$\newcommand{\id}{\operatorname{id}}$It can be proved that identity map $\id: S^1 \to S^1$ does not lift to $\widetilde{\id} : S^1 \to \mathbb{R}$ such that $e^{\widetilde{\id(z)}i} = z$.
The standard proof is based on injectivity contradiction (the lift can not be bijective because otherwise it will be a homeomorphism and it is not injective than for some $z_0, z_1 \in S^1$ we have $z_0 = e^{\widetilde{\id(z_0)}i} = e^{\widetilde{\id(z_1)}i} = z_1$. Contradiction).
It seems like this contraction doesn't work for, say, $z \mapsto z^2$. So does this map lift to $S^1 \to \mathbb{R}$?
In a word, "no".
If $\tilde{f}:S^{1} \to \mathbf{R}$ is an arbitrary continuous map, then its image is a compact, connected set of real numbers, i.e., a point or a closed, bounded interval. In any case, $\tilde{f}$ achieves an absolute maximum at some point $z_{0}$ of $S^{1}$. If $\pi:\mathbf{R} \to S^{1}$ is the covering map $\pi(t) = e^{it}$, then $f := \pi \circ \tilde{f}:S^{1} \to S^{1}$ fails to map a sufficiently small neighborhood of $z_{0}$ to an open set in $S^{1}$.
Particularly, $f$ cannot be a covering, or even an open map. Contrapositively, no covering map $f:S^{1} \to S^{1}$ lifts to a map $\tilde{f}:S^{1} \to \mathbf{R}$.