Note that this reference documentation is identical to the help that is displayed in MATLAB when you type “help music”.

  MUSIC source localization using MUltiple SIgnal Classification
  This is a signal subspace method, which covers the techniques for
  multiple source localization by using the eigen structure of the
  measured data matrix.
  Use as
    [dipout] = music(dip, grad, headmodel, dat, ...)
  Optional input arguments should be specified as key-value pairs and can be
    'cov'              = data covariance matrix
    'numcomponent'     = integer number
    'feedback'         = 'none', 'gui', 'dial', 'textbar', 'text', 'textcr', 'textnl'
    'reducerank'       = reduce the leadfield rank, can be 'no' or a number (e.g. 2)
    'normalize'        = normalize the leadfield
    'normalizeparam'   = parameter for depth normalization (default = 0.5)
  The original reference is
    J.C. Mosher, P.S. Lewis and R.M. Leahy, "Multiple dipole modeling and
    localization from spatiotemporal MEG data", IEEE Trans. Biomed.
    Eng., pp 541-557, June, 1992.