Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[slibenli]

I haven't tried to run Reshade myself.

Unfortunately I can't really tell from the log what's going wrong. It seems that some GL error is generated by Reshade, which is ignored by Reshade but detected by IL2GE.

[Gattobuono]

Ok, thanks for trying, I really appreciate it.
I'm still enjoying GE and looking forward to future developments!

[slibenli]

That's how the new mesh renderer looks like in-game:

[Gattobuono]

That's how the new mesh renderer looks like in-game...

Could you explain in simple terms what this actually means please?

[Stainless]

Your sampler states are wrong 

That shouldn't take you more than ten minutes to fix

[Hylo]

Some pretty awesome progress though


Sent from my iPhone using Tapatalk

[slibenli]

Your sampler states are wrong 

That shouldn't take you more than ten minutes to fix

The bigger problem ATM is that the meshes wobble slightly back and forth - there must be some floating point precision issue.

[Stainless]

When ?

Have you looked at them when they are stable, and compared to when in flight?

I am just thinking if it is to do with physics or a rendering issue, I doubt it is a floating point accuracy issue.

I have seen cases when bad collision meshes cause wobble as the physics system is getting constant collision events between meshes within the game object

[slibenli]

In flight. Collision detection isn't yet implemented - so if anything it would be the lack of collisions which creates problems.
In the standalone mesh viewer there is no wobbling issue.

Maybe I should just ignore the issue for now and see if it fixes itself when all the missing stuff is implemented.

Update: is also happens when the game is paused - so it seems that physics can be ruled out.
Maybe the camera wobbles around.
Or it's due to z-near/far values.

Update 2: It looks to me like there is an issue with the world->view->projection matrix. It may also be present in the standalone viewer, but less apparent, because the models are located more closely to the zero coordinates.
Update 3: Yea - that's it. It also happens in the standalone viewer when I move everything 100km away from the coordinate system center.

Update 4: I suppose there is some issue with how the view (https://gitlab.com/vrresto/render_util/-/blob/81a05ad9eaaa82df82c0b2f7c41b9c397fbe1272/src/camera.cpp#L216)
and projection (https://gitlab.com/vrresto/render_util/-/blob/81a05ad9eaaa82df82c0b2f7c41b9c397fbe1272/src/camera.cpp#L80) matrices are calculated.

[Stainless]

Is that your camera code or a port of the old IL2 code?

Because I wouldn't ever do a camera like that, it's very old school.

Yaw, pitch , and roll used like this are full of problems and yes you will end up with accuracy issues the farther away from the origin you get.

In fact when I was doing some tests I found flight impossible once you get over 100,000 metres from the origin.

I would look into a more modern camera class, maybe a quaternion based system

[slibenli]

It's my code, but since IL2 passes yaw/pitch/roll to the native code, that's what I have to work with

[Stainless]

Well you can work with it, you just need to change the code.

Basically your projection code is fine, but I would change the rest to a look at matix.

I think you have GLM, not sure, but look at this anyway

https://learnopengl.com/Getting-started/Camera