One thing that might help to reduce the degree of randomness and the amount of trial and error in the process would be to search "virtual systems" for the specific make and model of the two components that are being connected, and seeing what kinds of cables other people are using to connect those specific components.
I would suggest focusing particularly on the make and model of the CDP, because to the extent that cable differences are dependent on external factors, they are likely to depend more on the component that drives the cable than on the destination component.
Regards,
-- Al