Edtv passthrough

[goutsune]

No description provided.

[gingerbeardman]

Please consider merging this @ramapcsx2

[ramapcsx2]

Ah, I looked at this and there were problems. A lot of talk required here ><

[goutsune]

Ah, I looked at this and there were problems. A lot of talk required here ><

I'm all ears, I have been using gbs-control like this for about an year with PS2 and Wii, but perhaps I forgot to change some pre/post preset flags?

[ramapcsx2]

Just found this old PR ><
Okay, I think my worry here was that I have too many obfuscated program flow paths. I'm not sure modifying the conditionals will work in all other cases (other than EDTV passthrough).
This is a bad code problem, and I can't even test it all now.
So basically, I'm worried that this will break something unintentionally, and be forever hard to debug.

[goutsune]

I understand that code is messy and mine no better, so I think it's ok to let people test it and decide :D
I have been using GBS like this for a few years (+ I think I reduced some of input signal test delays to make CRT go to standby faster that in 5 or so minutes).

On a side note, I have been pondering around possibilities of true dynamic preset genetaion based on analyzed input signal — I noticed there are quite some test registers on GBS and I think it is possible to exploit clockgen remaining channels to try and catch pixel clock, kinda how LCD monitors do that when auto-matching screen.

[ramapcsx2]

I think I've probed the registers thoroughly, but it remains truly black box reversing. If you find anything new that can be used, great, but I also already tried hard :)