..:: PCSX2 Forums ::..

Full Version: Shadow of the Colossus !!!
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
Pages: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37
(04-04-2010, 12:22 AM)rama Wrote: [ -> ]You guys don't know about the new hacks in GSdx, am I right?
Add "allowHacks = 1" to your gsdx.ini, then you get new items in the configuration dialog.
GOW2 for example gets perfectly fixed with the offset hack enabled. I imagine SotC will look
good or perfect as well.

Wrong.
I know about the UserHacks_HalfPixelOffset=1. I use it, but i doesn't give nearly as good results as the old scaler hack.
The results for both GOW2 and SOTC when using the HalfPixelOffset=1 are:
In the exact center of the picture the layers are perfectly mixed, but the more distant we get from the center, the more badly are the misalignments, making this blurry, while the old hack made super-sharply perfect mixing everywhere.
I'll make some comparison screenshots to make you see my point...
Pattern:
First picture - new scaler WITHOUT HalfPixelOffset=1
Second picture - new scaler WITH HalfPixelOffset=1
Third picture - old scaler. (gsdx 2406)

x4 scaler used for all images, not custom resolution.
(there's a GOW2 example too, since you said that it fixes it perfectly. I wouldn't say that that GOW2 looks fixed in the second picture.)
(03-17-2010, 05:15 AM)Linear Wrote: [ -> ]Okay, rama told me that it's impossible to use initiall pixel offset hack again, so our only hope is - modified scaler :Р

Now the question is... where can we get the source code of gsdx-r2406 ?
On the google code site of pcsx2 I could only download the latest revisions source code...
And what part of the source code determines the scaler?
Any ideas ? Anybody?
Ok, I should've listened to pseudonym and never coded any of these hacks.
(04-05-2010, 02:54 PM)rama Wrote: [ -> ]Ok, I should've listened to pseudonym and never coded any of these hacks.

I don't think your manner solves anything. How would omitting to create a useful feature move things forward?

We're merely trying to achive a custom goal here, that seems quite possible to us. That's it. Don't balme me for saying the truth. UserHacks_HalfPixelOffset=1 doesn't fix offset errors completely. Thats simply reality. Sorry for being honest.
It's not your place to blame my decisions as bad or not helping, and you certainly are in no position to call on my manners.
For your information: I'm the only person on the dev team that supports these hacks, you wouldn't get anything otherwise.
You can either accept what you get, ask for a fix (I denied it, general compatibility is more important that your game) or fix your problem yourself.
I'll leave this thread open for people OTHER than you to contribute.
Maybe someone comes up with an acceptable patch to the problem even.

Oh btw, I couldn't make the game look like in your screenshots at all. It's perfect here.
(04-05-2010, 03:53 PM)rama Wrote: [ -> ]or fix your problem yourself.
That's exaclt ywhat i'm trying to do.

(04-05-2010, 03:53 PM)rama Wrote: [ -> ]Oh btw, I couldn't make the game look like in your screenshots at all. It's perfect here.

Then I must find out what i'm doing different. Any ideas of you?
I use the UserHacks_HalfPixelOffset=1 and the scaler. This is the result with either 2x 3x 4x 5x 6x. (second pictures....)
win7 dx11 mode (since dx10 was removed)
HOLY SH!#%^ !!! IT WORK !!! IT REALLY WORK !!!! OMG !!=) I've just include that allowHack line in gsdx.ini, just as Rama said, then I activated that offset checkbox and it solved all that problems !!! So now SotC fully and PERFECTLY works !!!! Innuendo, Rama, I'am happiest man in the world now !!! Offset fixed, fairy shading and bloom fixed, shadows are fixed - so now I might rest in peace ! (^_^)
Thank you Rama !! Many thanks to you and all people who helped to create this project !!!
PERFECT !! JUST PERFECT !!!
DEVs - you are our heroes !!!
Whoa, Linear, you don't have those small misplacings as I did in the seconds of the three pictures ?
Then why do my results differ ? : \
Linear, you've mentioned earlier you're "still using simple disk changing in startup in order to fool gsdx".
How is that procedure exactly, if I may ask ?
Pages: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37