The Ti-Al-Ta-N coating was applied using the R-HiPIMS method (bias voltage -80V, substrate temperature 425 °C, power 1 kW). In a frequency range of 1-10 kHz, there is no oxygen in the coatings sputtered at a duty cycle of 10%, but at 5% duty cycle its amount reaches 20 at.%. Also the appearance of oxygen was detected at a duty cycle of 10%, when the discharge power was reduced to 0.5 kW, or the pulse frequency was less than 1 kHz. In contrast, oxygen was not found in the coatings deposited in the DCMS mode, as well as in Ti-Al-Ta coatings obtained without the reactive gas. What is the possible reason for this effect and how to prevent incorporation of oxygen into the coatings at low duty cycles and frequencies?