I was playing around with the flight model while trying to set up my pressure sensing stick, and I realized that the flight model, in particular the roll axis, seems to be incorrect.
The DCS Viper had always seemed to fly a bit odd, not intuitive like a real airplane, but I could never figure out why. Something is off about it. I read numerous comments on the forums, and watched several videos, that have this odd pitch-up moment when rolling out of a turn to straight and level. It's done it to me, and real aircraft (even other DCS aircraft) don't do this. Why the DCS Viper? I've trained my muscle memory to pitch down when rolling out to hold an altitude, but this is not a trait with real airplanes.
I think I've found out why it acts this way. I was doing some high AOA flight to sort out stick forces to hold different steady nose-high attitudes, and discovered that ED has programmed the rolling axis of the Viper to be on the Flight Path Marker, not the gun cross in the HUD. In any airplane, there are three axes of rotation (pitch, roll, yaw) that all intersect at the Center of Gravity. Every airplane, no matter the pitch condition, the ailerons (and stabilators) will only roll the aircraft about this lateral axis, which runs down the center of the fuselage from front to back, parallel to the long axis of the fuselage and does not move or change with AoA. This axis is depicted by the gun cross on the HUD. If I pitch up to 30 degrees AoA and then roll the airplane, the roll should always center on this axis (the gun cross), not the FPM which in this case will be well below. In this case, I would end up in a nose high turn and the nose would not roll through the horizon like it does now.
But the DCS Viper rolls around the FPM with high AoA. You can see this by setting up a high AoA flight and inducing a roll. The gun cross will move laterally around the FPM, when it should roll around the gun cross and the FPM will then adjust to the direction of flight.
There is one other thing that can contribute to this behavior. In the flight control software, rudder is added to the roll input to assist in maintaining coordinated flight. This is designed to counteract the situation of adverse yaw (where the nose points opposite of the direction of turn due to drag increase on the outside wing). Most airplanes compensate for this in various ways, and the Viper is no different. Rudder is added to the roll input by the flight computers. Again, in a real airplane I can manually duplicate this odd behavior by inputting positive rudder with aileron. But this would technically be a poor technique, as adding that much rudder would induce a skid, which is too much rudder input for good, coordinated flight.
I'd like ED to have a look at this and see if I'm correct. It may be simply adjusting (reducing) the rudder input in the FM. The biggest issue I've seen is the weird pitch up when rolling back to wings level and having to add significant down stick to maintain an altitude.
Cobb