Hello ED Team,
I think it would be verry good if you can repair this and I think it's important for this product, and I will explain why I think so: This is a good point where also people who can't fly your great but complex aircrafts can make the experience in VR to "fly with them", and maybe later they decide to buy the modules ;-)
I want to share my experiences to help you to hopefully to enclose the issue (All points are just guesses and the result of long year observations ;-) ):
I found that there are still four noticed reasons for brocken track files:
1. By taking of from a carrier in a F/A-18, the "brocken track issue" happents mostly to me. When I only take off and try to land on the carrier again immediately, there is an "offset" of the delta between the carrier and the aircraft. This could have different reasons:
A) The course of the carrier is different in the mission and in the replay (Maybe the orientation could be the same but the "direction of moving" not?)
B) There is something happening with the aircrafts orientation/position/velocity/acceleration during takeoff (leaving the boat) (E.g. affecting wind effects?...)
C) A difference in the positioning / orientation / velocity / acceleration on initialisation of ship and/or aircraft between mission and replay
When I take of from a land based runway and land on the Stennis, this issue is not (Only by taking off from it).
However, I also have a "little" offset if I have a mission when the ship is holding it's possition all the time and without any wind... But it's not so big...
[Edit] I just had to try out another Idea and I hope this can help you defenetly for this issue:
When I tried out to use the carrier "CNV-70 - Carl Vinson" instead of the Stennis, this issue is no issue and the replay works perfect! So it seems that it's depending on the ship!
[Edit 09.10.2019] I just tried it again using the CNV-70 - Carl Vinson and using this ship the track is definetly not brocken again, but using the Stennis it is (with one exception where I landed on the deck I the replay but in another cable than in the mission, so a small error was still there...)
I flew ~5 minutes and landed in the replay on the Carl Vinson exactly as I did in the mission. So the workaround to create a good track with carrier take off and landing could be to use this ship :-)
D) Another idea I just had is that the error also could occurs during the plane is anchored onto the catapult. A little difference here could result in a big error by flying arround a bit...
2. Another issue seems to occur if wepons are used. I noticed that when I fly around even with AI aircrafts, it seems to be ok as long as nobody shoots. I don't know how the DCS engine works but maybe you are creating a new AI aircraft for the missile? You are also deleting aircrafts when they have been destroyed I guess, so do you have respect this on all points of the reading/replay/track file etc.? Just an Idea... So you should make sure that the recorded inputs are always put in into the right vessel. (==> No confusion with the array index of the object / "max index" etc.? If you are working with arrays, you need to "copy the upper indexes down" in the mission and replay and change the new "max" value of the array, if an object has been deleted. However, please make sure that the inputs are always used for the correct object(s) on replay, even if new vessels have been created and if vessels have been destroyed... ;-)) You also need to make sure that the behaviour of AI aircrafts and missiles is the same in mission and replay (No time based random for weapons and AI aircrafts. If AI aircrafts react on weopons and the weapon behaviour is different between mission and replay, this could lead to a different behaviour of the AI aircrafts also).
(Maybe this is what Dougall reported and means with 1vs.1 dogfights)
3. Module based issues (Multicockpit?): I notices that I had a brocken replay issue each time I changed the seat in the C101 (Maybe also in the F-14?). In a C101 replay, even the takeoff in the replay is not working good and can end into a crash if you have played it in the front seat an if you change to the back seat in the replay (C-101). You need to make sure that the inputs can be made and are used for "the right seat" in the replay. My feeling is that the "center of gravity" and or the "aerodynamic focus" is set relativ to the pilots seat and if you change the seat, it could be wrong for the replay, but it's just a guess and maybe I'm wrong...
4. Time acceleration: If you have a different time acceleration in the mission and in the replay, the replay is not correct. So you should deny to change the time acceleration in the replay as long as this cannot be fixed.
I also noticed that if you
- Create a mission with rain and much wind
- Start the mission and wait some seconds
- Note that the raindrops falls with an angle
- Exit the mission to the mission editor
- Change the wind to '0' in the mission editor
- Start the mission
Note that the raindrops falls with an angle first. Only after a while, they are falling the corect 90° towards earth (because there is no wind anymore).
That means for me that it will take a while until all wind/weather effects are "reseted" to the values which are set in the mission editor (Maybe the weather conditions are not "destroyed" on mission ending?).
That brought me to the idea that If the wind in the replay is different to the played scenario, this could also break the track file... So you should make sure that the weather conditions will be set to the ME values immediately and each time on mission start and replay start (Not only for wind, same for air pressure and maybe more...). The wind must also fit immediately on replay start for the deck of the carrier (The carrier is moving so the wind for the deck is different to a land based runway).
Also please make sure that if in a mission "wake turbulences" are used (or not), they also must be used (or not) in the replay, no matter what the setting in the options is to avoid a different behaviour!
I don't want to tell you how you can realise it, but did you ever thought about to save/load all positions, orientations, current accelerations and velocities in a "timestamp" from time to time (e.g. each second) and set this values during replay instead of just recording all the inputs?
Same for events like "missile hit" or damage things and that... Because sometimes I have destroyed an aircraft in the mission but the missile pass it in the replay, so the aircraft shoot me down in the replay! Thats not funny! ;-)
I guess that not all guesses are right ;-), but I guess they are worth to be checked.
I Hope this will help a bit...