> On 2017-11-06, at 16:54, Mia Magnusson <mia@plea.se> wrote: > >>>> 1) why make it more difficult and expensive while at the same time >>>> potentially less reliable and slower to lock? >>> >>> If you only feed h+v-sync into the computer but let the computer run >>> the pixel clock of it's internal oscillator then you will have >>> jitter of atleast one 28MHz clock cycle. >> >> Now it's me "?" ... > > Which part is hard to understand? I don't understand why would one want to do either this, or the second one (clock only). Maybe for engineering fun, I agree. But practically? The brilliant engineering guys at CBM allowed Amiga to be genlocked and it worked very well. I didn't want to feed only syncs and let the machine run on its own clock as it has no means of adjusting the clock internally. This just doesn't make much sense to me. I wrote that you should be able to get a good lock w/o feeding the Amiga with reference pulses but you still need to feed your innovative genlock peripheral with them so that it could have something to compare with and lock on to. So it's basically adding complexity and intermediaries for no gain, isn't it? >>>> 2) what for - if there is nobody willing to listen to those sync >>>> pulses anyway? >>> >>> ? >> >> You wanted to sync the output pulses from the computer. The question >> is what would you use those for if nobody is willing to listen to >> them. The whole "house" listens to "house sync", not to what one of >> the slave appliances in the house is trying to say. > > A PLL is used to control a 28MHz VCO feeding the 28MHz input so the > outputed h+v-sync from the computer is in sync with the "house sync". That's what I understood, and I agree that it should be possible (provided Amiga is capable of accepting clock and still emitting syncs). But you still need to first "align" the two. Meaning you have to slow down or speed up the Amiga clock and wait for the syncs to align their phases. This takes time. Not much but still. Then you need to constantly monitor the two for drifting and react accordingly by either speeding up or slowing down - basically a form of PLL. > This is the only way to generate a jitter free genlocked picture. ? I guess either it is not the only way or the jitter you refer to is negligible. > Of course the computer would sync faster if you also feed it h+v-sync, > but that isn't strictly necessary as you anyway need to feed 28MHz. 1) smf wrote that there may be no way for the Amiga to accept the clock and output the sync. I haven't tested that but if true, then this alone might void the whole idea. 2) Even if it is possible for the Amiga to accept external clock and emit sync at the same time - what would be better if it was done this way rather than giving it an accurately calculated clock and reference sync to lock on to? -- SD! - http://e4aws.silverdr.com/ Message was sent through the cbm-hackers mailing listReceived on 2017-11-06 19:00:03
Archive generated by hypermail 2.2.0.