In the paper "The conjectural connections between automorphic representations and Galois representations" by Buzzard and Gee, it is said
"We say that $\rho$ is crystalline/de Rham/Hodge–Tate if for some (and hence any) faithful representation $H \rightarrow GL_N$ over $\mathbb{Q}_p$ , the resulting $N$-dimensional Galois representation is crystalline/de Rham/Hodge–Tate."
Here $\rho$ is a homomorphism from the absolute Galois group of a finite extension of $\mathbb{Q}_p$ to a reductive group $H$ (defined over some algebraic closure of $\mathbb{Q}_p$). How to prove the "hence any" claim?
Second question: given that the definition does not depend on this choice anyway, is there a way to give a definition not involving this choice (or is there a natural choice here)?