@zing Thanks! I've just done it
Greetings!
@zing Thanks! I've just done it
Greetings!
Hi everyone!
I was testing newer versions of those cores for using on the GPI Case (pi zero), and they have some benefits over recalbox current versions.
snes9x2002:
recalbox version: a869da7f22c63ee1cb316f79c6dd7691a369da3e
proposed version: latest (48b67ee60cf32d587b2d3e9f4cd37c84c647666e)
The most important change is the "Backport colour operations from Snes9x 1.60" which greatly increase performance, making almost as fast as pisnes with the newer features of retroarch
snes9x2005:
recalbox version: 743ceea46479bafd3aeb3dbd32daeb716dff7456
proposed version: latest (71db1be367d780caa6128bc2a3c3bd33567baf5a)
The "Backports: Colour operations from Snes9x 1.60" and "Add optional automatic frame skipping" changes make it faster.
Even though its not as fast as the 2002 core, Yoshi´s Island works good enough (i tried only one level).
The auto-frame skip feature requires a newer retroarch (1.91 i think)
picodrive:
recalbox version: a869da7f22c63ee1cb316f79c6dd7691a369da3e
proposed version: lastest (8b01ec7e28bf87800d633bc3ed2f0dbb01fb7237)
The most important feature is the implementation of chd format which reduce the size of Sega CD images (almost 50%). I didn't notice a performance degradation with the terminator game.
I could successfully build both Snes cores using the recalbox scrips, and to get bit better performance i set the dispmanx retroarch video driver)
But with the Picodrive core i had to use the retropie rp1 binary (i don't know how to update recalbox script to build it successfully).
if its needed i can create an issue on gitlab for asking the update of said cores.
Best regards and thanks for your great work!!
@danixu86 Hi! I think i found a workaround to make it work
<command>fbset -g 320 240 640 480 16 & python /usr/lib/python2.7/site-packages/configgen/emulatorlauncher.py -system %SYSTEM% -rom %ROM% -emulator %EMULATOR% -core %CORE% -ratio %RATIO% %NETPLAY%</command>
binddev = sdl:Microsoft X-Box 360 pad
bind backspace = Fast Forward
bind escape = Enter Menu
bind e = player1 L2
bind t = player1 R2
bind \xA0 = player1 CIRCLE
bind \xA1 = player1 CROSS
bind \xA2 = player1 TRIANGLE
bind \xA3 = player1 SQUARE
bind \xA4 = player1 L1
bind \xA5 = player1 R1
bind \xA8 = player1 SELECT
bind \xA9 = player1 START
bind up = player1 UP
bind down = player1 DOWN
bind right = player1 RIGHT
bind left = player1 LEFT
bind f1 = Save State
bind f2 = Load State
bind f3 = Prev Save Slot
bind f4 = Next Save Slot
bind f5 = Toggle Frameskip
bind f6 = Take Screenshot
bind f7 = Show/Hide FPS
bind f11 = Toggle fullscreen
Then the emulator works and play games.
But i found that games that dont use 320x240 resolution get gfx corrupted. Im not sure why but even though the games i tried are working at 60fps (frameskip = 3), the sound on my gpi case has a lot of cackling (worse than the libretro core)
i hope it cant be fixed , it would be great to play psx on my gpi case.
Greetings!
Thanks for your answer Zing!
its a new emulator to me, so i wanted to share what happend on my netbook.
But i understand your point, hopefully pcsx reamed can be fixed on the gpi case.
Greetings!
Greetings everyone!
I recently tried DuckStation on an old x64 netbook (dell mini 1012), and it performs better than pcsx-reamed.
The under-clocking option for the psx cpu helps to make more games playable.
For example castlevania s.o.n. works fine in duckstation but its too slow on pcsx reamed.
Im no sure but maybe DuckStation can be added to de pi zero image to make possible psx emulation.
On my gpi case, the standalone pcsx rearmed carsh and the libretro core is too slow with sound cracking.
im using recalbox version 7.2.1 on both machines.
Best regards!
Hi Everyone!
First of all i would like to thanks to the recalbox team for this great system. I use it on my gpi case
Im planning to upgrade to version 7.2 but im wating the fix for the performance issue for gpsp core:
I think its related to the disabling of dynarec informed on the 7.2 changelog:
I think that issue was fixed on upstream recently, maybe upgrading the core to the last version could fix it.
The lastest version was reported by the libretro blog to be even faster.
I hope this information can help.
greetings!