NX 16mm f2.4 Different Performance With Different Bodies
I got some suggestions from the Samsung forum at DP Review, that NX 16mm performs better on older generation bodies – NX10/11 and NX100. I found one NX11 for a good price, so decided to make a fast test to see if there is any truth behind.After I posted results, following discussion on DP Review Samsung Forum indicates, that the problem might be linked to the RAW developing software. I noticed later, that in LR 3.6 and 4 Beta, RAW files from NX200 and NX 16, were corrected for lens issues (distortion, CA) automatically. Not so with NX11 and NX16mm.
Here are two screen shots with zero settings for the lens correction from LR 4 Beta. (Same is with LR 3.6)
 |
| NX200 and NX16mm in Lightroom without lens correction |
 |
| NX11 and NX16mm in Lightroom without lens correction |
Members from DP Review Samsung forum, namely Olivier 777 and PDM, suggested that issue can be caused by Lightroom and PDM proposed to use Raw Therapee v4 64. That suggestion brought me to the results presented at the bottom of this post, showing that indeed, problem is with a LR, instead of sensor/in camera processing.
I believe, that in time, LR will release update to correct this bug.
I will not comment software performances here, everyone has different work-flow, so you can make conclusion by yourself.
The first bunch of comparison shots, shows the result when I first tested cameras, developing RAWs in Lightroom 4 Beta.
First test scene was with cameras on a tripod at the angle of approx. 70 degrees, toward the ground. Distance from the focus point (middle of the frame was approx. 1.5m) Composition is slightly different between the cameras, but everything else is the same. It was shot in aperture priority mode, and exposure was left on the cameras (matrix). Single center focus point.
Results Center (Focus Area)
Results Top Left Corner
Results Top Right Corner
Results Bottom Left Corner
Results Bottom Right Corner
I also tried to downsize the NX200 image to the size of a NX11. It didn’t help much. Here is the sample at f2.4 of the most problematic top left corner.
Here is another scene – this time hand-hold. I will show results only at f2.4, as the difference remains more or less the same as in a previous case. Distance from the focus subject was approx. 2.5m. There was no wind at all.
Results at f2.4
Finally, here are new comparison shots of both cameras using both – Raw Therapee v4.0 64 and Adobe Light Room v4 Beta.
As noticed above, Lightroom applies lens correction automatically on NX200 and NX16mm RAW files, and I didn’t find a way to disable it.
Raw Therapee on the other hand allow automatic (but not profiled) lens correction on NX200 and NX16 RAW files, but not on NX11 and NX16 RAW files.
This finding proves at least one thing… Many test results on web are linked to one or another RAW developing software, and as seen here, differences are visible but also unpredictable. Therefore it is a good idea (but time consuming) to try yourself different softwares for different setups, to find the best possible combination.
All files here were processed in Raw Therapee and Lightroom using their basic settings and only auto WB correction was applied. Raw Therapee choose colder WB settings, but this can be easily corrected.
I think that it might be another issue. While NX11 (and other previous NX) sensor is very very old design back from Pentax K20D, NX200 is new sensor.
I think issue is same as between NEX-7 vs NEX-5N. 5N performs very good with old wide-angle lens (in both terms, sharpness and low color shift), 7 isnt so good with them and show color shift and most lenses are sharp only in middle (talking about rangefinder lens).
All works perfect with regular dSLR lens (or older SLR).
Tho question is why Samsung didnt optimized NX200 sensor for his own wide-angles. That blur shows that problem is much more than 6 mpix difference should be, eg. its simply sensor not optimized for short flange focal distance lenses. Even Samsung own..
But just my idea, nothing more.
Thanks Corwin for comment.
Right now, I found that developing RAW files from Samsung NX200 in LR3.6 and 4 beta, left files with in camera lens correction for some reason. I am trying to download other converter (DxO trial) to see, if this is Adobe bug, or NX200 applies lens correction in RAW. If the second is the case, than all comparison are useless and question is – did Samsung made it intentionally, or is it a bug.
I will come back with my findings hopefully soon.
This is very interesting post. I just bought Samsung 16mm lens and I’m using it with my nx10 body. The sharpness of the lens is excellent but distortion in Lightroom 4 is just intolerable. There was no problem If I only could have the 16mm lens profile for LR4. But sadly I’m still looking for one.
Hello. I found a way to recover real raw files. If you use ACDSEE 5.3 to convert raw file, than you can get a real raw file as captured in NX10 or older generation bodies.