# Electric field discontinuity

When there is a charged spherical shell for example, I can calculate its charge using: $$(\vec{E_{out}}-\vec{E_{in}})\cdot \hat{r} = \frac{\sigma}{\epsilon_0}$$ We generally ignore external electric field if it present because it cancels out, but what about if the spherical shell is inside a conductor? then $$\vec{E_{out}}=0$$ and we get something completely different - now the external field doesn't cancel. What have I done wrong?