I felt it was time to make my return to the forums with an AMAZING idea...

  • Thread starter Thread starter yarLson
  • Start date Start date
Status
Not open for further replies.
that is very strange it didn't compile for me in dolphin. which revision are you using
 
3.5-1552 It's only a few days old. I tend to update every 2-3 days.
 
hmm well it compiles in dolphin with no errors but still doesn't seem like it is working for me.  This is strange maybe my drivers are still jacked.  The text should not be pixelated after a 5x upscale of xbr
 
I have the internal resolution maxed out and "force texture filtering" is UN-checked.

Edit: This filter doesn't like to behave. I tried it with Metroid Prime, Wind Waker, and Twilight Princess GC afterwards. None of them showed any difference besides my original test on Animal Crossing.
 
Last edited:
how strange. This has really been giving some headaches.  I wish someone with more experience would get involved
 
how strange. This has really been giving some headaches.  I wish someone with more experience would get involved
Click to expand...
Me too... All of this is exactly what I've experienced. There are a couple shaders I can get to work in that they do SOMETHING with aalis driver, but they mostly look like junk or run REAL slow. There is a good effect you can get off the text from one or two of them, but they grind my system too bad. They work faster with forced AA off but that's a step backwards.

This would be particularly useful for playing the new releases as they are. They are perfectly fine... EXCEPT for the backgrounds.

I was thinking lets just make a 2GB lgp file or whatever but I don't actually know the structure of the backgrounds in the file system, and I know that would probably not be possible just because of the size.
 
The current enhancement (Omzy's field pack) is 3GB. That's not a problem because they're loaded as .png's from the mods folder. For an unconverted re-release, it would take a bit more work to batch-upscale them and then put them back (there's no lgp size constraint though). A shader would be more ideal.
 
Last edited:
The current enhancement (Omzy's field pack) is 3GB. That's not a problem because they're loaded as .png's from the mods folder. For an unconverted re-release, it would take a bit more work to batch-upscale them and then put them back (there's no lgp size constraint though). A shader would be more ideal.
Click to expand...
I have to agree with you because forced-AA doesn't work right with the unconverted steam release, but I haven't tested the one from last year in that regard... I wish someone would make a really smashing shader just for aali's driver, but it is possible the shaders, no matter how specifically written, won't work because something isn't right with the custom driver's implementation of postprocessing itself. And then, we have AA working correct and Omzy's fieldpack.

For the unconverted rereleases, is it possible to rewrite the shaders to do all this complicated upscaling?
 
You don't need to "re-write" per say. You can just use any HLSL shader (I think that is what the re-release uses) and rename it to replace the default one.
 
Totally OT but I have to ask:
Is there a list of emulators supporting that filter?
Esp. links are welcome.

I'm currently trying to play FF IX with that ePSXe thingy (lags like hell, since the drive isn't constantly spinning... but I don't know, if I'm allowed to rip the game even if it's my personal original...) so that filter would be a very nice addition.
 
Does anyone know exactly which shader version the new FF7 steam is using? It is HLSL, I am assuming FX_4_0.

I ask because I've almost ported a shader to the steam version.  Hitting an annoying issue tho, need to dig deeper.

It might be ps2_0 if so I might be hitting instruction limit...
 
Last edited:
Here's what i have to say: Do it yourself. Seriously. Anyone can have ideas, does neither mean yours is either more amazing nor does it mean it's more important. And it doesn't mean other people have implement them for you. People have to do sh*t. Before you have a big mouth, think about doing it yourself. Nobody here is your caregiver.

Instead of begging for realization, what about doing it yourself, smarty pants? ;)

Anyone here has great ideas, myself included, doesn't mean anyone other than you will put them to reality. Also there are more efficient upscaling methods, see: http://www.extremetech.com/gaming/84040-depixelizing-pixel-art-upscaling-retro-8bit-games

http://web.archive.org/web/20070805030008/http://www.hiend3d.com/smartflt.html

http://www.redfieldplugins.com/filterPerfectum.htm
 
Last edited:
I am severely confused by the meaning of insulting him over a question about technical information.
 
Ditto, I'm confused as well. Fischkopf, who are you talking to? If it's the OP, then you need to reread it carefully.

You linked the Perfectum filter as a "more efficient upscaling method" when Omzy and myself have already implemented the Perfectum 2 filter to field, battle, world, and magic textures. The OP is not fully satisfied with this method (which I agree) because it is NOT more efficient and is trying to show us a better upscaling algorithm that can be implemented within a shader, or in our case Aali's Driver, to enhance textures without having to deal with the manual labor and file size increases with PNG upscaling.
 
Last edited:
Besides the whole point of the forum is for community and peer input on the projects people are working on. Your close minded comments basically negate the purpose of the forum. You don't have to agree with what others are doing, but sprouting nonsense like this is just unnecessary.

I'd personally like to contribute to this project but I don't know enough about the tools being used. Yet another good reason for asking technical questions.
 
Here's what i have to say: Do it yourself. Seriously. Anyone can have ideas, does neither mean yours is either more amazing nor does it mean it's more important. And it doesn't mean other people have implement them for you. People have to do sh*t. Before you have a big mouth, think about doing it yourself. Nobody here is your caregiver.

Instead of begging for realization, what about doing it yourself, smarty pants? ;)

Anyone here has great ideas, myself included, doesn't mean anyone other than you will put them to reality. Also there are more efficient upscaling methods, see: http://www.extremetech.com/gaming/84040-depixelizing-pixel-art-upscaling-retro-8bit-games

http://web.archive.org/web/20070805030008/http://www.hiend3d.com/smartflt.html

http://www.redfieldplugins.com/filterPerfectum.htm
Click to expand...
if this is indeed an insult towards me I have a few things to say.  I am still working on implementing it and thanks to my recent studies I think I am getting closer, I might even have finished it already if there was a specification for aali's shader.  I already have the shader I need in modern glsl format but it doesn't seem to work.  In the meantime I have been working on building my programming chops in general.  The whole point was to just guage the reaction of the community and see if anybody wants/cares enough for me to bother.  In any case, the newer xbr level 4 shader actually surpassed the quality of the "depixelizing pixel art" algo, but there isn't too much info or screens available to show you (I only have it since the original author was nice enough to share it with me, it is not yet complete and therefore not available), maybe I'll post some of my own later.  In any case even the publicly available level 3 version is very amazing but its new so not too many screens of that one either.  Just fireup RetroArch and download the lvl 3 multipass cg shader from the repo and you can see for yourself.

In any case the algorithm you linked to isn't really productive much since there is no public implementation and I, being still an intermediate programmer, probably couldn't pull off implementing from scratch without any help.  And it would be rather pointless even if I did since the authors themselves say it is too slow for real time implementation.  Xbr is certainly the best choice for this experiment.  I don't know why you had to go blowing smoke before you even know what your talking about.
 
The whole point was to just guage the reaction of the community and see if anybody wants/cares enough for me to bother.
Click to expand...
Me, me, me. I care, which is all the reaction you need to do a project  :P
 
Status
Not open for further replies.
Back
Top