11 hours ago, SayAgain said:

I would imaging your doing that to keep FPS higher?

What value are you using for TAA?  Important, because any TAA value above 100 switch’s to SSAA … very different AA processing and SSAA is much sharper (higher image quality at lower FPS).

I don’t think many read the hover over on graphics settings for TAA … like I’ve said before TAA up to a value of 100.  If you set TAA above a value of 100, you are not using TAA and are using SSAA (very different AA processing with far better results at the cost of FPS).

 

However, now that I know that objects in the far distance are less/not showing with TAA I will retest.

DLSS on the 2 view server looks good. On the Left view client it is just the edges of the clouds that look better with TAA than with DLSS.

15 hours ago, ryanbatc said:

Ewww.  DLSS blurs glass gauges and when values change they lag badly.

It's fine on analog aircraft but no go with glass.

Not with DLAA and version 310.3 DLL it's razor sharp here (2K 1440P)

46 minutes ago, GSalden said:

 

However, now that I know that objects in the far distance are less/not showing with TAA I will retest.

DLSS on the 2 view server looks good. On the Left view client it is just the edges of the clouds that look better with TAA than with DLSS.

Have you duplicated the improvement in detail at distance with your own settings?
 

As I previously commented, the only thing seemingly demonstrated by the guy in the video is the effect of AutoFPS, which he says he’s using.  This reduces the level of detail at distance based on system demand - and as TAA is more demanding than DLSS, naturally it will trigger AutoFPS to intervene sooner than DLSS.

Edited August 13, 2025Aug 13 by tfm

33 minutes ago, virtualstuff said:

Not with DLAA and version 310.3 DLL it's razor sharp here (2K 1440P)

Indeed, it has been ever since DLSS 3 was released. 

9 hours ago, Ron Lefebvre said:

Ok I will ask.

Why 105 scaling if you are already at 4k

It cleans/sharpens up some more edges, I use a large 42" display, that increases the perception of aliasing. Even with frame gen I'm limited by CPU anyway.

6 hours ago, SayAgain said:

Because that switch’s to SSAA not TAA.  SSAA is an AA process that produces better image quality (less blurry).  Like I keep saying TAA value > 100 switches to SSAA … not sure how many ways I can say this … even Asobo put that information right next to the TAA slider.

I think you misunderstood the information text in the settings, Render Scaling > 100 does not "switch" from TAA to SSAA. TAA and Render Scaling are applied sperately for a combined result. The information text just states that it is recommended to use TAA when using Render Scaling < 100, as hard upscaling is increasing aliasing. But that does not mean you can't use it as well when downscaling (super-sampling).

Edited August 13, 2025Aug 13 by touchdown84

1 hour ago, tfm said:

Have you duplicated the improvement in detail at distance with your own settings?
 

As I previously commented, the only thing seemingly demonstrated by the guy in the video is the effect of AutoFPS, which he says he’s using.  This reduces the level of detail at distance based on system demand - and as TAA is more demanding than DLSS, naturally it will trigger AutoFPS to intervene sooner than DLSS.

Tried to see the difference at FT Amsterdam Schiphol in objects in the far distance (without AutoFPS) and saw no difference 

The default EHAM Is disabled.

Regarding AutoFPS : I do not think that he used AutoFPS for these tests. 

Also AutoFPS might renove objects depending on OLOD but does not make the object disappesr half (cranes) at 1 ml distance ) as far as I have seen it 

1 hour ago, touchdown84 said:

I think you misunderstood the information text in the settings, Render Scaling > 100 does not "switch" from TAA to SSAA. TAA and Render Scaling are applied sperately for a combined result. The information text just states that it is recommended to use TAA when using Render Scaling < 100, as hard upscaling is increasing aliasing. But that does not mean you can't use it as well when downscaling (super-sampling).

Well, using anything above 100% render scaling AUTOMATICALLY forces to use SSAA, because SSAA is nothing else than rendering the image on a higher resolution and displaying it on the native (or selected) resolution, e.g. downscaled. That is why using 105% forces SSAA automatically. At least this is how I understand SSAA.

What I still do not get: theoretically, using 105% and thus SSAA should make it possible to select "none" as antialiasing method, no? Because if you use TAA PLUS 105% renderscale, what happens in such a situation? Is the renderscale applied first? I guess so, so this means you actually get an SSAA image which is THEN also processed by the TAA algorithm? Not entirely clear to me... it could also be vice versa: first, TAA is applied and then the image is upscaled using SSAA approach. You would then end up with a TAA image processed by SSAA in addition...

@AnkH Well this applies to the vast majority of people I think, and has been the case from the start really (MSFS2020).

It wouldn't hurt for Asobo to reserve half an hour at one of the Twitch sessions to go through the settings that they are providing us with, and fully explain how they work. 

That would be of real value to us.

 

Edited August 13, 2025Aug 13 by bobcat999

Just tried TAA after a few years of DLSS. Flew the COWS DA42 and then the FlightFX C750. Pretty glass heavy. 

TAA looks no better or worse than DLSS here, even after lifting up my varifocals and zooming in as far as possible to compare. Except my machine runs about 5 degrees warmer. 

I see no reason to change back to TAA. 

Your experience may differ. 

3 hours ago, GSalden said:

 

Regarding AutoFPS : I do not think that he used AutoFPS for these tests. 

 

When showing his graphics settings, he says "these settings are a little funky because I am using AutoFPS".

Zooming into the displays is exactly the thing you shouldnt do, because this alleviates the effect. I also use 4K with DLSS DLAA and those numbers are only blurry with the default view point. Especially visible in the Fenix, but also in the ini A350 and PMDG 777, to a lesser extent in GA planes because the G1000 is comparably huge inside the cockpit.

5 minutes ago, AnkH said:

Especially visible in the Fenix, but also in the ini A350 and PMDG 777,

Ah well, there you go, I don't fly any of those so that might explain it. Only GA with and without glass and a few old school light-heavies such as the RJ and the F28. 

9 minutes ago, St Mawgan said:

Ah well, there you go, I don't fly any of those so that might explain it. Only GA with and without glass and a few old school light-heavies such as the RJ and the F28. 

Yeah, those are really great and easy to read.

Do you own the CRJ by any chance. This is what I used to test. No glass, but the type is small and always hard to read. Especially the Altitude. 

6 minutes ago, Ron Lefebvre said:

Yeah, those are really great and easy to read.

Do you own the CRJ by any chance. This is what I used to test. No glass, but the type is small and always hard to read. Especially the Altitude. 

I don't own the CRJ.