[ALREADY REPORTED] DCS World Net code Bug (Aircraft Misplacement)

[PanicMan15]

INTRODUCTION:

 

For several months now DCS net code has been an issue. Whether it be combat or aerobatic flying. There is a trend where the person that you are flying with, tends to not be accurate to what the user on the other end is seeing / doing. For us aerobatic pilots in the community. It never used to be this bad as it's gotten over the past month. To explain what is happening seems to be an issue with the system that is implemented to help "smooth" out the movements of users over the internet. However, that system has since become flawed and exaggerated that dampening. It is now becoming apparent that under G load. To put it into terms so there is an understanding. When I say USER, I am referring to the person that is witnessing the event. Usually the person that is watching their own track back. When I refer to the CLIENT, I am talking about the person that is being witnessed and is having the issues. So to continue, under G load, the USER witnesses the CLIENTS aircraft move vertically down and forward. This becomes very apparent under loops. We are not to sure if airspeed has a factor but we have tried this with the F/A-18C, L-39C and YAK-52. All at airspeeds ranging from 100 knots to 350 knots.

 

Progressing as to what we know. We know that this started in 2.5.5 and has been exacerbated in 2.5.6. We also know this is an effect of net code. We have thought it could be internet but after multiple test and multiple flights with perfect internet from users and poor internet from users. It seems to happen every time.

 

EVIDENCE:

These two tracks are from the date 8/2/2020. It is 2 L-39C flying formation. The reference for formation that is being used is the pitot tube on the wing. Stabbing the pilot in the back of his head. You should see in the USER track, that the CLIENT is flying forward and down.

 

USER TRACK:

https://drive.google.com/file/d/1T3Z0VoX3tRQInYP50iDFU-KNABuUs00N/view?usp=sharing

CLIENT TRACK:

https://drive.google.com/file/d/1EjlXU8MfRBi0GPS8EivxULe9ZfXTPyHz/view?usp=sharing

 

----------------------------------------------------------------------------------------------------------

 

- Best Regards, Panic

Edited August 6, 2020 by PanicMan15

[VanekK30]

They were informed about this bug about 2 months ago. (It started in the middle of May (about when SuperCarrier came out). But for some reason they still haven't been repaired. And soon the show ...:(:helpsmilie:

[PanicMan15]

They were informed about this bug about 2 months ago. (It started in the middle of May (about when SuperCarrier came out). But for some reason they still haven't been repaired. And soon the show ...:(:helpsmilie:

 

Yes sir they have. However, this issue just developed recently and became noticeable within the month so it's gotten worse unfortunately. :(

[CamelVA]

The guys over at Virtual Aeroshell have also experienced this issue and we are forced to fly on 2.5.5.41371 because of it.

 

When G is added the wingmen shift lower than they actually are. When speeds get slow at the top of loops, often the aircraft move forward as well.

 

I've created a comparison video between the two versions of DCS. This is not just one flight but all flights that take place in OpenBeta.

 

 

Hopefully the Netcode will return back to normal soon.

 

It's important to keep in mind that our wingmen are supposed to stay at Wing Level or above at all times. Sometimes they dip below the wing but nothing like this.

Edited August 3, 2020 by CamelVA

[VH-Rock]

I can confirm that the Virtual Horsemen are also experiencing some sever netcode issues since the release of the Super Carrier module... This has unfortunately resulted in us having to use an older version of DCS (like Camel) to rectify the issue.

 

Initially we were seeing the aircraft sit 2-3ft low in a loop from the external view when the pilots were actually maintaining the correct position. Obviously very frustrating when precision formation flying is the reason you fly DCS...

 

In recent weeks, the issue has become more pronounced and we have been experiencing several phantom crashes or contacts per flight as the sim thinks we have crashed...

 

This is the position the pilot sees and maintains:

 

 

This is what the lead aircraft sees at all times. (Noticeably further forward than the wing pilots view):

 

 

This in itself is bad enough as it ruins the look of the formation. However, the biggest issue is that as G is applied, the position of the aircraft shifts in sim (Even though the wing pilot is maintaining the position showed in pic 1):

 

 

 

As you can see, the sim thinks we have made contact, but the wingman is still holding his correct position on the wing.

 

Please, please look into this and get it fixed as soon as possible. This is causing some major disruption to our hobby...

[Coop]

I'll add to this issue. We noticed this "position lag" right after the super carrier release. Before then, client aircraft matched location in reference to other aircraft in a formation. I've provided some screenshots below that I will elaborate on to show the issue at hand. I listed the time of the track and positioned myself in IDENTICAL spots on the ground to take these screenshots, to demonstrate the effect this lag has. I use the head of each pilot as a reference to indicate jet position throughout this post.

