Embodiments are disclosed for enhancing engine sound. An example method for a vehicle comprises acquiring a signal including harmonic content generated by an engine of the vehicle, upmixing the signal into a plurality of channels for a given number of engine orders, adjusting an order filter for each engine order of the given number of engine orders based on operating conditions of the engine, filtering each channel of the plurality of channels with the corresponding order filter, mixing the filtered channels into a mono output, and outputting the mono output to at least one speaker in the vehicle. The mono output is delayed based on a position of the at least one speaker such that an occupant of the vehicle perceives the mono output as originating from the engine.