Quote from FreeScirocco :

Lots of people with powercards forget to realize that LFS is pretty GPU friendly, in a way that the GPU goes into sleeping mode by clocking itself and its memory down if it detects it has not have to do much. Without monitoring GPU-Z (which has a clock history graph) its impossible to say what is going on. When in fullscreen or windowed/borderless screen and post processing on/off the drivers can decide to be in a different power state and therefore cause the 10x framedrop. Its then your card which is "sleeping" but to be sure about this; monitor GPU-Z.

Yes, I know that. But you do not know my configuration / resolution.
Anyway, Scawen provided me with the necessary input. Everything is fine now.

Quote from Scawen :

The reason for that is, I believe 4xAA is the best result for the least minimal extra GPU usage. That's why mirrors and VR render target use 4xAA and that's why the postprocessing one does as well (as it was a quick add-on using some existing code). My understanding is the render target uses 4x the size of Z buffer when you use 4xAA. Higher AA settings use bigger and bigger Z buffer and it starts to become a memory access and processing issue. And regarding anisotropic filtering, higher AF settings cause the card to produce its own special anisotropic versions of textures in memory and the amount used is unpredictable (at least to me it is).

Thanks for the explanations

This makes the topic a bit more understandable for me.

Quote from Scawen :

I'm not sure if you did a typo there. Did you mean AA in the final sentence above? As in antialiasing, not anisotropic filtering?

The fps is low if I select "Full-scene antialiasing" with 8x. If I switch off "post-processing shader" with this value fps fall down to around 5 fps. If "Full-scene antialiasing" is at 4x the framerate is stabil. It doesn't matter if "post-processing shader" is enabled or not.

Perhaps you can only gray-out instead of hidding "Full-scene antialiasing" if "post-processing shader" is enabled? Or do you plan to change this dependency between "post-processing shader" and "full-scene antialiasing"? After Talking about this issue I realize, that "Full-scene antialiasing" is hidden while "post-processing shader" is active. :-)

I'm sorry for the following off-topic question (you can ignore it): Why "Field of view" is limited to 90° in multi-monitor usage? For single monitor configuration I can already use "unrealistic" values of max. 160°. Will be nice if viewing area for multi-monitor usage with more degrees (max. 110°?) are possible in the future.

BTW: Thanks for your fast support with my "little" (known) issue.

Quote from Scawen :

What is your setting for Full-scene antialiasing? (Options - Graphics)

It was at 16x. The issue exist also at 8x.

Quote from Scawen :

So maybe try turning down AF and limit your AA to 4x.

With 4x AF switching between off and on works without frame rate drop.

@Scawen:
By the way I DISABLED post processing and then I had the bad frame rate.

I can not imagine that it is related to the vram. The graphics card has dedicated 4 GB of RAM. According to LFS, the textures memory usage is around 324 MB. There should be enough space/ram, right? I don't know if that so easy to compare. I don't understand enough about vram and usage.

Sorry - This issue is not test patch related. I can reproduce this issue in the current public version too.

I don't know if this issue is since my graphic configuration has changed from a 2 screen to a 3 screen setup.

If I switch off post processing while I'm on a track the frame rate drops down over 10 times lower value then before. I mean that I switch OFF the post precessing. Before it was enabled. If I switch back ON, everything works fine.

If I'm not on track, I can switch between on and off without this frame rate drop.

BTW: I can reproduce that only in fullscreen! In window mode or in borderless mode I can't reproduce it. Very strange behavior.

Used graphic card: GTX980 / current NVIDIA drivers (397.64).

Quote :

It will not be developed or included in future versions

Too bad! :-(

That was a long time ago. But it's nice to see where LFS stands today... and will stand in the (near) future. I'm looking forward to the next updates.

Your questition is a long time ago... without any answer.

Please check your LFS settings. With a Dual Monitor Setup you can configure which one is the main and which one is on the right or the left side. You can configure that under "View" > "Multiple screen layout". Perhaps that solve your issue?

Quote :

Hope you're hearing this mate @Scawen.

Check out the latest comments of Scawen here:
https://www.lfs.net/forum/post/1933469#post1933469
and
https://www.lfs.net/forum/post/1933477#post1933477

Do you plan any new features for LFS Top?

Perhaps your antivirus blocks something? But in this case - I hope - you should get an appropriate message.

Create a checksum of your setup file. Try this tool https://www.microsoft.com/en-us/download/details.aspx?id=11533 from Microsoft itself.


You should get this checksum: 3d685a49561fc783ed8358df67e6825a

You will find this hash on the download page of LFS, too (https://www.lfs.net/downloads). If you get another hash, the file is incomplete/corrupt.

In the meantime I get the new AntVR II (Cyclop) (as a backer: for free).

