Spatial audio capture

A method of spatial audio capture comprising: determining a 1st direction parameter through modifying a pair of audio signals, and determining a 2nd direction parameter based on the modified signals. Modifying the audio signals may comprise: maximising the correlation between the pair of signals through applying a delay to one of them, identifying and subtracting a common component of the signals, and removing the delay from the delayed one of the pair. The common component may be multiplied by a gain value. The method may be applied to individual frequency bands of the sound signal and may only be applied if the frequency band is below a threshold. A 1st sound energy parameter may be determined based on the audio signals, and a 2nd sound source energy parameter may be determined based on the audio signals and the 1st energy parameter. The energy parameter may be a direct-to-total ratio.