Entropy is often used in polarimetric classification al- gorithms, for example by unsupervised Wishart classi- fication in alpha entropy feature space. In this frame- work, entropy is supposed to be low for man-made tar- gets. However, on most examples of classification results on San Francisco images, this parameter fails to well clas- sify the SOMA district, which contains a lot skyscrapers with a particular orientation. Even very recent studies fail to compensate the orientation effect on this area. More- over, TerraSAR-X images are also difficult to handle be- cause they show a high entropy with poor contrast be- tween natural and deterministic targets. Then, this paper investigates the reason of these issues. Several aspects are investigated: does entropy depends on noise ratio, wave- length, resolution size, orientation effect or complexity of the medium?