DXO-Mark review of the Lumia 1020

[arni99]

Nokia Lumia 1020 overview: Has the best got better? - DxOMark

PROS:

• Excellent detail preservation both in low and bright light conditions.
• Very low noise level with no chroma component.
• Accurate exposure in extreme low light conditions (where other camera phones fail).
• Pleasing photo rendering with flash: good exposure, color, white balance, texture and noise.
• Fast and precise autofocus.
• Good texture preservation in bright lighting conditions.

CONS:

• Strong color shading with indoor lighting and occasionally outdoors.
• In low light conditions, overly long exposure times lead to motion blur.
• Slightly tendency to underexposure outdoors.
• Occasional white balance inaccuracies outdoors or with fluorescent lightings.
• Substantial flare noticeable in presence of a strong light source.
• Strong row noise evident
• Visible blocking artifacts
• Walking movements remain uncorrected by the video stabilization
• Autofocus oscillations are visible during refocusing.
• Over-correction of video stabilization when steadied (on a tripod).

• 1. Nokia 808 Pureview
  
  DxOMark Mobile:
  77
• 2. Apple iPhone 5S
  
  DxOMark Mobile:
  76
• 3. Sony Xperia Z1
  
  DxOMark Mobile:
  76
• 4. Samsung Galaxy S4
  
  DxOMark Mobile:
  75
• 5. Nokia Lumia 1020
  
  DxOMark Mobile:
• 74

 

[Tech friend]

This test is pretty balanced and objective, its results coincide to a great extent with my own assessments.

Somehow I feel my disappointment regarding the 1020 confirmed, after having been very enthusiastic initially.

 

[vlad0]

right.. they gave 86 on the flash test to the Z1 , 79 for the 1020.. and 81 for the 808, which should be 95+ in relative terms to the other phones

and here it is

Z1

Note 3

 

[blue1k]

Coming from an S4 I can't believe it rated higher than the 1020. No way. Lol

 

[Chazn]

So, the video quality is kinda poor and the optical image stabilization didn't do squat to dampen hand-shake.
Any idea if Nokia would be able to fix the row noise and artifacts in videos in an upcoming firmware update?

 

[Tech friend]

In the meantime Steve Litchfield commented on the DxOMark test, questioning especially the video part.

I am drawing new hope from his overall upshot:

The Lumia 1020's hardware (as with the 808's before it) is tremendously capable and it will probably take an update or two to really get the most from it.

In the meantime, the Nokia 808 still sits top of the DxOMark league table for smartphone cameras, with the Lumia 1020 in fifth place overall. I have every confidence that a re-test in a few months will see the 1020 significantly higher.

 

[Nick_1020]

It seems incredibly harsh for the 1020 to be rated 5th.

It's probably tied first with the 808 for me. Some things the 808 does better and other things the 1020 wins. Very surprised about the criticism of the video quality as I find it to be outstanding.

 

[Bahamen]

Many so-called research houses may conduct tests and publish reports that sound scientific, but in the end, there will still be a lot of subjectivity involved in assigning scores to image quality and to assign weights to each category. Ultimately if these results fail the common sense test it means something is wrong with it. Just a few examples of their test which I find to be quite extraordinary:

1. 1020's video stabilization score: 49 with Optical Image Stabilization (OIS). Z1's video stabilization score: 59 with no OIS(!!!). A freaking 10 points higher score just by having no OIS!!!

2. Flash: 1020's xenon: 79, Z1's LED: 84. Huh(??).

3. 1020's video noise: iPhone 5S has 75 points. Z1, having a larger sensor, has 85 points. The 1020, with sensor 230% the Z1's and 400% the 5S's, scores a whopping 53 points!!

4. The 1020 has slight tendency to underexpose outdoors? Now, this can be very subjective, a correctly exposed image for some may be under/over-exposed for others. In Steve Litchfield's comparison at AAWP, the 1020's image is correctly exposed while the Z1 is quite badly overexposed. Anyway, that's what the Pro Cam's exposure adjustment is for, isn't it?

5. DXO deducted quite a lot of points for the 1020's colour. The huge penalty dragged the overall score down by at least 3 points for stills. Anyway, that's what the Pro Cam's WB adjustment is for, isn't it?

 

[tgr42]

Common sense also dictates that specs aren't everything and actual results are what matter. I don't know about the other points but when it comes to video noise, the 1020's low score is well deserved in my opinion. I don't know why but video on the 1020 is incredibly noisy, worse than a webcam. I have to run it through a strong temporal filter to get something that looks half decent when viewed at full resolution.

 

[Bahamen]

Common sense also dictates that specs aren't everything and actual results are what matter.

Of course, but such wide disparity in the specs should at least trigger some serious questions. It's like someone with a stick winning against someone with a gun - of course that can still happen.

I don't know about the other points but when it comes to video noise, the 1020's low score is well deserved in my opinion. I don't know why but video on the 1020 is incredibly noisy, worse than a webcam. I have to run it through a strong temporal filter to get something that looks half decent when viewed at full resolution.

