The speed of light is an important physical parameter. Currently it is a common belief of the constance of the speed of light regardless of the relative velocity between the source and the observer. Because the speed of light is very fast, if the relative velocity is small compared with the speed of light, it is difficult to detect the effect of the relative velocity on the measurement of the speed of light. In this paper we present a method of comparing the speeds of starlight and the light emitting from a terrestrial source. We use a telescope to collect the light from the star having significant relative velocity with respect to the earth, e.g. Capella. Then we modulate the starlight and the light emitted from the local source into pulses i.e. these pulses leave the modulator simultaneously. After travelling 4.2 km, these pulses are detected by a receiver. If the starlight and the terrestrial light have the same speed, then these pulses must arrive at the receiver at the same time. Our results show that the arrival times of the pulses of starlight are different from that of the local light. For example, the Capella is leaving away from the earth. The Capella pulses arrive later than the local light pulses. It indicates that the speed of Capella starlight is slower than the common believed value, c. The presented method uses one clock and one stick, so the clock synchronization problem and any physical unit transformation can be avoided.