[RESOLVED] Blinking autogen

[mfannee]

Hello,

I'd like to report a bug in v4.4.16.27077.

My specs:
Operating system
Win 10 up-to-date
Add-ons installed
Ultimate Terrain V2, all regions (developer: scenerysolutions.com)
Ground Environment P3D (developer: flight1.com)
Hardware
GPU: MSI GTX 1050 ti OC [4 GB]
CPU: i7-7700 processor (up to 4.20 GHz)
RAM: 16GB DDR4 2400MHz
Storage: SSD 500 GB
No change at all to my system (no overclocking, ...)
No tweaks to Prepar3D's settings.

My own P3D settings are:
Resolution: 1920x1080 (default resolution for my ProLite B2283HS monitor)
FXAA: Off
AA: 4xMSAA
Texture Filtering: Anisotropic 4x
Texture Resolution: Medium - 1024x1014
Target frame rate: 30 (I also set 30 fps (Frame Rate Limiter v2) in NVIDIA inspector, otherwise I get stutters!)
VSync: On
Triple buffering: ON

Level of detail radius: Max
Tessellation Factor: Medium
Mesh resolution: 10m
Texture resolution: 30cm
Use high-resolution terrain textures: NO

Water detail: Low
Reflections: Clouds only

Scenery Complexity: Extremely Dense
Autogen draw distance: Medium
Autogen vegetation density: Extremely Dense
Autogen building density: Extremely Dense
Dynamic 3D Autogen vegetation: NO

Special effects detail: Low
Special effects distance: Low

HDR Lighting: NO
Shadow quality: Low
Shadow: Internal Vehicle (Cast/Receive), External Vehicle (Cast) only

Cloud draw distance: 60 mi

My NVIDIA Profile Inspector settings
2 - Sync and Refresh
Frame Rate Limiter : '30 fps (Frame Rate Limiter v2)'
Maximum pre-rendered frames: Use the 3D application setting
Triple buffering: On
Vertical Sync: Use the 3D application setting
3 - Antialiasing
Antialiasing - Mode: Enhance the application setting
Antialiasing - Transparency Supersampling: 4x Sparse Grid Supersampling
4 - Texture Filtering
Texture Filtering - Negative LOD bias: Allow
Texture Filtering - Quality: High quality
5 - Common
Multi-display/mixed-GPU acceleration: Single display performance mode
Power management mode: Prefer maximum performance
Threaded optimization: Off

Summary:

Before the upgrade to v4.4.16.27077, I was using v4.3.29.25520, which worked absolutely fine on my machine with the settings hereabove (no performance problem at all).

But as soon as I upgraded to v4.4, I noticed a bug I never had before: some of the autogen buildings were blinking. What I mean by that is: most of the autogen buildings are just fine, but SOME random buildings just disappear for a few seconds, ... then come back, and disappear again… very weird...
Maybe interesting to know for the developers: the buildings that disappear are not at the same location (they can be far apart) and they all disappear at the same time, and then all reappear at the same time too, so they behave as a group, somehow.

I first tried to solve the issue by following the steps in 'General Debugging Steps / How to Post Issue Reports'
https://www.prepar3d.com/forum/viewtopi ... 2&t=123246

So I updated my video card driver yesterday. Today, I uninstalled P3D, removed all the P3D related folders (in %appdata% etc). Then I re-installed P3D and tested the application WITHOUT any add-on: the issue was still there.

Do you have any idea how to solve this?

[Martyson]

Do you have a video of the problem?
Does it happen at the same location / time of day?

Maybe other pilots could try to duplicate / verify by going to the same default P3D location and phase of flight (gate, taxi, approach etc.).

[mfannee]

Hello Martyson,

Thank you for answering.

I don't have a video of the issue yet. But I can tell the blinking happened at different locations. I tested at Nice/Cote d'Azur (LFMN) and also at Bilbao (LEBB). I'm quite sure I tested at two different times too: around 2PM in spring and around 9AM in summer. It made no difference.

[mfannee]

Just a thought:

I read in the release notes of v4.4, section Fixes and Improvements:
"Fixed issue that could sometimes cause missing autogen at scenario start."

I suppose this new fix might be the cause of the new bug, because v4.3 doesn't have this fix nor the blinking autogen bug I described above.

