Quantization watermarking is a technique for embedding hidden copyright information based on dithered quantization. This non-blind scheme is only practical for watermarking applications, where the original signal is available to the detector as for a fingerprinting purpose. The goal of this paper is to analyse the quantization watermarking in the three-dimensional wavelet transform. We consider the nonlinear effect of dithered quantization in the time-domain representation of the filter bank. We derive a compact and general form for distortion in the host video due to the encoding and embedding process. The formulation has the capacity to be simplified and optimized for different filter banks and dither signals. We provide some supporting experiments for the three-dimensional wavelet analysis of video signal.