Quel smartphone a pris LA MEILLEURE photo de l'éclipse ? (comparatif)

Quel smartphone a pris LA MEILLEURE photo de l'éclipse ? (comparatif)

Which smartphone took THE BEST photo of the eclipse? (comparison)

🎙 Frandroid 👥 625K 📅 August 12, 2026 ⏱ 13 min 👁 164K 📄 original study 🧭 2026-10-20
Available in: English (current) Français

Keywords

eclipsesmartphone cameraAI zoomXiaomiPixelSamsungOppo

Summary

The video tests four Android flagship smartphones (Samsung Galaxy S26 Ultra, Oppo Find X9 Ultra, Xiaomi 15/17 Ultra, Google Pixel 10 Pro XL) during a partial solar eclipse in Paris. The setup includes a tripod, identical framing, and a solar filter for each phone. The host explains the challenge: extreme zoom, high contrast, and a subject the AI training rarely includes. After taking photos in auto mode, the results are revealed and analyzed. The Xiaomi superimposes a moon texture on the sun, a clear AI artifact. Samsung’s photo appears saturated and lacking detail. Oppo shows a slight softness with HDR ghosting, though a later shot is sharper. Google’s Pixel 10 Pro XL produces the most realistic and detailed image, using its AI reconstruction. The final ranking places Xiaomi last, Samsung third, Oppo second, and Google first. The host emphasizes that while Google also uses AI, it works better and respects reality, unlike Xiaomi’s deception.

154 words

Critical Evaluation

Value of the Information & Strength of the Argument

The video provides original, hands-on evidence by capturing the same celestial event with four phones under a controlled protocol. The argumentation is strong: it demonstrates the influence of AI on computational photography, highlighting both benefits and ethical issues. The comparison is fair, as all phones use auto mode, and the host openly acknowledges limitations. The analysis is well-reasoned, with each photo examined critically against the observed reality. The conclusion is persuasive, backed by visual evidence and a clear ranking. The inclusion of a second photo set reinforces the consistency of results.

Scientific Rigor, Source Quality, Title Accuracy

The scientific rigor is acceptable for a consumer tech test. The host explains the role of AI training and post-processing, citing Samsung and Google’s statements (though without direct links). The protocol is replicable, and the limitations of the eclipse’s progression are mentioned. The sources in the description are mostly channel links and a guide page, not academic references, but the video does not claim to be a scientific publication. The title accurately reflects the content, and the adéquation is strong. The video includes a short ’ethics’ moment about AI deception. Overall, the methodology is transparent and the conclusions are justified.

206 words

Title / Content Match

The title directly matches the content: the video compares four flagship smartphones to determine which one takes the best eclipse photo. The revelatory result is delivered clearly.

Quality & Reliability

8/10

The video presents a clear comparative test with a controlled protocol (same scene, solar filter, identical framing). The analysis is transparent about limitations (eclipse progress, auto mode) and the results are based on actual photos. The channel discloses non-sponsorship. Some subjective judgments are present, but overall methodology is sound for a consumer tech test.

Chapters

Cited Sources

Concurring Sources

Dissenting Sources

  • No direct contradictory sources found — The video's findings are based on original tests not yet contradicted by other sources. However, the general debate on AI authenticity in photography could be seen as a counterpoint, but not specific to this test.

Contribution & Novelties

The video’s main contribution is a practical, side-by-side demonstration of how AI-based computational photography can distort reality, specifically using a solar eclipse as a rare and challenging subject. It exposes that Xiaomi’s camera algorithm misidentified the sun as the moon and applied a synthetic lunar texture, raising ethical questions about photo authenticity. It also shows that Google’s AI reconstruction, while also synthetic, produces a more faithful result. This provides concrete evidence for ongoing debates about AI in photography. The test is timely, given the rarity of eclipses.

Pour aller plus loin :

  • Computational photography — Overview of techniques used in smartphone cameras, including multi-frame processing and AI enhancement.
  • AI hallucination — The phenomenon where AI generates incorrect or fabricated content, relevant to Xiaomi’s moon texture.
  • Google Super Res Zoom — Google’s approach to zoom enhancement using multi-frame and machine learning, explaining why Pixel processed the eclipse well.

147 words

Radar Profile

The radar profile shows a balanced performance with strengths in information quantity and quality, and slightly lower in technical depth and reliability. The high quality of the visual evidence and clear analysis compensate for the lack of statistical rigor typical of a consumer test.

Reliability 7/10

💬 The climate is predominantly positive with a touch of humor and irony. On the 30 comments analyzed, many viewers appreciate the transparency and the exposure of Xiaomi's AI deception, while some express amusement at the 'moon on the sun' effect. A few commenters note the distinction between astronomy and astrology, and others share personal experiences, indicating engaged and satisfied audience.