1024, 2048, 4096 texture resolution

[N95KF]

Good morning,

Question for Skuzzy and whoever else has knowledge on this topic.  I am running AH3 at 4096 texture resolution in the main Video Settings.  Since I only play on a 1920 x 1080 screen, does this have any effect?  Thanks

[Chalenge]

Yes, it affects the quality of the textures applied to objects in the game, and the higher the setting it also consumes more memory. You may also gain a few frames (FPS) by dropping the memory size down. Each increase in texture size requires four times the memory for each texture, for instance.

[Skuzzy]

The larger format allows for more detail and better smoothing.

While this was done for Aces High II, it still shows how texture size impacts the visual quality.

https://www.youtube.com/watch?v=aqS6-QfWbHQ

(We will be updating all those videos to AH3 standards).

[Warty]

I get very jaggy, and a bit shimmery effects from the in-cockpit shadows. Would bumping max res up to 4096 help with that? I honestly can't tell difference so far from 1024 to 2048, for looking at cockpit, my wings, etc. But the shadows are very jaggy. (but cool

[100Coogn]

I get very jaggy, and a bit shimmery effects from the in-cockpit shadows. Would bumping max res up to 4096 help with that? I honestly can't tell difference so far from 1024 to 2048, for looking at cockpit, my wings, etc. But the shadows are very jaggy. (but cool

That will probably help some.
My shadows are always jaggy though, even at 4096.

Coogan

[N95KF]

I get very jaggy, and a bit shimmery effects from the in-cockpit shadows. Would bumping max res up to 4096 help with that? I honestly can't tell difference so far from 1024 to 2048, for looking at cockpit, my wings, etc. But the shadows are very jaggy. (but cool

Try setting your AA settings in your AMD or NVidia control panel too.  See if that helps.  Make sure your resolution fits your monitors native resolution as well.

[Skuzzy]

The shadows are gong to be jaggy.  We are not crazy about it either.

[Dobs]

The shadows are gong to be jaggy.  We are not crazy about it either.

Is it due to a higher workload required to smooth the shadows that they are going to be jaggy?   I'm just curious because the objects casting the shadow look smooth.

[Bizman]

Is it due to a higher workload required to smooth the shadows that they are going to be jaggy?   I'm just curious because the objects casting the shadow look smooth.

I've been wondering exactly the same.

[Pudgie]

Is it due to a higher workload required to smooth the shadows that they are going to be jaggy?   I'm just curious because the objects casting the shadow look smooth.

Well I've been doing some testing and when I've applied GPU side supersampling AA to enhance the in-game post processed FXAA the jaggies were greatly diminished vs using the post processed FXAA alone..............(Crimson driver AA set to "Enhance Application Settings" then AA method set @ Supersampling)
I've also disabled the in-game FXAA then set AA up at the Crimson driver level to "Override Application Settings", set max AA level @ 8xEQ (same as 16x), max AA filtering @ Edge Detect and max AA method @ Supersampling (all this is applied at the GPU side) and the game graphics look the absolute best including the jaggies when AA is applied thru the GPU vs FXAA on the post processed side (shaders). Now when I use these settings w\ AH III Dx11 version I get no GPU performance slowdown at all but when I run these settings thru AH III Dx9 version the GPU performance tanks badly to utterly unplayable levels...less than 10 FPS. I have recorded graphs of all this as well for records. This is also why I suspect some issue(s) between the 2 versions of Dx....IMM the GPU performance disparity shouldn't be this wide between the 2, especially w\ the greater performance disparity being towards the older Dx version.

Would\could this also be just due to post-processed FXAA not being as able to do as good a job of antialiasing as the GPU applied antialiasing at this time? 1 of the advantages of using post processing graphics rendering techniques is to offload the GPU of this work to improve overall graphics performance............correc t?

Or is it just a matter of the choice of FXAA to use? There are several levels of FXAA to choose from as I understand it (also have seen thru using Radeon Pro w\ earlier Catalyst drivers and thru research on FXAA) to be used w\ varying pros\cons.............

Just putting this out here to chew on................

 

[Chalenge]

I tested the same way and saw no performance hit for AF and other settings, but DX9 suffers (I think) from using earlier methods of AA. So, it was my conclusion that confusing the AA versions in DX9 by forcing an override was causing a performance hit, especially when I added SLI enhanced AA. At least with DX9 it seems like a good idea to leave everything to AH, except perhaps AF.

[Pudgie]

I tested the same way and saw no performance hit for AF and other settings, but DX9 suffers (I think) from using earlier methods of AA. So, it was my conclusion that confusing the AA versions in DX9 by forcing an override was causing a performance hit, especially when I added SLI enhanced AA. At least with DX9 it seems like a good idea to leave everything to AH, except perhaps AF.

Agreed, but as you also know this will require some setting changing from 1 version to the other to take advantage of it which could be a turnoff.
What I did was to set all this up to find an acceptable compromise between the 2 to alleviate this for the time being but I'm wanting to lean towards going w\ AH III Dx11 w\ driver applied AA to get the best of all but the freezing\pausing was stopping me from doing that.