ENBSeries

The Elder Scrolls Skyrim (TES Skyrim, Скурим Свитки)

Graphic mod ENBSeries / patch ENBoost 0.227

Added water category in enbseries.ini. Fixed some particles to allow edit them.
Water have many bugs which almost invisible in vanilla game, but not allow to modify easily, so i'll add various features for it from time to time.

Particles work again!

Thanks a lot Boris, and looking forward to this new water stuff!

This question is for Boris or anyone else that happens to know. I performed a moderate search and couldn't find the answer to my question. This was mentioned a few releases ago, so this forum seemed the appropriate place to ask.

My question is about SSAO SizeScale and SourceTextureScale. I understand that the lower they are (under 1.0), the greater the likelihood that shadow artifacts will occur. It also seems to have a greater effect on performance than Sampling Precision and Quality. Boris stated in a previous thread that lower values can be used at higher resolutions, yet I see ENB authors using all kinds of values: 0.57, 0.81, 0.87...as if some formula were being used to determine these values. Others use seemingly arbitrary values. I'm curious if there is an actual formula that I can use to determine the appropriate scales to use for a given resolution. I usually play at a resolution of 2560x1600, and I'm assuming that most presets are configured for 1920x1080, so it would be helpful to understand whether this number can be set to an appropriate value for a given resolution and how that is determined.

Regards

Maeldun
I asked the very same question a while back.
Take a look here:
viewtopic.php?f=28&t=2403

Sry for off-topic.

Maeldun wrote:This question is for Boris or anyone else that happens to know. I performed a moderate search and couldn't find the answer to my question. This was mentioned a few releases ago, so this forum seemed the appropriate place to ask.

My question is about SSAO SizeScale and SourceTextureScale. I understand that the lower they are (under 1.0), the greater the likelihood that shadow artifacts will occur. It also seems to have a greater effect on performance than Sampling Precision and Quality. Boris stated in a previous thread that lower values can be used at higher resolutions, yet I see ENB authors using all kinds of values: 0.57, 0.81, 0.87...as if some formula were being used to determine these values. Others use seemingly arbitrary values. I'm curious if there is an actual formula that I can use to determine the appropriate scales to use for a given resolution. I usually play at a resolution of 2560x1600, and I'm assuming that most presets are configured for 1920x1080, so it would be helpful to understand whether this number can be set to an appropriate value for a given resolution and how that is determined.

Regards

Well, I've been the first to use values like " 0.7071067811865475 " (wich equals 0.5 computation, value beeing its square root) and everyone laughed at me at that moment
The fact is there is no perfect formula, only choices.
Boris explained long time ago that all that was about a balancing act between scales and filter / quality levels.
You can choose between what's *enough", like these 0.7071067811865475 scales coupled with 1 filter and 0 quality at close range (0.25) @1080p with Skylighting to complement shadowing on wider range in exteriors, and what can be gorgeous, like 1.0 scales (full computation) with 1 filter and quality with a bit wider range (0.3)...
@1600p, 0.5773502691896258 scales (0.3 computation) coupled with 0 filter and Q are ENOUGH, math wise.
But fact is what's NOT recommended can be often fantastic
For example, I'm currently using 1.183215956619923 scales @1080p, 1440p for my gaming and when I want to achieve gorgeous lighting for particular screenshots, I've been pushing my scales MUCH further, like 1.414213562373095, wich means 2x computation, so the algorythm computes 'outside' of screen size, given pixels number. But it'll KILL your performances, litterally.

Boris told me one day " Free your mind ".... I remembered this
Experiment, trial and error, and trust your eyes !

Oyama
You really do like long numbers... =P

I think scales and SSAO performance will differ from rig to rig and config to config. My best performance is from (i think it was) 85% of screen, SamplingQuality=1 while SamplingPrecision and FilterQuality was set to 2. Range was 32 I believe.
Great blackness with the best performance for my setup and no flickering when camera moves, even better with the new noise filter.
Like Oyama says: trial and error and chose the setup that looks best in your eyes.

Version updated, download again. Fixed that damn sky for water reflections when panning third person camera. Unfortunately too far movement from player do not work (may be i'll fix later, game simply won't render anything, except top of cubemap).

Thanx Boris !

@Oyama and insomnia_jt: Some excellent tips to add to my growing list of bookmarks. I swear I need to make a collective post since so much good info is spread throughout the screenshot thread.