smoothrandonce.sh

1)  when averaging: velxp and the direct smoothed velpdirect are close (however)
		    obvious changes in signal , mean is about the same
			does not see much of layering effect

2)                   velzp is about 100 m/s slower than the direct smoothed, see		     s the effect of layering.


smoothrandonceslwo.sh

	looks much closer  , now I smooth direct the slownesses
	makes sense since v=v(z)   not v(t)

1)	velxp > velpdirect ca 60m/s
2) 	velzp < velpdirect ca 50m/s

3)    velzp != velxpnmo      not pseudo isotropic  when vp/vs ratio changes