That has certainly not been my experience so far. If your 1020 video is worse than webcam, I strongly suggest you bring it to Nokia Care Center.

 

[vlad0]

The dxo results lost all credibility in my book ... those flash scores are silly. Also, the video on the 1020 is not all that noisy.

I don't know what criteria they are using, but .. you can always tweak your criteria to match the desired results.

I am very surprised that they still have the 808 on top, granted they've made the difference seems as small as possible, but I guess they couldn't bring it down further because it might really hurt their credibility.

 

[Tech friend]

I don't know about the other points but when it comes to video noise, the 1020's low score is well deserved in my opinion. I don't know why but video on the 1020 is incredibly noisy

Other owners reported this too, at least for low-light recordings:

Nokia Lumia 1020 color accuracy, low-light video performance called into question

There’s also word of middling performance when recording video in low-light conditions. While the 1020 shines at still images in low-light environments, that doesn’t seem to carry over to video, and the resultant output is riddled with noise. That’s a little surprising, and we wonder if such performance could similarly be improved via update, or if this could be a limitation of the phone’s hardware.

 

[Tech friend]

If your 1020 video is worse than webcam, I strongly suggest you bring it to Nokia Care Center.

Bahamen,

do you have a static tone in your videos when recording in a quiet environment?

I would like to hear whether there are 1020s not suffering from that issue, so that a replacement of my device could make sense.

Thank you in advance.

 

[Tech friend]

The dxo results lost all credibility in my book ... [snip] I don't know what criteria they are using, but .. you can always tweak your criteria to match the desired results.

So do you think they are biased?

If so, I would wonder why.

 

[vlad0]

^ Yes, but its all conspiracy... I don't really have any solid proof that they are It just that the scoring doesn't make any sense to me since I've seen what the 5s , the 1020, and the 808 can do... and from image quality perspective and overall camera flexibility (IQ in difficult conditions, etc.) the 1020 and the 808 are way ahead.

It should be..

808: 77
1020: 70
5s: ~57

and what device is 100 ? Are they waiting for a phone with 1" sensor or something ?

gsmarena didn't find anything wrong with the video recording on the 1020

Six-way camera shootout: Smartphone roulette - GSMArena.com

But they don't have the raw video files for download..



Does anyone here care to upload a short low light video clip ?

 

[Bahamen]

Bahamen,
do you have a static tone in your videos when recording in a quiet environment?
I would like to hear whether there are 1020s not suffering from that issue, so that a replacement of my device could make sense.
Thank you in advance.

Yes, there is some static sound, it is quite audible when played over the phone. But it is barely audible when played on the PC unless you set the volume at very high levels. It is somehow louder if you set bass filter to "Strong". But it wouldn't bother me since it is still quite soft and should easily be covered by ambient sound. You can check out the sample video here.

Does anyone here care to upload a short low light video clip ?

Here. It is quite dark.

http://sdrv.ms/16mTcxp

 

[Bahamen]

Anyway, link below to a spectacular demonstration of how webcam-like "noisy" the 1020's video, and how poor the stabilization is:

Free runtastic! Is this the coolest video ever shot on a Nokia Lumia 1020? – Nokia Conversations : the official Nokia blog

Choice quote from "one of the world’s top extreme sports film makers Claudiu Voicu":

Auto-exposure and focus works great, and it didn’t once fail when we taped it to the bottom of the camera crane (where it recorded non-stop for 45-minutes). It’d be nice if my DSLR had auto-focus that good in video mode…

I'm going to bet that a certain somebody at WPC is going to immediately dismiss this outright as "blatant publicity" and accuse Mr Voicu of lying for money.

In a world that perhaps not everybody is necessarily always a liar, perhaps Mr Voicu may have a point. I mean, how else you shoot some fast-moving people jumping all over the place without decent auto-focus? Now that brings the question, is the 1020's video auto-focus so bad that DXO gave it only 45 points, much lower than even Blackberry Z10's 64 points and lower than even the iPad??

 

[vlad0]

Here. It is quite dark.

https://skydrive.live.com/pagenotfounderror

Thanks!

yes, there is some noise but.. isn't that normal ? I will have to test the 808 and see what I get in low light.

 

[Bahamen]

Thanks!

yes, there is some noise but.. isn't that normal ? I will have to test the 808 and see what I get in low light.

Yes, that is to be expected for that kind of lighting and not something that justifies such severe penalty by DXO. The 808 has similar level of noise in lowlight, although by default the 808 will choose to underexpose the scene. Once you increase the exposure, the noise will be quite visible.

 

[vlad0]

^ yes, the 808 does seem to underexpose for some reason.. I will compensate for that

Here is another test .. note 3, z1, 1020

Cameras Compared: Lumia 1020 vs Xperia Z1 vs Note 3 Photo Test Part 1 | SoyaCincau