We report on the influence of pulse duration, bandwidth, and phase shaping on femtosecond laser induced breakdown spectroscopy (LIBS) and micromachining of metallic samples. Shorter pulses gave a lower threshold. Different phase functions were tested, and sample-dependent differences were found. Phase dependence was greater near threshold. When 30fs pulses were stretched to 10ps by linear chirp, little or no effect was measured on the LIBS signal, seemingly contradicting the advantages reported for femtosecond pulses. We determine that the bandwidth of the laser pulses is inversely proportional to the LIBS threshold.