Speaking of mobile phone photo rankings, many people first think of DxOMark 's ratings and "photographed ladder." But is the camera's photo level really reflected by a simple score? How high is the credibility of DxOMark and many reviewers' photo placement rankings and ratings?
Professional evaluation agencies also have falling chains
Founded in 2003, DxOMark is a website that measures and ranks camera and lens imaging quality according to industry standards. Today, it is already one of the most famous and professional camera testing agencies on the planet.
It has a customized testing room. The testing room is covered with matt materials, the temperature is controlled at 21-25 degrees, and the humidity is between 30-70%. The on-site light source even uses spectrum equipment to perform color temperature correction to strictly control the entire test environment.
They have their own testers and calculation formulas that can process the RAW format files generated by the camera and produce fairly objective and valuable data. With the open test instructions, the generated data can be used to describe the machine's performance indicators like the parameter list. Of course, the prerequisite readers understand their test environment and have a certain technical basis.
Even if you can't read it, it doesn't matter. Just know that their camera review is very professional and meticulous. However, in the years when the quality of smart phone pictures has skyrocketed, DxOMark has also joined the testing and scoring of mobile phone pictures. There are many unanswered mysteries and legends:
Why is Sony's mobile phone camera scoring incredibly high?
Why can the Sony Z5 score surpass the Samsung S6 edge and iPhone 6s Plus ?
Why does S6 edge+ and Note5 have different scores when hardware and software are highly consistent?
What happened exactly? Now we begin to approach science.
Simple score judgment is not feasible
DxOMark's score table for each phone is as above, divided into "exposure and contrast, color, focus, detail texture, noise control, Artifacts, flash effects" 7 items. Artifacts is a problem of deterioration of the image quality caused by sensors, lenses, electronic noise, and algorithms, including purple fringing, sawtooth, noise, sharpening halo (white edges caused by excessive sharpening), and moiré.
If there is more than one machine, the above chart will be used to visualize the pros and cons of different items in the mobile phone.
The domestic media usually intercept only the general score of DxOMark and will not explain the small items of the score. Therefore, there are often theisms. For example, the Sony flagship can always shoot the first few in the same generation...
In addition to the "out of context" scores of different media for scores, refer to the seven sub-items with greater value, their scores are also doubtful. The first is the specific definition and standard of different sub-items, followed by the sub-item's scoring weight. In addition, this total score also incorporates video shots that are rarely compared.
Intercepting Samsung's flagship scores for the past few years here, you can see that Samsung’s flagship is improving every generation, but careful analysis will reveal that there are many strange places (red spots): S6 edge+, Note5, and S6 edge hardware are exactly the same , but why are the focus scores significantly different? The first two are even the same as the software. Why are the image stabilization and texture detail scores different? S5 with phase focus, why is the focus score even far worse than S4?
Compared with different models of different brands: modern models such as G4, 1020, and Z5, why does the Artifacts score lag behind old models? Why is the focus and flash score of Cable Dafa Z5 so high?
Nexus 6, Droid Turbo 2 is the same generation of moto works, the former is the optical image stabilization Sony IMX214, the latter is 2100W IMX220, Moto X Sytle is the Sony IMX230 sensor; Blackberry Priv rumors are intercepted IMX230, Nexus 6P is 13 million pixel Sony IMX377.
Comparison of device scores at the top of the photo rating and popularity chart
There are many doubts inside. There are many things here.
In addition, some of the tests released in the sample also hit a lot of DxOMark believers heart, above the G4 lens is clearly not clean and cause the light ... ...
Even if these doubts are ignored, the scores are not very meaningful to normal users. The score system is not divided according to the scene. The night scene and low light performance, which are the bottleneck of the mobile phone camera, are directly known.
The machine-friends value the analysis (corresponding to the texture and detail scores). It looks like the weight is not large, and the precision is not high: S6 edge with sharpness can actually be the same with Note5? The G4, the peak of daytime resolution in active duty, is only the same as Note5 and other machines. The item with the highest score is Nexus 6P (light and dark detail bonus from latitude).
where is the problem?The key to objective evaluation is to try to exclude people from the influence of subjective factors. DxOMark's expertise in the field of camera and lens evaluation is unquestionable. Their professional performance is in a strictly controlled testing environment. They have the means to convert proofs directly into data. They can try to rule out subjective human factors, but they are debatable in the field of mobile phones.
