It is known that in general, the moduli space $M$ of (Gieseker) stable sheaves do not have a universal family. Therefore, a map $f \colon S \to M$ does not necessarily induce a family of sheaves.
On the other hand, there are references (the book by Huybrechts and Lehn 2nd edition on p.139 or Simpson's paper Theorem 1.21) that state that a universal family exists étale locally.
I was wondering what this meant precisely.
Does it mean that for any scheme $S$ and any map $f \colon S \to M$:
- There exists an étale surjective map $\alpha \colon U \to M$ and our "étale-local universal family" $\mathcal{E}$ on $U \times X.$
- There exists an étale surjective map $\beta \colon T \to S$ and a map $g \colon T \to U$ such that the map $f\beta \colon T \to S \to M$ corresponds to the family $g^*\mathcal{E},$ i.e., the morphism induced by $g^*\mathcal{E}$ is isomorphic to $f\beta?$