The sunlight experiment
In 2011, S. Karmakar, Y. Zhai, H. Chen, and Y. Shih reported The First Ghost Image Using Sun as a Light Source at CLEO. Their research abstract discusses two-photon ghost imaging with solar illumination and how its correlations might be interpreted. This was the researchers' work, not an experiment performed by the blog.
The resolution comparison
The conference abstract compares potential resolution with a conventional optical system imaging an object 10 kilometers away with a lens 92 meters in diameter. This is a comparison reported by the authors. The full experimental methods and assumptions were not examined here, so it should not be read as proof of a general-purpose camera with that performance.
A later lensless experiment
In 2014, Liu and colleagues reported lensless ghost imaging with sunlight in Optics Letters. They selected a narrow spectral region of sunlight, measured correlations, and reconstructed an object with two apertures. This is a separate laboratory result, not proof of unrestricted long-distance imaging.
The use of sunlight followed earlier experiments with thermal light. In 2009, Chen and colleagues demonstrated lensless ghost imaging with a thermal light source. Their experiment did not magnify the image. The 2014 study by Liu and colleagues demonstrated the technique with filtered sunlight. These results establish feasibility under specific experimental conditions, not a universal advantage in spatial resolution.
Quantum and classical correlations
A 2004 theoretical analysis by Gatti and colleagues compared classically correlated thermal beams with entangled beams. Natural illumination motivates questions about the interpretation of optical measurements, rather than establishing entanglement in every arrangement.
How to evaluate claims about performance
The 2014 experiment used filtered sunlight and a target with two openings. This is a useful demonstration of feasibility, but a two-opening target is not a detailed scene. A comparison of imaging techniques should therefore specify the target, distance, illumination, measurement time, and reconstruction quality.
Spatial resolution and image quality are related but distinct: resolving two features does not establish that the resulting image has high contrast or low noise. For a practical use such as remote observation, the relevant question is whether the correlation method offers a measurable benefit under the same constraints as another optical system. The conference report and the later lensless experiment should be evaluated independently rather than combined into one claim of performance.
Practical boundaries
Noise, available light, detector properties, measurement duration, and the reconstruction procedure all affect performance. Whether correlation imaging offers an advantage depends on the specific application and the conventional system used for comparison.
Read the companion article
Ghost Imaging (I): Light Correlations and 1995 Experiment
Editorial revision: November 15, 2026.
Explore further: Explore more science articles in the Article Library.