Topic: Not clear asymmetry (conversion of Haar)

I can not understand, why for a square-topped symmetric signal  conversions of Haar it turns out asymmetrical. As far as I know, expansion on  - is unique, so why then there is a symmetry (after all, on idea then it is possible to receive mirror display the second variant of expansion): the source code in Mathematica: lst: = Table [0, {40}] ~Join~Table [1, {10}] ~Join~Table [0, {40}] ListPlot [lst] DiscreteWaveletTransform [lst, Automatic, Padding-> 0]; pic = WaveletScalogram [%, ColorFunction-> ColorData [{"DeepSeaColors", "Reversed"}]]