return b;
}
+/**
+ * Converts a double precission floating point number to a AVRational.
+ * @param max the maximum allowed numerator and denominator
+ */
AVRational av_d2q(double d, int max){
AVRational a;
int exponent= FFMAX( (int)(log(ABS(d) + 1e-20)/log(2)), 0);