Recently we put some extra time in developing a new feature for our site. With the cameraphones making such rapid progress, people are starting to take more and more interest in them. Trying to keep up with the growing demand we came up with a new test for the camera sections of our reviews.
For our test we use a standard ISO12233 chart to evaluate the synthetic resolution of digital cameras. It is an excellent tool for measuring pure horizontal and vertical resolution and also offers a good reference point for resolving power across a sample of imaging devices.
The long parallel lines serve as good measurement of resolved detail. The point where the lines are no longer clearly separated is where the resolving capability of the camera maxes out. There's a number scale alongside so you can even assign a numeric value to the camera's performance.
The several different line sets are taken from various places across the frame so you also have a nice illustration of how the image resolution deteriorates at the edges of the frame (poor optics are far blurry towards the ends than in the center of a frame).
We have also added a few touches of our own by including a color and a grayscale palette to the chart. Those are standard charts too, although their application in digicams testing is somewhat different.
With cameraphones, we are using the color chart to illustrate the amount of noise in different colors. This is a value that tends to quite vary from phone to phone depending on the sensor and the noise reduction technique applied.
Meanwhile, the thin lines at the top of both the color and the grayscale charts illustrate the actual effect of noise reduction. You will notice that the numbers are barely readable with some phones and more easily discernible with others.
The gray scale is there to trigger the usually more powerful noise reduction, which kicks in when shadows are involved. Once that happens, you will notice complete loss of detail towards the darker end of the scale. Usually with a poor camera you can't really distinguish the shades of black, while in reality there is obvious difference.
Bear in mind though that some aspects of the camera performance that you normally look for aren't relevant in this studio test. For example getting proper exposure doesn't matter here, as phones aren't tuned for taking photos of such charts. How it behaves here usually doesn't have anything to do with a handset's real life performance. Wrong exposure doesn't affect the synthetic resolution either, so if one of the handsets fails to get a proper exposure it isn't in a disadvantageous position.
On the following pages you will find the samples of all the cameraphones we have tested so far, grouped by camera resolution. You can also follow the links below if you are looking for a specific phone's results.