google Play Store listing, June 2026. https://play.google.com/store/apps/details?id=com.irishcoffee.ghostseeker (Accessed June 27, 2026).
- **"Structured light."** Wikipedia. https://en.wikipedia.org/wiki/Structured_light (Accessed June 27, 2026).
- **Ashford, Sam.** "What Is An SLS Camera? And How Does It Work?" SpiritShack blog, April 6, 2023 (updated October 11, 2025). https://www.spiritshack.co.uk/blog/ghost-hunting/what-is-an-sls-camera/ (Accessed June 27, 2026).
- **Wood, Jon.** "Let's Talk Tech – 5) 'The SLS Camera.'" jonwoodscience.com, December 31, 2024. https://jonwoodscience.com/2024/12/31/lets-talk-tech-5-the-sls-camera/
Edited on June 30, 2026
Fact Checked on June 30, 2026