New beta update 1.37.8.0 out

[Alvega 2,124]

11 minutes ago, Dillon said:

Did the northern Caribbean finally get addressed?

No, it didn't. 🙄

[bobcat999 8,484]

25 minutes ago, Dillon said:

Did the northern Caribbean finally get addressed?

No.  It was initially stated it will be fixed in a few days.  Maybe they meant a few months.    

[MarcG 2,997]

Off work with the lurgy so thought I'd run a SU14 > SU15 beta benchmark, download started on fire and thought "ooohhh this is nice" but then SPLAT, down to the standard 2mbps 🙄 

#795 of the things they need to get sorted for MSFS2024

[mmcmah 408]

4 hours ago, virtuali said:

Alright, I never got that message, so it recognizes the new MSFS beta correctly.

 

I think I understood what's happening, and I suspect there might be a problem with the "lock at 50% of monitor rate" option in MSFS when used together with the "TLOD Min on Ground/Landing" in AutoFPS.

I'm not using Frame generation, and my monitor is 60hz, so the fps is locked and steady at 30 fps and I set 30 fps as target in AutoFPS. The MSI Afterburner overlay graph confirms a 30 fps with a flat line, and I could keep 30 fps even if I closed AutoFPS and set TLOD manually at 200.

But since I have the LOD min at 25 so, and I'm on ground, TLOD is 25, it's working as expected.

However, after takeoff, if the fps in the sim is locked at 50% of the monitor rate, AutoFPS doesn't even try to increase it: it just shows TLOD 25 in red and it stays that way. Cloud quality was set to Ultra and never reduced (Decrease Cloud quality is enabled), and I was flying in clear weather anyway.

As soon as I changed the fps lock option to 100% of monitor rate, fps went up, and AutoFPS quickly raised TLOD up to 200 and kept it that way.

If I set the frame rate lock to 50% monitor again when in flight, TLOD stays at 200 as it should.

It seems that, when the fps is already locked at 30 fps in the sim, so you are already on target, if the TLOD Min on Ground/Landing option is used, it never tries to raise the TLOD after transitioning from ground to flight. Here's a video I made:

 

 

That makes sense. The app doesn't see any headroom between where you are with your TLOD min on the ground and where you are higher up, since you have the same FPS vs your target.

How does it work if you uncjeck the TLOD min on ground/landing option?

[virtuali 4,421]

1 minute ago, mmcmah said:

The app doesn't see any headroom between where you are with your TLOD min on the ground and where you are higher up, since you have the same FPS vs your target.

I understand that. It seems that, while there's clearly a logic that decreases TLOD if you are below the target frame rate or increase it if you are above, there's nothing that *tries* to increase it if you are locked exactly on the target, because it thinks you have no headroom for doing that, when in fact you just locked the fps.

[mmcmah 408]

8 minutes ago, virtuali said:

I understand that. It seems that, while there's clearly a logic that decreases TLOD if you are below the target frame rate or increase it if you are above, there's nothing that *tries* to increase it if you are locked exactly on the target, because it thinks you have no headroom for doing that, when in fact you just locked the fps.

I think it might if you uncheck the TLOD min on ground option.

[DuckZilla85 17]

I played with the AutoFPS settings and noticed the same behavior as @virtualinoticed. TLOD won't go UP if your FPS are locked-in to 30, 60, 90 etc... The way I found out how to make the TLOD raise with a locked FPS is using RTSS and instead of locking my FPS at 60 by exemple I simply lock it at 63FPS which allows the AutoFPS apps to "see that you have a headroom." 

[MarcG 2,997]

Well this doesn't bode well, as per ever all benchmarks run identically as possible, I thought there was a discrepancy with the first SU15beta benchmark so ran it again....same result 😐

[DuckZilla85 17]

5 minutes ago, MarcG said:

Well this doesn't bode well, as per ever all benchmarks run identically as possible, I thought there was a discrepancy with the first SU15beta benchmark so ran it again....same result 😐

 

Thanks for running this. I'm patiently waiting for the final SU15 before updating and was hoping the performance would be better as stated previously by Asobo themselves. Someone else stated the had better performance with the Beta so I was optimistic but looking at these charts I have questions....

The Frame time graph on SU15 seems much better and less erratic than SU14 doesn't it? However, the Pie charts says otherwise?! (more >12ms variance)?

Frame time/stuttering is my biggest gripe about MSFS. 

Also, what is your in game settings as I feel it would add to the clarity of these benchmark.

 

Thanks

 

 

[MarcG 2,997]

17 minutes ago, DuckZilla85 said:

 

Thanks for running this. I'm patiently waiting for the final SU15 before updating and was hoping the performance would be better as stated previously by Asobo themselves. Someone else stated the had better performance with the Beta so I was optimistic but looking at these charts I have questions....

The Frame time graph on SU15 seems much better and less erratic than SU14 doesn't it? However, the Pie charts says otherwise?! (more >12ms variance)?

Frame time/stuttering is my biggest gripe about MSFS. 

Also, what is your in game settings as I feel it would add to the clarity of these benchmark.

 

Thanks

 

 

forgot to add my benchmarks are in VR (as that is what I always fly in), I am however running a 2D benchmark in a minute to see if there's the same issues. (I'll post settings afterwards)

[mobeans10 197]

58 minutes ago, DuckZilla85 said:

I played with the AutoFPS settings and noticed the same behavior as @virtualinoticed. TLOD won't go UP if your FPS are locked-in to 30, 60, 90 etc... The way I found out how to make the TLOD raise with a locked FPS is using RTSS and instead of locking my FPS at 60 by exemple I simply lock it at 63FPS which allows the AutoFPS apps to "see that you have a headroom." 

A frame time measurement would be a better variable to control LOD if it was observable to the app since it would show headroom available even with capped FPS.  

[MarcG 2,997]

Did a quick fly about in SU15beta (VR) and it's a stuttering mess as above benchmarks would indicate, went back to SU14 and it's back to being OK (still a regression on pre-SU14 performance lest we forget). I never tried the early builds of this beta so maybe I missed the good performance increase, Asobo have royally screwed something up somewhere down the line since as this is currently a massive downgrade - lets hope whatever they did they fix it.

[virtuali 4,421]

2 hours ago, mmcmah said:

I think it might if you uncheck the TLOD min on ground option.

Then I would lose the ability to give priority to the airport and the ground services while on ground, where I might not need an high TLOD (unless the airport is close to Photogrammetry buildings, which really look bad at low TLODs).

I think being able to use a different strategy if the frame rate is locked might be a useful feature.

Edited March 13 by virtuali

[AnkH 2,268]

The question is: do you really still need the framerate lock if you are able to get more than those 30 FPS when using the AutoFPS tool? Back in P3D, I only locked my frame rate because my rig was not capable of achieving much more and it helped to stabilize the FPS. But nowadays, using frame generation and MSFS, I have my FPS unlocked and thanks to AutoFPS, the framerate anyway never dips into a region critical, e.g. for my G-Sync to be disabled (running from 48FPS upwards). 

[mmcmah 408]

5 minutes ago, virtuali said:

Then I would lose the ability to give priority to the airport and the ground services while on ground, where I might not need an high TLOD (unless the airport is close to Photogrammetry buildings, which really look bad at low TLODs).

I think being able to use a different strategy if the frame rate is locked might be a useful feature.

I don't recall if you did already, but can you please post a screenshot of your AutoFPS settings? CC'ing Reset as well  @Reset XPDR