The high noise sensitivity of the Wigner distribution makes smoothing a necessity for producing readable time-frequency images of noise corrupted signals. Since linear smoothing suppresses noise at the expense of considerable smearing of the signal components, we explore two nonlinear denoising techniques based on soft-thresholding in an orthonormal basis representation. Soft-thresholding provides considerable noise reduction without greatly impairing the time-frequency resolution of the denoised distribution.