I'm beginning the effort to add genlock/framelock support into our video design and am wondering how this functionality works in a system with multiple inputs and/or outputs. This has brought up two issues thus far:
1) The current system we have has 4 inputs and a single display. The technique for doing genlocking described in the documentation says to enable the SOF generation in the CVI and the SOF input of the CVO. What happens if I have multiple CVIs? The SOF connection is "invisible" to the user so I have no way of controlling how it flows from the CVI to the CVO.
I did notice that the CVI has the option to generate the SOF reference signals only. Is this telling me that I need a 5th CVI to use as my genlock source and that I must mux my inputs externally to the Qsys block to pick off the one I want as the reference? If so then this begs the question as to why I would burden the system with a VIP switch and multiple CVI blocks if I have the mux outside...
2) How does this work if you have multiple CVO blocks? We have the potential to require 2 or more CVO blocks that need to be locked together. Assuming I resolve issue# 1 then by enabling the genlock feature in the CVOs is the input side correctly routed to all of the CVO blocks? If so, and if they're running off of the same clock, how close can I expect their alignment to be? Should they be exactly in time with each other or is there some skid expected between them?