[ThomasEDDL]

Hi,
same problem here. ORBX base, lc, vector, 1070 Ti, no hdr, max LOD and autogen draw distance.
Greetings Thomas

[keenrw]

Im seeing the Autogen blinking too , dont recall it being like this in 4.3

[Saldo]

Seeing it here to.

[dabba]

I also echo the original posters bug report
Prepared 4.4
windows 7 x64
gtx 1060
followed debugging
https://www.prepar3d.com/forum/viewtopi ... 2&t=123246
It appears only to be buildings, not trees
It appears all areas
Default,
Orbx,
amongst other Autogen sceneries

I don't use Nvidea Inspector
no alterations to Cfg files

deleted SceneryIndexes_x64 folder, The Shaders folder, The Prepar3D.cfg file

Thanks

[Martyson]

@ mfannee,

"Before the upgrade to v4.4.16.27077, I was using v4.3.29.25520, which worked absolutely fine on my machine with the settings hereabove (no performance problem at all). "

"So I updated my video card driver yesterday. Today, I uninstalled P3D, removed all the P3D related folders (in %appdata% etc). Then I re-installed P3D and tested the application WITHOUT any add-on: the issue was still there. "

Ok, default P3D v4.4 setup with no sliders etc changed, no add-ons / scenery/aircraft and you seen the problem right away.
Correct?

Where is the best place to see this problem ... airport, gate, runway etc?

Can it be see while inside the aircraft .. like during pre-flight , taxi etc, take off runway etc.?

[mfannee]

Hi Martyson,

The best place to see the problem is from the air. I saw it while inside the aircraft, just after take-off and during flight. From the air, I could clearly see several buildings blinking "together", exactly in sync (t.i. disappearing and reappearing at the same time, even when those blinking buildings were located far apart from each other).

As stated before, I tested at different locations (Southern France and Northern Spain), with the default v4.4 installation without add-ons.
Since I wasn't able to solve the problem, I went back to v4.3.

[dabba]

Hi

it also occurs when autogen at dense, with a vanilla ( no addons) flight.
it appears to be similar to this issue only with autogen buildings and not trees

https://www.prepar3d.com/forum/viewtopi ... f&start=15

on a vanilla flight it is not as prominent as it is with addons ( orbx autogen) ect but still there.

Thanks

[Martyson]

Hi,

I just flew into LEAL P3Dv4.4
All scenery / autogen all ok.
I will keep an eye out on my next flights.

Showed my screenshots, information and comments in this posted topic:
https://www.prepar3d.com/forum/viewtopi ... 41#p192641

[mfannee]

Hello again,

Today I decided to give it a try again:

- I updated my video card driver (see the very first post for my system specs!)
- I uninstalled p3d v4.3 (which runs very well, but I want to upgrade to 4.4)
- I removed all Lockheed Martin folders in %appdata% and %programdata%
- I cleaned the Windows registry (with jv16 Powertools)

So I had a perfectly clean PC. Then I installed p3d v4.4 and tested the result without any add-on installed: the "v4.4 blinking autogen bug" I mentionned in my first post is still there. Nothing I can do about it.

LM doesn't reply in any way: are they investigating? do they care? No way to know if they keep silent.

So, switching back again to v4.3... Fortunately, 4.3 is doing the job for now.

[Martyson]

Hello again,

Today I decided to give it a try again:

- I updated my video card driver (see the very first post for my system specs!)
- I uninstalled p3d v4.3 (which runs very well, but I want to upgrade to 4.4)
- I removed all Lockheed Martin folders in %appdata% and %programdata%
- I cleaned the Windows registry (with jv16 Powertools)

So I had a perfectly clean PC. Then I installed p3d v4.4 and tested the result without any add-on installed: the "v4.4 blinking autogen bug" I mentionned in my first post is still there. Nothing I can do about it.

LM doesn't reply in any way: are they investigating? do they care? No way to know if they keep silent.

So, switching back again to v4.3... Fortunately, 4.3 is doing the job for now.

If you had a video using P3Dv4.4 other pilots could go to the same location / altitude / heading and check it out.

[mfannee]

Martyson,

The bug occurs everywhere. I tested before in Southern Europe (see previous posts). Today, I tested in Seattle in the US, and the bug occured there too. Location, heading etc. are not relevant in this case. Of course you need to have some altitude to be able to see it: I saw the bug occuring only after take off.