Unsolved 4.37.2: Cursor speed ramp-up erratic when using keyboard/button-button controls

[Huesudo]

When controling the SOI cursor with the keyboard or joystick buttons (not an analog control), there’s suposed to be an “acceleration” or speed ramp-up.
However, this “acceleration” is missing when making a cursor movement in one axis (up/down), and then attempting to move it in another axis (left/right). The acceleration is present when making movements on the same axis (moving the cursor up and down or left and right repeatedly).

[airtex2019]

@Huesudo to clarify … are you invoking ( SimCursorUp + SimCursorRight ) at same time … or are you invoking the newer SimCursorUpRight callback?

does one method work better than the other?

[Huesudo]

@airtex2019 I am using SimCursorUp/Down, releasing it, then using SimCursorLeft/Right. Never using the diagonal directions.
Edit: To clarify even further, the problem also appears when using SimCursorUp/Down after having used SimCursorLeft/Right.

[airtex2019]

@Huesudo oh, that’s weirder than I thought. interesting

[RingoSurf]

I also noticed the exact same problem

[RingoSurf]

Please Fix, it’s very difficult to use TGP

[airtex2019]

@Huesudo @RingoSurf I spent a few min trying to repro, and not sure that I could.

I was using FCR (AA mode) and HSD. The “acceleration” there is very slight… so maybe I’m missing something. But it always seemed ok.

Can you guys clarify more? Is the problem the cursor movement starts slow and remains slow? Or does it start fast and remain fast?

Is this a TGP-only problem, or general SOI cursor slew inputs?

If TGP… is it when the TGP is ground-stabilized or when panning? What zoom mode? etc

I think you either need to post detailed repro steps, or post a short 30 second video.

[RingoSurf]

The problem only using the TGP, the cursor darts away fast in the X axis. I use keyboard keys but it’s the same thing using the joypad. The cursor movement starts fast and remain fast, it happens either when it is stabilized on the ground or when panning in all zoom modes.

[Huesudo]

@airtex2019 I don’t think the problem can be well represented in a video clip.
Also, the AA FCR feels like the least affected cursor.

To replicate, bring up the TGP. Using SimCursorLeft and Right, move the view left and right. You should notice the view accelerates from zero to full speed.
Then, press SimCursorUp or Down. You should feel like the view starts moving almost or at full speed right from the beginning. But if you continue moving the view alternating between SimCursorUp and Down, you should feel the same acceleration as when moving it left and right

[Huesudo]

@RingoSurf I’m pretty sure all the cursors are affected, but the TGP is where it’s most noticeable.

[jayb]

There are some options g_expcursoraxis or thereabout. They slow down initial movements with actual axes, perhaps also the keys

[RingoSurf]

So is this problem reproducible? Has anyone else noticed? solutions ? as far as I’m concerned it occurs only in the TGP (X axis), while in the FCR everything seems ok to me.

[Huesudo]

@jayb Those options are for controlling the behavior of an analog joystick axis used to control the cursor, for cases of physical ministicks on joysticks not working quite right.

Unfortunately for our case, it doesn’t affect “digital” cursor controls.

[airtex2019]

@Huesudo @RingoSurf I mostly fly with analog thumbstick for cursor, so hadn’t noticed – but I just switched back to try the new 8-way hat bindings.

I do confirm the repro case for this. It does make TGP very hard to use, for any sort of CAS or scanning around a target area.

It seems like several small problems combine to make a pretty wild, inconsistent experience.

1- the sudden unexpected acceleration when panning side-to-side after having panned up/down (or vice versa)

2- sometimes the hat input seems to “stick” when moving around… maybe when diagonals are involved? hard to repro 100%, but this greatly exacerbates problem 1

3- (this also seems to affect analog axis input) … I notice the TGP sometimes seems to get “hung up” on ground terrain… even when it’s not ground-stabilized yet. I think I’ll start a separate thread to track this, since it’s obviously not related to input modality – but the effect can further add to the jerky experience.

[Tumbler31]

Keyboard inputs for axes (slew, throttle, pitch, roll) has always been jerky as far as I can remember. This is still somewhere on the todo list.