Which primordial power spectrum is fed into CMBFAST?

[Yu-Hsiang Lin]

Hello,

I'm sorry that I'm asking a question about CMBFAST instead of CAMB, because I don't know anywhere else I can ask for help. I hope someone could give me some advice.

I need to feed different primordial power spectra into the CMBFAST code (version 4.5) to compute the temperature anisotropy, but I'm not sure which primordial power spectrum is the correct one.

For example, should I adopt the spectrum of the metric perturbation "psi" in conformal Newtonian gauge, or the spectrum of the "curvature perturbation" (a linear combination of the metric perturbation and the inflaton field -- the one that Weinberg proved to be conservative far outside the Hubble horizon) in synchronous gauge, or something else?

Thank you!