An asynchronous chorus method and device, which are used for solving the problems of a poor chorus effect, a complex processing process and high costs. The method comprises: after receiving an audio file uploaded by a terminal which requests the use of a first obbligato file to participate in a chorus, marking a part in the audio file, which has a mixed sound and is not marked, as a part which has already been chorused (101); and determining the marked audio file as a second obbligato file (102). Therefore, chorists can sing their own parts at different times and different places, which will not lead to a poor chorus effect due to crowding or that the sound has a difference in volume and distance. In addition, when singing the parts of their own, each user can sing many times or separately process the part sung thereby, without influencing the parts sung by other users, therefore it will not lead to the situation where the whole song has to be sung again because of the poor performance of one user.
展开▼