Because AntVR make use of SteamVR, the headset works in "VR Headset" mode without problems. But because AntVR solves the position tracking only with the carpet (with the internal cam), headtracking is like DK1 (forward/backward and side moves are NOT tracked). Thats really bat. But normal headtracking (looking around) works really good. The Headset is ok. But currently the latest consumer Oculus Rift is cheaper then AntVR and works better.

Perhaps the license upgrades/costs can be better described in the current table: https://www.lfs.net/shop/licenseselect

It's not clear enough in this table which upgrade costs between s1>s2 or s2>s3. Currently I see only upgrades from demo version to s1, s2 or s3. I think thats the reason for the first question of JJ2478?

Did you buy the seat now or are you still indecisive?

Currently I use the "old" Rseat Evo R3.
http://lfs-svrt.ch/index.php/live-for-speed/seats-und-wheel-stands

@Scawen:
Will the physic update also "solve" the following problems or - in other words - will future car damage have more influence on the driving behavior?
https://www.lfs.net/forum/thread/11421-Car-damage-in-LFS-explained

Scawen: Is it possible that AI driver are less aggressive or reckless in the new version? Last race evening with AI drivers was really relaxed? :-D

I can not find any clues in the change list.

One year later... Any news about progress of HTML5 LFS Remote version? On github still the same "old" version is available.

No problem - OK, that makes sense

Then I will still be a "bit" more patience.

Hello

I have created a simple script that changes random the current car (simulate pressing the tab key) and the view (F1 and v/V) with AutoHotKey (http://ahkscript.org/).

Features

• createte a random "decision" for a car switch
  
• simulate a keypressure of key "Tab" (random 1 to 3 times)
  
• simulate a keypressure of keys "F1" and v/V (currently 3 views are defined)
  

Views

• Cockpit
  
• Behind the car
  
• TV cam
  

The result of the script you can see here:
https://gaming.youtube.com/watch?v=lNhYdS_8nbs

The position list comes from LFSlazy (https://www.lfs.net/forum/thread/85771-LFSLazy) (TV mode).

Keypressure is only simulated if Live for Speed window has the focus. The script (source or compiled) can be paused and resumed at any time (Icon in the system tray).

This script is USELESS if you drive on the same PC! Use this script only on a dedicated (streaming) PC.

Download (contains compiled binary exe and source script)

• SwitchCarandView-1.0.zip
  

Last edited by Pasci, Sat 21 Jan 2017, 10:26.

Does not work!

...
16.01.2017 19:47:07.851 INFO database saved: /data/lfstop/databases/RO11.dat
120 /opt/rh/devtoolset-2/root/usr/include/c++/4.8.2/ext/new_allocator.h: No such file or directory.

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0xf6affb40 (LWP 17590)]
0x08237626 in __gnu_cxx::new_allocator<unsigned int>::construct<unsigned int, unsigned int const&> (this=0xf6afebe8, __p=0xf6113200)
at /opt/rh/devtoolset-2/root/usr/include/c++/4.8.2/ext/new_allocator.h:120
Missing separate debuginfos, use: zypper install libgcc_s1-32bit-debuginfo-4.8.3+r212056-2.4.4.x86_64 libstdc++6-32bit-debuginfo-4.8.3+r212056-2.4.4.x86_64 libz1-32bit-debuginfo-1.2.8-5.8.1.x86_64
(gdb) generate-core-file
Couldn't get registers: No such process.
(gdb) generate-core-file

Quote :

This is gonna be fixed tommorow hopefully. Big thanks for the trace

Sounds great - thank you for the fast fix

Quote from vitaly_m :

No need for debuginfos really.., but you forgot to type 'bt' command right after the crash, and then there will be full backtrace.

Ow - sorry

Now you will find the backtrace in the log file attached.

Quote :

I am sorry, and thank you for reporting.

Thanks for the hint to gdb

I don't know what I have done while lfstop crashing. But perhaps following infos are useful any way.

...
15.01.2017 15:56:52.829 INFO database saved: /data/lfstop/databases/RO11.dat
120 /opt/rh/devtoolset-2/root/usr/include/c++/4.8.2/ext/new_allocator.h: No such file or directory.

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0xf6affb40 (LWP 29790)]
0x08237626 in __gnu_cxx::new_allocator<unsigned int>::construct<unsigned int, unsigned int const&> (this=0xf6afebe8,
__p=0xf610edf0) at /opt/rh/devtoolset-2/root/usr/include/c++/4.8.2/ext/new_allocator.h:120
Missing separate debuginfos, use: zypper install libgcc_s1-32bit-debuginfo-4.8.3+r212056-2.4.4.x86_64 libstdc++6-32bit-debuginfo-4.8.3+r212056-2.4.4.x86_64 libz1-32bit-debuginfo-1.2.8-5.8.1.x86_64
(gdb)

Do you need more (debug) infos? See the "missing separate debuginfos" hint (no problem to install this libs too)

Thanks for the update

But the programm still crashes:
25949 Segmentation fault

But I don't find any hint in the logs.

OS: Linux