2015 PC Screenshot Thread of the Only Place Where Compression Isn't

NattyDread said:
22764859095_b20d11463c_h.jpg
Click to expand...
Epic
dr_rus said:
avp_dx11_2015_11_06_2z0saq.jpg
Click to expand...
...C...Cortana??
 
Auto-Reply said:
I've played this on 360 back in the day, would love to replay it again (not the best game in the world but also not terrible) on pc but it doesn't seem to be on Steam :(
You are playing the retail copy or ist it available digitally somewhere else?
Click to expand...

I think the gamer edition, the 360 port rather than the xbox/ps2 one, was only available as a digital download. Even when it was out it was hard to come by, so good luck finding it. :(

But yeah, I thought it was pretty good. Certainly one of the better movie tie-ins ever. (Though, that's not saying much.)
 
I can finally contribute to this thread. From the Street Fighter 5 beta with some custom colors done by me. Of course my pic looks like crap compared to y'all stuff. Resolution: 1920x1080p
 
First of all: hats off to you, Natty for the magnificent shots of e.g. TW3! I feel like a fumbling amateur with my pictures haha.

NattyDread said:
It was hard for me, trust me. :) Nobody wanted to help on Reshade forums so I had to figure it out myself.
Click to expand...
Hmm. Yeah they're not very 'friendly' at times and very protective of what they release themselves (but is often made by others). Also odd at times. E.g. I ported Marty's new Advanced DOF to reshade framework, but they didn't want it, as they're busy rewriting things. That's cool but people now can't use that new DOF he wrote. ....

I tried adding depth = -depth; below the linearization code just now and it did something to the DoF but the AO was still inverted (shadowing on the inside of the objects).

Thanks for pointing out that dither method. I was seeing some noise in the "near blur" area of DoF when using low nearblurcurve values in games using log depth. Changed dither_bit = 8.0 to 16.0 and now it's perfect.

Does that thing I did in McFx Util.h make sense? It did fix the AO for me but is there a better way of doing it?
Click to expand...
Could you point me to the thread? (the forum there doesn't allow me to search it seems).

The SSAO.h uses the GetLinearDepth function in McFX\Util.h. To get inversed depth in SSAO, you have to change the contents of that function (so the lines inside {}) to
Code: 
	float toReturn=0.0;
	#if RFX_LogDepth 
		toReturn = (exp(depth * log(2f)) - (1f-0.0045*(1f-pow(abs(depth * log(300f)),100))))/0.056f;
	#else 
		toReturn = 1 / ((depth * ((zFarPlane - zNearPlane) / (-zFarPlane * zNearPlane)) + zFarPlane / (zFarPlane * zNearPlane)));
	#endif
	// inverse
	return -toReturn;

22878939031_eaf3f770e4_o.png
 
OtisInf said:
Hmm. Yeah they're not very 'friendly' at times and very protective of what they release themselves (but is often made by others). Also odd at times. E.g. I ported Marty's new Advanced DOF to reshade framework, but they didn't want it, as they're busy rewriting things. That's cool but people now can't use that new DOF he wrote. ....
Click to expand...
So that's what it was...I'm such an idiot, thought I was being ignored. :)
I can understand that. Although they did use a lot of other peoples work - it probably took a lot of time to port and optimize all that to ReShade.


OtisInf said:
Could you point me to the thread? (the forum there doesn't allow me to search it seems).
Click to expand...
Here.

OtisInf said:
The SSAO.h uses the GetLinearDepth function in McFX\Util.h. To get inversed depth in SSAO, you have to change the contents of that function (so the lines inside {}) to
Code: 
	float toReturn=0.0;
	#if RFX_LogDepth 
		toReturn = (exp(depth * log(2f)) - (1f-0.0045*(1f-pow(abs(depth * log(300f)),100))))/0.056f;
	#else 
		toReturn = 1 / ((depth * ((zFarPlane - zNearPlane) / (-zFarPlane * zNearPlane)) + zFarPlane / (zFarPlane * zNearPlane)));
	#endif
	// inverse
	return -toReturn;
Click to expand...
Great. Thanks. Looks more civil than what I did. :)
I noticed for a while now a cutoff on the near depthbuffer (something like 1m from the camera) and managed to fix it by setting log(300f) to log(1f) comparison
Hopefully that didn't mess up anything else.


Darksiders 2 || 2160p || ReShade beta || jim2point0's CT || nvidia insp SSAO & AF || Panorama (12 shots) ||


Darksiders 2 || 2160p || ReShade beta || jim2point0's CT || nvidia insp SSAO & AF || Panorama (20 shots) ||

^not the Deathinitive Edition
 
NattyDread said:
So that's what it was...I'm such an idiot, thought I was being ignored. :)
I can understand that. Although they did use a lot of other peoples work - it probably took a lot of time to port and optimize all that to ReShade.
Click to expand...
Not sure, probably. But wont cost them extra time to stay polite, something they sometimes forget.

Here.
Click to expand...
Thanks. I've answered there and corrected a small error you made.

Great. Thanks. Looks more civil than what I did. :)
I noticed for a while now a cutoff on the near depthbuffer (something like 1m from the camera) and managed to fix it by setting log(300f) to log(1f) comparison
Hopefully that didn't mess up anything else.
Click to expand...
This is for AO I presume? The linearization math is completely different in the SSAO code than in the DoF code (which uses the Reshade linearization logic). The numbers in the code look like results of trial/error to me and depend on the game used (as the logarithmic spread of the depth differs per game, e.g. what's defined as 'infinite'). So changing it for the witcher might result in good results there but not the same results in e.g. skyrim or GTA ;). But good you fixed the issue :)

22889809871_af2a74fbfd_o.png

22878645485_4b54a082b9_o.png
 
Clean pic (not on full max options [not even on 1080p]. I don't have a PC for that :c )

310560_2015-11-08_0001asra.jpg





And with Reshade (with Film Grain and CA on. Sue me!)

drt2015-11-0818-18-07n6soj.jpg
 
OtisInf said:
Hmm. Yeah they're not very 'friendly' at times and very protective of what they release themselves (but is often made by others). Also odd at times. E.g. I ported Marty's new Advanced DOF to reshade framework, but they didn't want it, as they're busy rewriting things. That's cool but people now can't use that new DOF he wrote. ....
Click to expand...
Please note that this attitude to you was only comming from one person, which I personally do not support in any way. I would have stepped in, but I was on vacation and read the thread (the discussion on reverse-engineering) two days too late. I'm on your side here, since ReShade would not have been possible without reverse-engineering at all.
I'm also truly sorry your DOF port wasn't used and while I would have, I don't have the sole power anymore: I handed over the management of all shaders to the ReShade Framework team, so I personally can concentrate on ReShade itself.
In short, please don't generalize the attitude of a few on the entire project. It does hurt to see that. =)

lifeisstrangepfskn.png
 
You must log in or register to reply here.

Similar threads

  • Poll
2025 PC Screenshot Thread of No Compromises | "Can't see fake frames on screenshots"
|OT|  2 3 4
195
14K
CGNoire replied
"Only in gaming" thread
2
65
2K
JustaStupidBoulder replied
Steam is the worst place to play retro PC games
4
503
Krathoon replied
Rumor: Gears of War: Reloaded physical coming to PS5 & PC only
2
82
400
DenchDeckard replied
Why is Digital Foundry the only place that feels like it has a standard in the gaming journalism sphere
2 3
129
245
Rocco Schiavone replied
Top Bottom