In this paper we present three new algorithms for sparse multichannel blind deconvolution. The first algorithm is based on a cascade of a forward and a backward prediction error filter. The second consists in an alternating minimization algorithm for estimating both the reflectivity series and the seismic wavelet. The last one is a regularized version of the so called Euclid deconvolution, solved by the well known fast iterative shrinkage-thresholding algorithm (FISTA). Simulation results illustrate that the three algorithms presents some quite interesting results in both synthetic and real data.