(Screenshot above ^) This is a screenshot from my track, me flying the left wing #3 jet. Based off the way this "position lag" acts, I will be in the proper location I saw from my aircraft in relation to the #1 lead aircraft. You can see that my head is located a few feet forward of the aft part of the left missile rail. #2 and #3 wingtips are "in line" meaning we have symmetrical distance forward/aft the lead aircraft.

(Screenshot above ^) This is a screenshot from the #2 pilots track. You can see his Actual position is much further back than what my track indicates. My head (#3) is also further up the missile rail, nearly touching the front of the leading edge and the missile rail joining together.

My assessment of the issue is non client aircraft will fly roughly 5-10 feet forward of their respective client positioning.

To re-enforce, this is the SAME flight, same time, same ground camera position, where the only variable changing these screenshots is net code/lag issues.

(Screenshot above ^) This final screenshot is from the lead aircraft #1 Flex track. Both wingmen #2/#3 are off from the client track. Between this screenshot and my client screenshot entitled "#3 Coop Client Track", there is nearly 10 feet difference forward and aft, based on my head location.

 

-------

 

This screenshot shows my client viewpoint (#3 left wing) to prove my location on matches with only my track screenshot.

I am flying WELL behind the leading edge and missile rail connection, which aligns with my first screenshot entitled "#3 Coop Client Track".

But based off the screenshot entitled "#2 Razor Client Track", it has me flying about roughly 5 feet forward of that position.

This screenshot (^) is taken from the Client track of the #2 Jet. He is flying well aft the center of the missile rail, as indicated by his client screenshot entitled "#2 Razor Client Track".

But reference the screenshot entitled "#3 Coop Client Track" and you can see he is roughly a 1/3 of the way up the missile rail.

G loading in this maneuver is consistently between 1.8 - 2.0 Gs, which is a relatively small amount in terms of the entire show. Adding G appears to cause a "drop" like Camel's post suggest. 5-10 feet in precise formation flying, particularly when wingtips and canopies can be within that 5-10 feet from other aircraft, this is a huge issue. We measure maneuvers based off inches, and moving 10 feet can be catastrophic to the formation. Based off the issues noted here, we ran into other issues that I presume to be associated with this "position lag".

${1}

This video showcases an issue I believe is related to the "position lag" and shows the results of mismatched location.

This is a video recorded from #2 Razors track, and showcases an abrupt, uncommanded nose down pitch attitude. We have had this issue occur several times, on this maneuver.

We believe it is caused by the "position lag" inaccurately positioning us, in turn, causing an invisible or non existent collision, so to us, it appears there is a random abrupt nose dive.

 

I'd like to re-emphasize all these screenshots are taken from the exact same time, same location, same angle.

I hope these screenshots and description can help allow a solution to be made to fix the issue, because it is a significant hindrance to formation flying in DCS World. I'd be more than happy to answer more questions about our experience with this issue to help find a solution.

Thanks - Coop

Edited August 4, 2020 by Coop

[VH-Rock]

Great post, Coop. I'm hoping we'll get a few more coming through posts like this so people can see how wide spread the issue is. It's really made flying our demo in the latest versions of DCS impossible :(

[Tom Kazansky]

I'm pretty new to this subject but since I plan to use multiplayer in DCS much more from now on I'm very interested in this issue.

 

 

Is this "just" a matter of inaccurate track replays or does a live observer see the inconsistencies in positioning too?

[VH-Rock]

I'm pretty new to this subject but since I plan to use multiplayer in DCS much more from now on I'm very interested in this issue.

 

 

Is this "just" a matter of inaccurate track replays or does a live observer see the inconsistencies in positioning too?

 

Live inconsistencies. The screenshots I posted were taken whilst flying live. The wingman sees himself in the correct position. The leader sees the 2 wingmen sitting much further forward than they actually are. As we pull through a loop, it gets worse and worse to the point that DCS thinks we have collided... From the wingman's perspective, he has maintained the same position all the way through the loop.

[Tom Kazansky]

Live inconsistencies. The screenshots I posted were taken whilst flying live. The wingman sees himself in the correct position. The leader sees the 2 wingmen sitting much further forward than they actually are. As we pull through a loop, it gets worse and worse to the point that DCS thinks we have collided... From the wingman's perspective, he has maintained the same position all the way through the loop.

 

 

Thank you. That's a really big issue for the kind of multiplayer gaming I'm about to start. Hope for a very soon fix.

[PanicMan15]

Can a ED staff member reply with a link to where this was reported? We want to make sure this is the same issue and that we have a clear understanding of what is happening. We want to be able to provide the information but there was no followup from a team member. We just want to be sure this has been reported properly before. Thanks.