Partial formula selection for DxOMark camera test
The core issue here is that DxOMark's package is based on the processing of RAW files. However, many mobile phones cannot output RAW-format photos and cannot use the "sample data" method on the camera. Each proof is a product of the joint impact of the lens, CMOS, ISP, and algorithm. In addition, there are HDR, night scene noise reduction, and other algorithms for multi-frame synthesis of the “original filmâ€.
In the field of photography, the specific score is a very sensitive minefield. There are too many factors for taking pictures, and proofs and EXIF ​​information are objective. However, these so-called objective scores are subjective. The official did not publish the standard and reference documents (perhaps the author did not find English well?) and the original Zhang, we can not verify the authenticity. In a nutshell, DxOMark's mobile camera test is still a “black box†that cannot be reproduced and snooped by the outside world. The credibility itself has been discounted.
Of course, even if you ignore this large number of problems, its professionalism should still be more professional than the majority of domestic mainstream media camera evaluation.
Domestic evaluation agencies also have many minor problemsRelatively famous peers have stepped on the "grading system" pit. Although most of the test items that everyone pays attention to will be involved, which of them are more important? How specific weight? These are undecided items, and it is unlikely that there will be weight indicators that everyone agrees on. After all, the demand for photographs is different for everyone. Perhaps this person likes to use the analytical power of Mao Zedong, and that person only needs to be lovable. Color and contrast are enough to make a circle of friends.
In addition, the issue of the weight of the rating item will lead to many paradoxes, such as the scoring system in the verbal battle between Xiaomi 3 and MX3. The effect of hacking is as above.
Recently, another colleague (pictured above) accidentally stepped on this pit. We all said that the good Note5, actually scored a big loss to 6s Plus, or even Sony, believe that even the cable powder will not believe that ... ...
And even more evaluation agencies have even started to compare photos with light, scenes, shooting positions, and focus points. Even more frightening is that some of the cameras have not been wiped clean.
Blind rating total score comparison
Rating of votes for different projects in blind review
Since professional organizations may be tricky, wouldn’t the blind test be more credible? Unfortunately, it does not. Blind voting is conducted through web pages, but many voters do not have the mind or the speed to click to view the big picture, and some blind assessments are not even provided by the original Zhang.
This can only compare obvious items such as photometry, contrast, and color saturation. Numerous machine-friends' ambivalence, noise control, and even white balance cannot be considered (and there is no standard sample available to confirm the true white balance).
The final conclusion is very metaphysical, very simple, very hasty, very nonsense: To learn to use a dialectical perspective to see the comparison of pictures, there are many traps and even testers themselves did not notice. In addition to the direct comparison of the original photographs with good environmental control, otherwise there is no absolute conviction. However, if you simply use a simple score or ranking, you want to judge the level of the camera's camera. This kind of thing is all about entertainment topics.
The picture in the article is partly due to the network
WeChat scan, call the author ~ 
Weibo: 
Lynx Box M13
Installation Guide>>
Huawei Glory Box Pro
Installation Guide>>
Millet box 3
Installation Guide>>
Infinik i9
Installation Guide>>
KKTV
Installation Guide>>
Letv TV S50 Air
Installation Guide>>
My phone is my master, what APP is worth loading two: Do ​​you know these apps can be more refreshing? SioeyeIris4G Blink sports live camera evaluation VR machine Oculus Go listing, the country's summer trip out of the box one: BOSE SoundSport Pulse wireless Bluetooth sports headset out of the box do not run wrong: WeLoop Hele 3s sports watch out of the box comment 
Canon EOS M6 camera start experience from entry to give up: Synology Synology DS416play NAS network storage usage report (on) TITAN X Star Wars Collector's Edition release even AMD's red are used to help you Province 200: crack ZTE ZTE B860A IPTV TV Box tutorial Black Group DSM6.1 installation tutorial (with download) Latest equipment
Smart TV TV boxSmart TV / Box Reviews
Installation guide
Find the model to find the methodapplication
Rabbit video 
QQ Music HD 
Bird flying 
Rainbow music 
JJ Landlords 
Egg plan Snake 
Plants vs Zombies Hentai Edition HD 
RUI TV Desktop 
TV jailbreak tool 
Ten drops of water 
Phi film and television HD 
My Applications Hot topics
