The paper aims to study an effective and precise method of preparing samples with uniform and small color differences among color patches to achieve effective and precise preparation. Firstly, samples were created by assigning the delta a* values between each adjacent columns and delta b* values between adjacent lines of each sample. Then visual experiment was executed to evaluate color differences between adjacent columns and adjacent lines. According to the delta a* values between adjacent columns and delta b* values between adjacent lines of the desired color difference, the delta a* between adjacent columns and delta b* values between adjacent lines were adjusted. After two rounds of adjustments, the third round samples for each color center were gotten. At last, the best sample for each color center was selected in terms of two evaluation index. After two rounds of adjustment, the samples of the second and the third rounds were improved and had similar results. Their evaluation results for accuracy were 1.75 and 1.90. The evaluation indexes for uniformity were 0.73 and 0.98. The method adjusts the delta a* values between each adjacent columns and delta b* values between adjacent lines of sample in terms of the algorithm based on visual results. The workload is appropriate and the samples are qualified.