I am working with SyncIn and SyncOut chops and have been wondering about the optimal practices for applying the sync frames for video playback & other processes.
For example my current set-up is…
Host machine has a /base with a time component running independently; through a timeline operator, we feed current frames into a SyncOut chop.
Server(s) have the realtime flag off.
They each have a /base with a MODIFIED time component.
The modification comes from inserting the SyncIn frames from the Host into the switch operator called op(‘independent’) and having that switch toggled to pass the sync frames.
This appears to work as intended, and the idea being that we understand this to mean all processes within the base are then being cooked in syncronization with the Host.
Is this accurate, is this ideal, and if not - or even if yes - what are some other practices used for applying sync frames?
Hey,
That workflow is fine, if everything in your file is working based on the me.time.frame number. Often what people do is leave the timeline alone since they arn’t using it directly, and instead pipe the sync values directly to the nodes that are being driven by them (frame number for a movie, transform values for an object, etc).
Or, like malcolm says, not add Time COMPs to your receiving processes, but send a frame number you have more control over, like from a Timer CHOP in the sending process, through Sync Out CHOP, to the Sync In CHOPs in the receiving processes/machines.
Thanks guys! I think I have something working well for this process now - I have gone with supplying the frames directly to the video playing back - more in line with what you are suggesting Greg!
I may dig back into syncronizing clocks in the future and circle back if I run into any snags.