What is the best independent way (solution) to measure digital camera image quality? Not its control, auto focus or light-meter but the outcome from its sensor.
DxO Optics has put a lot of effort into that... DxO Mark.
But, honestly, the best "measure" is to shoot as you normally do, print the images, and if they look nice, then you're good to go
Posted by Richard Wong on December 9, 2009 at 2:27 PM
I agree with Floris. Shoot a landscape that has grass in the foreground then print the image to the largest size you think you will need then see if it meets your standards or not. If you don't own the camera, then rent it.
"But, honestly, the best "measure" is to shoot as you normally do, print the images, and if they look nice, then you're good to go "
I totally agree with Floris' statement. The only true way to gauge the performance of your system is to make a print. If it meets you expectations, you're in good shape.
Keep in mind that the sensor (or film) is only one peice of quality puzzle. If you're not getting the results you want, you'll need to look at your entire system and workflow to see which part(s) of it are letting you down and then work to fix the problem area(s).
-P
Posted by Bryan Maltais on December 9, 2009 at 6:39 PM
You can't really perform a test without comparing camera A to other cameras; the other cameras being the standard benchmark, or control. The only way you could test a camera by itself would be if you're already an expert on what an image "should" look like.
That said, the test would be to shoot the same scene, at the same moment, with the same lens (which is not always possible). Then compare the RAW images at 100%+ magnification side by side.
Posted by Buck Ward on December 9, 2009 at 7:43 PM
DPReview does test shots and compares standard images between cameras, but when I've researched a camera on DPReview I haven't really paid a lot of attention to the image comparisons.
Posted by Kent Staubus on December 9, 2009 at 11:28 PM
The problem with comparing images from different cameras online is you are really testing your monitor. And, at only 72 dpi. Let me go to a wider perspective here. I so far haven't selected a camera based just on sensor quality. Other things enter in. How much does it cost? How much better is the more expensive one than the one just under it in price? What lenses are available? How about the rest of the camera system? I'm thinking that all of the mid priced and above cameras in current production are so good I won't see much if any difference in real life.
Discussion
6 repliesBut, honestly, the best "measure" is to shoot as you normally do, print the images, and if they look nice, then you're good to go
I totally agree with Floris' statement. The only true way to gauge the performance of your system is to make a print. If it meets you expectations, you're in good shape.
Keep in mind that the sensor (or film) is only one peice of quality puzzle. If you're not getting the results you want, you'll need to look at your entire system and workflow to see which part(s) of it are letting you down and then work to fix the problem area(s).
-P
That said, the test would be to shoot the same scene, at the same moment, with the same lens (which is not always possible). Then compare the RAW images at 100%+ magnification side by side.
Kent in SD