In ordinary LED testing system, the reflectivity and Lamber properties of the coating on the integrating sphere surface are not ideal. One reason is low reflectivity and the other is poor diffuse reflectance. The result of the low reflectivity of the integrating sphere surface is that the direct light of the LED gradually decays after several reflections. However, in the whole process of light mixing, direct light and reflected light occupy a large proportion, which is dominant. In some cases, low reflectivity materials can have a strong shadow effect on the back of the baffle probe. However, this is the linear reflection of the shadow effect resulting in inaccurate measurement.
In addition, the lower diffuse reflection rate will seriously affect the attenuation of the signal. In the course of light measurement, the light is reflected in the integral sphere for many times, each reflection will produce a certain attenuation, but the influence of the reflection degree on the light intensity will be strengthened after many times of reflection. For example, the reflected light is reflected 15 times in the integral sphere, and if there is a 5% difference between the reflectivity, the signal attenuation may be more than doubled. In fact, the difference in reflectivity of the integrating sphere is much greater than that.

The current LED test system is not yet used as a standard LED as a standard light source. The standard halogen tungsten lamp was selected as the standard light source during the measurement. Due to the large difference in the external structure of standard light and measurement LED, including the influence of LED lamp holder's light absorption and the difference between standard light installation position and LED installation position, all of these are important factors affecting the accuracy of test results.