Forum Twitter Facebook Google + RSS

Author Topic: New way to test mobiles needed? Or is DXO biased?  (Read 2898 times)

Offline mlife

  • Newbie
  • *
  • Posts: 1
New way to test mobiles needed? Or is DXO biased?
« on: April 11, 2014, 03:02:56 pm »
I have never even heard of this site up until this review... but I must say not the most positive introduction for a "credible" source.  It seems however this site earned its "credibility" while reviewing SLR's and hasn't put the same thought and effort into provide in depth reviews for mobiles. after reading the "sensor scores" and "in depth measurements" section on this site, I'm even more convinced they need to reevaluate what they are doing.

I know all the Sony Fanboys will jump all over my post but I would ask that you first think befor you blabber off some nonsense about DXO being "reputable" testers and me not knowing what I'm talking about... Consider this... On THIS VERY SITE for the PHOTO review of the 808 and Z2 they scored 81 points each - Do yourself a favor and pull out a calculator, tally up the scores, divide by 7 and tell me if they really each scored an 81 because according to my calculator you get 80.85714286 for the Z2 and 81.85714286 for the 808.  Now ask yourself "why would a REPUTABLE source roundup for one device and not the other?"  .... And I don't know what I'm talking about... ???

Okay, maybe the rounding issue was a MISTAKE or ERROR.... which I can accept because we all make them... but what I can't understand is some of the areas you would choose score a phone?  I get that while trying to leave a device intact, it is not realistic to perform actual sensor tests as described in the testing sections referenced above.  Also, I feel it IS IMPORTANT to evaluate the lens elements each manufacturer decided to use in their respective devices.  What I don't understand (if you're trying to determine actual quality) is the justification of;
EXPOSURE AND CONTRAST - IMO this is not valid due to MOST EVERY device having the ability to adjust this according to individual preference... maybe Dynamic Range would be a better measure.
COLOR - Most devices don't allow for RAW images so how is this rating the camera sensor when color is part of the .jpg processing algorithm.
AUTOFOCUS - I get this could be a great measure for a lens test (if evaluating separately) but HAS NOTHING TO DO WITH FINAL IMAGE QUALITY... maybe rate something of importance such as pin cushion, CA or vignetting.
FLASH - although can impact IQ (and I agree is a good category) I'm not convinced that if in a lab test is the best idea for an OVERALL rating unless you weight the average relevant to consumer use - day/night.

I know Ill get the usual ignorant Nokia fanboy comments but please try and understand that although I do love nokia devices, I also love the Sony brand.  I only use Nokia because camera is the #1 feature I shop for in a device and if Sony was truly better, I would jump aboard in TWO SECONDS flat.  I have no agenda other than to get some good quality testing on phones that would allow folks like myself (constantly seeking best IQ)  to have some real UNBIASED data to go on prior to making a purchase.
« Last Edit: April 11, 2014, 03:14:45 pm by mlife »

Offline riotsloth

  • Newbie
  • *
  • Posts: 1
Re: New way to test mobiles needed? Or is DXO biased?
« Reply #1 on: April 11, 2014, 06:25:40 pm »
The scoring does look a little odd.  Did you try adding up both camera and video?  If I added up correctly, I got 77.7 combined for the Z2 and 75 dead for 808.

So technically, the result stands, its just the scores shown are wrong.

Z2 is still the new highest rated smartphone for camera and video on DxO.