After a Reddit user proved that Samsung’s photo processing system “created” details of the Moon, the Korean brand gave an official response. Through a statement on its website, the company explained how the optimization of captured images works, reinforcing something it had already said a few years ago.
The published text is not exactly new, as a very similar version had already been published in Korean last year. However, with the recent controversies, the content was translated into English and slightly edited.
The part that has more to do with the controversy concerns the performance of artificial intelligence (AI) in the images. According to Samsung, the photos undergo a deep learning model, based on a variety of shapes and details of the Moon.
This feature is based on images already captured from the natural satellite, at the angles at which it is seen from planet Earth — whether in waxing, waning or full phase.
In general, Samsung’s announcement does not give many extra details about this technology, compared to what was known before. However, a graphic produced by the brand intends to make the understanding more intuitive:
Samsung even cited other techniques to generate better photos of the Moon, such as the Super Resolution mode. It is activated at 25x zoom or higher, and uses multi-frame capture to combine more than ten images — this way, noise levels are reduced and the image is sharper.
In addition, the brand explained that its technologies optimize the exposure of the Moon, so that it does not appear “blown out” in the dark sky. Meanwhile, “Zoom Lock” combines optical and digital stabilization to prevent the image from blurring.
understand the case
The controversy surrounding the photos of the Moon captured by Samsung devices has gained momentum in recent days, after the user “ibreakphotos” showed his experiment on Reddit.
To prove that the device’s system created details artificially, he reduced the quality of a photo of the Moon downloaded from the internet. After turning off the light in the room and reproducing the conditions of the night sky, he noticed that the result was much sharper than the image itself.
The controversy has been commented on by experts and Internet users in general. The most critical believe that the manipulation of images is a kind of “cheating”, while others understand that the heavy processing of photos is something that already happens in all kinds of scenarios, anyway, and in the end, the beneficiary is the user himself who receives a photo much richer in details.
In summary, the message that Samsung wanted to convey with the statement is: the image processing system will not change. Therefore, the Lua will continue to “gain” sharpness when using automatic mode with AI optimization, whether the user likes it or not. For those who want something more “real”, just use the camera’s manual mode or Expert RAW, where AI adjustments are disabled.
Source: The Verge
