Anonymous, 06 May 2020Excuses, excuses. You're making things a lot MORE COMPLICATED than they REALLY are. I never me... more"WHY does DxOMark need to HIDE their 'proprietary' and mapping' calculations when the subscores are already available for all to see"
I'm confused with what you're trying to tell me here, but isn't it to, as I already said, prevent companies from putting the most effort on the subscores that DxOMark focus on the most and to get an undeserved high overall score just for the sake of attention (like BS we saw in OnePlus & DxOMark partnership)? I apologise if I'm not interpreting your comment right.
Nick Tagataka, 06 May 2020I actually think sending an auditing team to DxOMark would be a good idea, as it will prove it... moreExcuses, excuses. You're making things a lot MORE COMPLICATED than they REALLY are. I never mentioned sending an audit team to DxOMark. All they need to do is SHOW how they arrived at the final scores given the subscores. Nobody is questioning the subscores because they have photos to share why they arrived at those. Users may have subjective assessments on the sample photos, but as for the FINAL SCORE which is 'calculated' from the subscores, how can DxOMark HIDE their calculations? WHY does DxOMark need to HIDE their 'proprietary' and mapping' calculations when the subscores are already available for all to see? SIMPLE really and you refuse to acknowledge that the reason is to avoid showing that they are 'FLEXIBLE' with the FINAL scores/calculations.
Nick Tagataka, 06 May 2020The example of students and teachers you used is utterly wrong. Unlike exams in school the tes... moreLook, you're making the analogy and comparison more difficult than it really is. DxOMark shows the tests they do and even shows pictures and verifiable results from the photos. How is it that the FINAL CALCULATION of scores from the SUBSCORES is not VERIFIABLE by anyone? Why? Because it gives them FLEXIBILITY. No one can question the final scores/ratings that naive users like to believe.
Anonymous, 06 May 2020A synonym for 'verifiable' is 'auditable'. DxOMark doesn't want audits because should anyone k... moreI actually think sending an auditing team to DxOMark would be a good idea, as it will prove its validity of their proprietary calculation method while keeping it confidential and taking away opportunities for companies to cheat at the same time. But then there would be other set of questions: How would you know that they are capable of objectively verifying their method and convince people that the auditors are not rigged?
And so far no organisation has put up their hands to check if their tests are fair and square, which kind of shows how difficult it could be to carry out the auditing properly.
Anonymous, 06 May 2020"In the moment it becomes "verifiable" it loses all the value.. shouldn't be that hard to unde... moreThe example of students and teachers you used is utterly wrong. Unlike exams in school the tests that all phone go through are identical, which means if a company knew a "mark scheme" (= how exactly DxOMark will calculate the subscores and the total score), they would pretty much nail the most important aspects of the test without putting much effort into others. Let's say they were actually using the fixed weighting system and somebody figured out that they put the strongest emphasis on exposure and noise sections. What do you think would happen? Companies would make their HDR algorithms more aggressive and apply a lot of noise reduction to get rid of grains just to receive high marks while skimping on other features, even if it meant that the overall image quality would go downhill.
The prime example is DxOMark & OnePlus partnership in 2017, where OnePlus teamed up with DxOMark and publicly stated that they worked together to optimise OP5's camera performance. Guess what happened? Complete garbage camera plagued with oilpainting issue somehow got the highest mark in their test. They got backlash from it and later they revamped the testing & scoring method to better reflect the real life camera performance.
And gosh please stop using capital letters everywhere BECAUSE it's HARD to READ.
A synonym for 'verifiable' is 'auditable'. DxOMark doesn't want audits because should anyone know how their scoring methods work, they'd lose credibility and flexibility in scoring. Also, they'll lose income from collaborative efforts with phone manufacturers who want their flagship products to get good camera ratings, and the ratings lose PHONE BUYER appeal. The ratings are designed in such a way that keep on increasing like PROCESSOR BENCHMARKS, because numbers SELL.
Nick Tagataka, 04 May 2020Of course they use the proprietary equations, their scores wouldn't mean anything if someone c... more"In the moment it becomes "verifiable" it loses all the value.. shouldn't be that hard to understand."
Amazing explanation! It's like a teacher telling students her scoring methods are 'not verifiable' by anyone because it loses value.
Actually, the REAL EXPLANATION is this:
The moment smartphone BUYERS find out just how INCONSISTENT (across phone reviews), UNRELIABLE AND UNREPEATABLE those 'proprietary' and 'confidential mapping' calculations are, DxOMark's PUBLICITY oriented camera scores lose value. It doesn't take a genius to figure out the 'proprietary and confidential mapping' into the final scores are UNVERIFIABLE by ANYONE because it would highlight the fact that they haven't been CONSISTENT with the 'Equations' used.
