From optical spectroscopy of X-ray sources observed as part of the Chandra Multi-wavelength Project (ChaMP), we present redshifts and classifications for a total of 1569 Chandra sources from our targeted spectroscopic follow-up using the FLWO/1.5 m, SAAO/1.9 m, WIYN 3.5 m, CTIO/4 m, KPNO/4 m, Magellan/6.5 m, MMT/6.5 m, and Gemini/8 m telescopes, and from archival Sloan Digital Sky Survey (SDSS) spectroscopy. We classify the optical counterparts as 50% broad-line active galactic nuclei (AGNs), 16% emission line galaxies, 14% absorption line galaxies, and 20% stars. We detect QSOs out to z ~ 5.5 and galaxies out to z ~ 3. We have compiled extensive photometry, including X-ray (ChaMP), ultraviolet (GALEX), optical (SDSS and ChaMP-NOAO/MOSAIC follow-up), near-infrared (UKIDSS, Two Micron All Sky Survey, and ChaMP-CTIO/ISPI follow-up), mid-infrared (WISE), and radio (FIRST and NVSS) bands. Together with our spectroscopic information, this enables us to derive detailed spectral energy distributions (SEDs) for our extragalactic sources. We fit a variety of template SEDs to determine bolometric luminosities, and to constrain AGNs and starburst components where both are present. While ~58% of X-ray Seyferts (1042 erg s-12 - 10 keV44 erg s-1) require a starburst event (>5% starburst contribution to bolometric luminosity) to fit observed photometry only 26% of the X-ray QSO (L 2 - 10 keV >1044 erg s-1) population appear to have some kind of star formation contribution. This is significantly lower than for the Seyferts, especially if we take into account torus contamination at z > 1 where the majority of our X-ray QSOs lie. In addition, we observe a rapid drop of the percentage of starburst contribution as X-ray luminosity increases. This is consistent with the quenching of star formation by powerful QSOs, as predicted by the merger model, or with a time lag between the peak of star formation and QSO activity. We have tested the hypothesis that there should be a strong connection between X-ray obscuration and star formation but we do not find any association between X-ray column density and star formation rate both in the general population or the star-forming X-ray Seyferts. Our large compilation also allows us to report here the identification of 81 X-ray Bright Optically inactive Galaxies, 78 z > 3 X-ray sources, and eight Type-2 QSO candidates. Also, we have identified the highest redshift (z = 5.4135) X-ray-selected QSO with optical spectroscopy.
