Normalization and conventions regarding mode expansions and spinor normalization in different textbooks

The mode expansion for the spinor field in Srednicki's book on Quantum Field Theory is the following:

$\Psi(x)=\sum_{s=\pm}\int\frac{d^3\vec{p}}{(2\pi)^3(2E_p)} \Big[ u_s(\vec{p})~ b_s(\vec{p})~ e^{ip\cdot x}+ \upsilon_s(\vec{p})~ d^{\dagger}_s(\vec{p})~ e^{-ip\cdot x} \Big]\\bar{\Psi}(x)=\sum_{s=\pm} \int\frac{d^3\vec{p}}{(2\pi)^3(2E_p)} \Big[ \bar{u}_s(\vec{p})~ b^{\dagger}_s(\vec{p})~ e^{-ip\cdot x}+ \bar{\upsilon}_s(\vec{p})~ d_s(\vec{p})~ e^{ip\cdot x} \Big]$ where $E_p=\sqrt{\vec{p}^2+m^2}$, and $b^{\dagger}_s(\vec{p})$ and $d^{\dagger}_s(\vec{p})$ when promoted to operators create fermions and anti-fermions with momentum $\vec{p}$ and spin index $s$. In addition, $u_s(\vec{p})$ and $\upsilon_s(\vec{p})$ are the spinor wavefunctions corresponding to a fermion and an antifermion.

It is known that this expansion is different in different books. For instance, if I introduce a normalization complex constant $\mathcal{N}(\vec{p})$, I can write

$\Psi(x)=\sum_{s=\pm}\int\frac{d^3\vec{p}}{(2\pi)^3(2E_p)} \mathcal{N}(\vec{p})\Big[ u_s(\vec{p})~ b_s(\vec{p})~ e^{ip\cdot x}+ \upsilon_s(\vec{p})~ d^{\dagger}_s(\vec{p})~ e^{-ip\cdot x} \Big]\\bar{\Psi}(x)=\sum_{s=\pm} \int\frac{d^3\vec{p}}{(2\pi)^3(2E_p)} \mathcal{N}^*(\vec{p})\Big[ \bar{u}_s(\vec{p})~ b^{\dagger}_s(\vec{p})~ e^{-ip\cdot x}+ \bar{\upsilon}_s(\vec{p})~ d_s(\vec{p})~ e^{ip\cdot x} \Big]$ I am writing down the mode expansion for the spinor field in all possible different conventions. Indeed, for instance, in Srednicki's book $\mathcal{N}(\vec{p})=1$, while in P&S it is $\mathcal{N}(\vec{p})=\sqrt{2E_p}$, etc. I would like to understand how to transition from one type of conventions to another.

My thoughts so far: it is known that the equal-time anti-commutation relations for the spinor field is (at least in Srednicki's book) ${\Psi_{\alpha}(t,\vec{x}),\bar{\Psi}_{\beta}(t,\vec{y})}= (\gamma^0)_{\alpha\beta}\, \delta^{(3)}(\vec{x}-\vec{y})$ So, I guess this must certainly be satisfied, and imposing this should give me a relation between $b_s(\vec{p})$ and $b^{\dagger}_s(\vec{p})$, provided I choose appropriately the normalization for the corresponding spinors $u_s(\vec{p})$ and $\bar{u}_s(\vec{p})$.

Is the normalization choice for the spinors something I choose? Therefore, I have the freedom to choose $\mathcal{N}(\vec{p})$ and the normalization for the spinors, i.e. $\bar{u}_s(\vec{p})u_r(\vec{p})=\mathcal{A}(\vec{p})\delta_{rs}$, for instance? Or is $\mathcal{A}(\vec{p})$ related to $\mathcal{N}(\vec{p})$ somehow, with only one of them really be independent?

What is the minimum amount of conditions that must be satisfied by the spinor fields in all conventions?

Any thoughts will be appreciated.