$U(\theta,2\theta)$: the likelihood falls in $\theta$ on $[X_{(n)}/2,X_{(1)}]$

Let $X_1,\dots,X_n$ be iid $U(\theta,2\theta)$. The maximum likelihood estimator of $\theta$ is: