By using master method the solution of this recurrence T(n) = 4T(n/2) + n2 is "n2*logbase2 N", but using recurrence tree; its tree height will be:
(No.of.Leaves)N=d^h ==> N=4^h ==> logbase4 N = h
and at-last T(n) = O(n^2*logbase4 N) will be its running time, but not "n^2*logbase2 N" as by master method. Am I right?
[Update # 1]
As I found, log base doesn't matter in case of Big O analysis. Thereby, O(n^2*logbase4 N) could be equal to O(n^2*logbase2 N) Asymptotically