I was spending some time thinking about how complex flight sims actually are. Does anyone know if infinite flight uses the real current place or does it use where it thinks you will be using a combination or speed,altitude,direction, and wind. And is there a difference between your aircraft and other users aircraft. Also can new features be added forever with a combination of stronger devices and backend optimization or will there come a point where more can not be added.
3 Likes
Hello there!
Infinite Flight’s “world” is a 15m resolution map of the real Earth. It also utilises real-world winds, etc.
Can you elaborate on what you mean by this?
There is naturally only so much you can do with a mobile simulator with the performance limitations of mobile devices. However, mobile devices are getting very powerful and capable now, so hopefully we’ll see some awesome features soon. It’s been hinted that we will be getting a graphics update soon!
7 Likes
Hi, Benny I was asking about the position of the aircraft and how it is determined. Ie does infinite flight use the current simulated position or does it use a prediction based visual after taking into account the winds, speed, direction, and altitude. And is there a difference between your aircraft and the way it is simmed and other nearby users shown to you.
The heart of my question is how does infinite flight know where your aircraft is and how is that simmed to the users.
1 Like
I guess this is closer. I’m not 100% sure as I’ve never made an advanced simulator myself before so don’t know the inner workings ;)
There are some differences - for example, live cockpits in other aircraft won’t show up on your end (for performance reasons). At the moment, ground services that are attached to other aircraft won’t appear on your end, however, the developers are working on adding this in.
4 Likes
I was interested in the inner working and have kind of gone down a rabbit hole.
The whole positioning of your aircraft is very likely a compile of advanced code which is beyond my pay grade when it comes to that area.
2 Likes
If only they made a course on some of the juicy insides of the sim.
2 Likes
Naturally - I don’t see them doing that…
Thank you.
1 Like
No problem! Any other questions feel free to send me a message :)
Have a great rest of your day!
1 Like
Is there a link that back up your statement about the hinted graphics update? Genuinely curious since that update would be a much needed addition to the sim.
2 Likes
I mean, the way I did it in creating a flight sim (there are a lot of inevitable universals of method here) was calculating all the physics that get represented in a set of state variables.
The state of one’s aircraft and any other aircraft would each have such a set of state variables.
(edit: of course, to display another person’s aircraft, you only need to be passed the position and angular orientation info that resulted from their device’s incremental calculations on their other state variables: you can be blind to those variables characterizing their current physics state.)
These set of variables would include all the attributes that you are simulating from the physical world and of course you need 3 variables for each spatial dimension: position: x,y, and z; velocity x,y, and z; rotation about x,y,z, rotational speed about x,y,z etc.
The physics modelling is characterizing your motion through 3d space by incrementing the above type of unique state variables according to the rules you’ve implemented in your algorithm.
You need to include relationships for acceleration (change in speed) from the mass and force relationships both linearly and rotationally (rotationally, for example: torque and angular inertia—>angular acceleration—>angular speed—>angular displacement).
You’ll also have dampening and oscillation effects you have to take care of, and relationships between those different variables that need to be modelled as necessary.
But the bottom line is the physics algorithms operate on your state variables as described, so you can think of the sensation that you actually view on your screen as a completely separate step - you could do the whole simulation without any visual representation.
But for visual representation the graphics engine receives the state variables and determines the scenery region loading and particular scenery within that region and the position and orientation that the scenery data gets processed with to give you the view that you see.
And the external views of aircraft will also start with what the state variables define for position and angular orientation and will use those along with the aircraft modeling and view orientation within the scenery, to “paint” the scenery orientation and the aircraft on top of that scenery orientation correctly, according to the characterization of your view choice and the implied view angle and displacement offsets.
I’m presuming the state variables from the physics modelling is processed on each user’s device for their own aircraft, and passed on to the server for the server to distribute to each user, for viewing other aircraft in your current space.
Whether this is true or not:
Say, in the sense of predictive positioning, to deal with server communication lag or something, is beyond my ability to speculate with confidence.
I don’t know if that answers your question?
3 Likes
This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.