If you don't consider that FISHY, I don't know what is.
Anonymous, 03 May 2020Straight from DxOMark's mouth:
"The Overall score is not a weighted sum of the sub-scores. ... moreOf course they use the proprietary equations, their scores wouldn't mean anything if someone can just analyse the correlations between the subscores and the total scores and guess the fixed weighting pattern. In the moment it becomes "verifiable" it loses all the value.. shouldn't be that hard to understand.
And this should also prevent some manufacturers to find out which subcategory is the most important one. For instance, if the function is written in a way that applies the least % weighting on the highest subscore, it makes it a lot harder for the manufacturers to cheat by sending them devices that are specialised in some areas and perform poorly in others. In such cases they won't receive good overall scores.
judging by the results in the dxomark website.it would be quite a while before video and audio surpass the 100 mark.
Don't really care much for scores awarded by DXoMark! Not an iPhone users/owner and don't really care much for foldables either so little merit or attention from me as far as their opinions or scoring goes. Prefer real life comparisons..!
Straight from DxOMark's mouth:
"The Overall score is not a weighted sum of the sub-scores. It is a proprietary and confidential mapping of sub-scores into a combined score. The Overall score is also not capped at 100. That just happens to be where some of the best devices are currently."
Translation: The resultant DxOMark PHOTO score, resultant DxOMark VIDEO score and resultant overall DxOMark CAMERA score are all calculated by some FISHY 'PROPRIETARY' and 'CONFIDENTIAL MAPPING' of subscores.
It means it's NOT a simple FIXED percentage contribution per TEST ITEM/CATEGORY SUBSCORE WHICH MEANS, the overall PHOTO score, the overall VIDEO score
and the overall CAMERA score (the INFAMOUS DXOMARK camera rating) ARE NOT AT ALL VERIFIABLE BY ANYONE for a FISHY purpose or reason!
Good grief and people here in gsmarena still quarrel over DxOMark's DUBIOUS camera ratings comparing that this or that smartphone is better or worse because DxOMark overall CAMERA score says so.
Jake420, 02 May 2020Yes as a photographer something I've not heard before I happen to have played with some camera... morePeople in gsmarena*
Shui8, 02 May 2020Lol I dont need to trust you. I trust my eyes as a photographer that owns Fuji & Leica cam... moreYes as a photographer something I've not heard before I happen to have played with some cameras myself the Panasonic Lumix gx85 a good example so I can also use my eyes as a photographer? People with dxomark have a camera and suddenly think they have 10× better vision than the entire planet.you could see comparisons between mate 30 pro and any phone the S20 has contented with and tell me the mate 30 isn't oversaturated when it in fact is.
Anonymous, 02 May 2020My favorite brand is not in top of dxo, so trust me, they are paid? Correct? The images shot b... moreMy favourite brand rn is actually apple.but a phone with a better camera than every phone from last year being ranked under said phones And although it's more popular and came out before most phones on the list it was reviewed almost two months after however they were reviewed the day they were announced almost seems like company in question is fishy
Your website has such a comprehensive photo review section. There is no need to keep referencing a website that often has odd results. Why not just give the camera section a rating yourselves, then compare to other phones reviewed?
WHOmark lol
Jake420, 02 May 2020That's not the point I think Samsung can outlay Xiaomi yet somehow that's not what we see and ... moreMy favorite brand is not in top of dxo, so trust me, they are paid? Correct? The images shot by those guys are there for all to see... So everyone can know which phone has better camera. Oppo at least does not have autofocus issues..... Unlike the Samsung at dxo labs or in hand of every reviewer
Jake420, 02 May 2020That's not the point I think Samsung can outlay Xiaomi yet somehow that's not what we see and ... moreLol I dont need to trust you. I trust my eyes as a photographer that owns Fuji & Leica camera myself.
Shui8, 01 May 2020So you think Samsung is very pure not to pay DxO even when they got the money? That BS lol. Ev... moreThat's not the point I think Samsung can outlay Xiaomi yet somehow that's not what we see and it shows that dxomark is getting paid so isn't trustworthy to begin with.
Trust me the S20 ultra has better software than Xiaomi and more especially find x2 pro the S20 ultra destroys find x2 pro in every comparison except low light video
Tip us
1.9m 150k
RSS
EV
Merch
Log in I forgot my password Sign up