Gambatte: change the order of output from "GB => resampler => metaspu => DCFilter" to "GB => resampler => DCFilter => metaspu".
This doesn't change anything under most circumstances, except when playing the emulator in slow motion (for instance, 50% throttle). There, the metaspu sometimes adds silence to the output, which isn't actually silence if it has a different DC offset than the audio it's being mixed with. Well, 50% throttle sound output will always suck anyway...
all axes are handled as boolbuttons with hardcoded deadzones.
all boolbuttons are handled as expected.
"slider"s are not handled because i had nothing to test against.
button names for axes have changed; button names for boolbuttons have not.
From my reading of the code, this should not affect movie playback; the saveram from disk is not used when starting playback or recording of a clean movie.
If for some reason I'm wrong though, that problem should be addressed directly instead of deleting the file.
bt&dd has a glitch that reads from $6000 when there is no ram installed, and crashes if something too large is returned. I think this chooses a frame for abobo, and you can actually see abobo's sprites glitch for one frame right when the wall finishes getting busted down (the same place where bizhawk was freezing)
-ImportLSMV can now import GB(C) movies created with LSNES's gambatte core.
--It doesn't sync perfectly because of the subframes, but it does for a while, so the conversion is sane.
--GB(C) .LSMV files currently always have SRAM files packaged with it, which is a bug with LSNES.
---Ilari told me that a fix for now would be to ignore 0 byte SRAM files, so that's what I did.
---There are some cases in which the spurious SRAM files aren't 0 bytes, but that's on Ilari to fix, so I'll let this case fail. Yay, I contributed to LSNES!
1) choose your SGB rom from the path config (should probably be an .smc file of about 256KB)
2) turn on Load GB as SGB from the GB menu.
3) load rom again.
To turn off, uncheck Load GB as SGB from the SGB menu.
Fixed Hex Editor's FindNext/FindPrev functions. FindNext/FindPrev now accept an additional argument (bool) that tells them to wrap around once. Fixed a bug where if the matched address was the 1st byte of the array it was ignored.
goyuken