reported TF-51 crashing into sea freezed the game

[dorianR666]

hello

 

i was just flying the TF-51 and crashed into sea after losing ailerons from overspeeding, syria map. as i crashed into sea, i was in f2 camera and this is the last frame i saw.

 

 

the game then froze like this.

 

notice the numbers in the bar - altitude is at least millions of feet and G is at least in thousands.

 

dcs.log

last lines are suspect:

2021-04-12 19:35:29.489 ERROR_ONCE WEATHER: getAtmoWind_w NOT FINITE POSITION
2021-04-12 19:35:29.490 ERROR   SOUND: invalid host_params(CANOPY_SHOST): position velocity
2021-04-12 19:35:29.490 ERROR   SOUND: invalid host_params(LEFT_GUN_SHOST): position velocity
2021-04-12 19:35:29.490 ERROR   SOUND: invalid host_params(RIGHT_GUN_SHOST): position velocity
2021-04-12 19:35:29.490 ERROR   SOUND: invalid host_params(TAIL_GEAR_SHOST): position velocity
2021-04-12 19:35:29.490 ERROR   SOUND: invalid host_params(woLA-16779520): position velocity

 

maybe numerical instability in physics engine creating values of infinity or NaN

[dorianR666]

okay, i stumbled upon something absolutely insane

 

replicating this with the following track

bizarre.trk

i found out that after the TF-51 crash, other planes in the game start deforming themselves and stop producing engine sound

 

 

 

 

 

some of the planes also start flying sideways, start falling into ground and dont collide with ground, just fly below it

Edited April 12, 2021 by dorianR666

[dorianR666]

got it on video

 

 

[Flappie]

Your tiny Bf-109 is so cute!

Joking aside, I cannot reproduce your issue when replaying your track.

(I mean no sound issue nor visual bug)

[dorianR666]

i have turned off the computer, booted up, ran slow repair and replayed the track

 

its still happening

 

it may be HW related but i think its not graphics related (card, drivers, ...), or at least not directly in sense of incompatibility.

this is clearly a problem with physics simulation, you can actually see the remnants of my plane flying up into space in the background at 0:50 in the video. the log is talking about values of infinity. something in the physics engine is dividing by zero or something close to zero.

 

maybe it is sensitive to FPS or perhaps stutter caused by loading sound effects of the impact (disk speed).

 

27 minutes ago, Flappie said:

when replaying your track.

one more thing: have you cycled through the bf-109s with f2 after the crash? its not part of the track i think. one of the bf-109s isnt getting deformed for some reason but its propeller effect and sounds are still broken.

Edited April 12, 2021 by dorianR666

[BIGNEWY]

Hi

 

is the mission a scripted one? 

 

thanks

[dorianR666]

1 minute ago, BIGNEWY said:

Hi

 

is the mission a scripted one? 

 

thanks

no

its as stock as it gets, you only need base game and syria.

[BIGNEWY]

 

I was able to reproduce in open beta, but the good news is in 2.7 it seems to be fine. 

 

I would suggest waiting for 2.7. 

 

I will have the team check it in the morning however 

 

thanks

 

ps cute isnt it, I want one

[dorianR666]

 

i used "take control" on the replayed track so it doesnt end when i ended it originally.

 

it gets even weirder
this thing is bf-109's wingtip, the plane gets many kilometers long... and rectangular. i think its because of lower LOD from distance.

 

 

 

 

2 minutes ago, BIGNEWY said:

I was able to reproduce in open beta, but the good news is in 2.7 it seems to be fine. 

 

I would suggest waiting for 2.7. 

 

I will have the team check it in the morning however 

 

thanks

 

 

alright

 

 

thank you both

Edited April 12, 2021 by dorianR666