5 minutes ago, mpo910 said:

. PMDG, FSLabs, QW and so far and so on use "own" textures and are NOT affected by this slider.

But they absolutely are. I'm literally flying the FSL 319 right now and can see the difference: https://imgur.com/a/48GKVC1

The difference is even more significant in the NGXu

10 minutes ago, Chapstick said:

But they absolutely are. I'm literally flying the FSL 319 right now and can see the difference: https://imgur.com/a/48GKVC1

The difference is even more significant in the NGXu

Ok....I never did noticed this. Sorry for that. Probably my fault. Learned something again and you reached me 🙂

 

Edited January 6, 20215 yr by mpo910

am i the only one who can't see a difference between these two?  Looking at it on a 4K screen. 

29 minutes ago, Chapstick said:

But they absolutely are. I'm literally flying the FSL 319 right now and can see the difference: https://imgur.com/a/48GKVC1

The difference is even more significant in the NGXu

I gotta say, I can see the differences but find them incredibly slight. I'm actually surprised at how small the differences that you show here actually are. The fact that you have to zoom in like that to make them clear further tells me that under normal operation I probably wouldn't have noticed.

But I already said it here and there: the interpretation of these sorts of things are terribly subjective. To me it is beyond understanding how somebody could not prefer v5's EA lighting over v4's, but some people simply don't seem to care too much. And in this case, the texture resolution seems to drive you nuts, @Chapstick, while I barely see the difference apparently.

EDIT: well, seems @micstatic also has trouble seeing it. The best showcase of it is probably around those buttons on the blue panels right above the main flight displays.

5 minutes ago, micstatic said:

am i the only one who can't see a difference between these two?  Looking at it on a 4K screen. 

 

1 minute ago, Benjamin J said:

I gotta say, I can see the differences but personally do not find them particularly earth-shattering. I'm actually surprised at how small the differences that you show here actually are... It shouldn't be surprising that for me 1024 textures generally have not been a deal breaker in the least.

But I already said it here and there: the interpretation of these sorts of things are terribly subjective. To me it is beyond understanding how somebody could not prefer v5's EA lighting over v4's, but some people simply don't seem to care too much.

This is one example of many. It's not even the worst I've seen. I just so happen to have the FSL up right now. 

1 minute ago, Chapstick said:

 

This is one example of many. It's not even the worst I've seen. I just so happen to have the FSL up right now. 

I'll say that for me the bigger problems have been in terms of scenery textures, and less so aircraft VCs. Particularly ground poly textures. I notice it often clearly there, but rarely care enough because the overall appearance still looks fine to me... The lighting is a much more important thing for me.

I set mine at 4096.  The only time I've ever noticed differences previously is on gate markings and such when looking outside.  But it is an interesting post to see the picture differences.  

There are 4096 airports that in reality are 2048 airports .... 

Like the v1.3 FT EHAM update.

For years I had textures set to 1024 and it never looked unsharp. I will go one step further : to high res textures can make objects look unrealistic.

Just look out of your window to a tree a few meters away.  You will not see the sane sharpness as you see at the edges of sim objects.

For the same reason I use 256 cloud textures to get a soft look instead of 4096 clouds with soft clouds on the texture. Much better franerate and looking better too.

 

I switched from 2048 textures to 1024 textures at UK2000 Bristol 2019HD a few weeks ago....just to see if there was a difference. Whilst the VC itself looked OK, the texture quality of the sign directly in front of my plane was significantly reduced. As soon as I noticed that, I switched back to 2048 textures. Thankfully, I did it so quickly that a full on panic attack was avoided

1 hour ago, GSalden said:

.I will go one step further : to high res textures can make objects look unrealistic.

  I know what you're saying, but that's not what I'm talking about. This is what I'm talking about: https://imgur.com/a/pWXtbUO. You can barely read Lockheed's own logo on the default F-35 at the default texture setting. Is that what they want to showcase?

13 minutes ago, Rob_Ainscough said:

Agree, you're settings are low and I can't see why they would use up that much VRAM.  What are your Traffic settings (Boats, Road, Airport)?  ASP3D will use some VRAM if it's on the same PC.  I don't have AIG so I can't test (use UTLive2) ... have you tested VRAM with AIG disabled and no AI traffic?  Are you using higher quality PBR AI models with AIG?

Cheers, Rob.

AIG doesn't really have an impact on VRAM from my testing. At LHR with 70 a/c, VRAM remain unchanged when I turned AI off.

2 minutes ago, Chapstick said:

  I know what you're saying, but that's not what I'm talking about. This is what I'm talking about: https://imgur.com/a/pWXtbUO. You can barely read Lockheed's own logo on the default F-35 at the default texture setting. Is that what they want to showcase?

AIG doesn't really have an impact on VRAM from my testing. At LHR with 70 a/c, VRAM remain unchanged when I turned AI off.

Agree. Very cpu bound. ORBX HD Buildings in use perhaps?

9 minutes ago, mpo910 said:

Agree. Very cpu bound. ORBX HD Buildings in use perhaps?

nope

9 minutes ago, Chapstick said:

  I know what you're saying, but that's not what I'm talking about. This is what I'm talking about: https://imgur.com/a/pWXtbUO. You can barely read Lockheed's own logo on the default F-35 at the default texture setting. Is that what they want to showcase?

AIG doesn't really have an impact on VRAM from my testing. At LHR with 70 a/c, VRAM remain unchanged when I turned AI off.

It surely depends on what you want. Higher res aircraft textures show a nicer ac.

Trees/Clouds and textures on objects further away at a lower resolution will up performance and from within your ac you won’t see it.

FT EHAM v1.2 had normal and higher res textures mixed. The road side of a terminal used lower res textures.

At this moment with my 3090s I see no difference in franerate with textures set to 1024 or 4096. Perhaps the scenery is not heavy enough for the cards..

Everyone do as you like it 👍

1 hour ago, Rob_Ainscough said:

That definitely should NOT be the case, AI off you should see less VRAM usage.

I tried to replicate with your settings and those add-ons you listed at that airport but I didn't get anywhere near your VRAM usage and that was at 4K res (even tried without Dynamic Texture streaming and 4096 x 4096 texture cap) ... any other things you might have not mentioned?  Mesh add-on?  HiFi settings?  NGXu settings?  Shadow quality tweaks?  TEXTURE_SIZE_EXP tweaks, MAX_TEXTURE_REQUEST_DISTANCE, UseGlobalTerrainView, NCP settings?  

Any change when you turn Wideview off?

Cheers, Rob.

 

No mesh adds, HiFi settings are default, NGXu default perf settings, no GFX tweaks in the CFG besides upping the two light point lines to 800. I use ChasePlane so the widescreen view option is cancelled out anyway.

4 hours ago, Rob_Ainscough said:

Are you using higher quality PBR AI models with AIG?

Yep, if he's using the default AI aircraft texture resolutions for most AI planes then that could significantly affect the VRAM usage.  Most of them use 2048x2048 textures, and some use 4096x4096.  Moreover, many don't have mipmaps.

I downsized all mine to 1024x1024 with mipmaps and noticed improved performance.

Dave