The drop hammer test is the easiest way to assess the sensitivity of explosive materials, but drop hammer results for low-velocity impacts have not been able to explain how explosives will react to other kinds of initiating stimuli. In order to do that, we have to understand the fundamental mechanisms of drop hammer initiation and how they differ from other initiation methods. For this reason, there is interest in instrumented drop hammers that help reveal what the drop hammer does at a fundamental level. We have developed a drop hammer that combines two types of mid-wavelength infrared (MWIR) imagers that, when operated simultaneously, can detect both the rapid explosion and slower combustion from impact-initiated polymer-bonded explosives with high time (1 μs) and space (15 μm) resolution. Results are presented that show how to vibration isolate the drop hammer to minimize MWIR image shaking during impact and to quantify the noise floor for MWIR temperature determinations via optical pyrometry. Experiments were performed on polymer-encased crystals of RDX ([CH2-NNO2]3) and HMX ([CH2-NNO2]4). Our experiments showed that drop-hammer initiated explosions occur in two phases with roughly 100 μs between explosions. Drop-hammer initiation is compared to an ultrasonic hammer, which initiates explosions by rapid frictional rubbing of the explosive surfaces against the surrounding polymer. The explosion rise time is faster with the drop hammer because the drop hammer inputs energy throughout the explosive volume, whereas the ultrasonic hammer produces localized heating and much more heat at the explosive surface.