FlyPT Mover

After one single ride in an early version I found the platform had moved quite a bit, 20-30cm from its original place. It had been quite a ride , even for me on the edge of being comfortable. Anyway, I started with smoothing out most harsh movements. In the last third of the 'Crystal Beach' track there are a few intentional bumps which I used for reference. For many other tracks you can probably 'pump up' a little bit...

Have yet to test the profile. I always test anything new with 0.1 gain and work up. I had to bolt my rig down to the wooden floor to keep the virtual tracker in position.

Despite that, I have discovered that the VT wanders over some hours, perhaps due to the nature of the expansion and contraction of wood, of which the gaming cabin is built. Those base station are attached to the rafters.

I experience the same as @deadlymanager. When DCS is loaded I can connect (with yellow borders), but as soon as I hit the 'FLY' button, source disconnects. Funny thing: the balance between Air and Ground changes by 20% every time I hit 'Connect'. Happens with youngest lua and older lua's.

Can someone advise me on how to apply a filter to the actuators in order that when reaching the maximum stroke the stop would be softer?
do you believe this would be a fine filter? EMABS(VALUE;1.5;100)
Thank you

Logistic

'Mix sway in roll' means that you ADD the result of the sway calculation to roll ─ this may end in exceeding the limits and harsh results. Just my 5ç...

I just uploaded a new version (v1.1) of my IL-2 profile. Apart from splitting up some Pose modules and giving them more commentarial names, I added:

• more 'float'. You'll notice e.g. on landing approaches. Besides, you can now easily tell calm air from 2m/s 'turbulence'.

Experimental effects for

• flaps (all aircraft)
• gear (all aircraft)
• gear jolt (only down position, only selected aircraft)

I'm a great and longtime fan of Simshaker, Andre is a very nice and helpful person, but that jolt is definitely better than his Jetseat. Not necessarily more 'real life', but better!
Anyway, these effects are experimental and you're welcome to work on all of them, including the motor idle, and make them more natural. I saw that Pedro included a bunch of motor types in all variations, but I don't know how to integrate them properly. Maybe anyone with better knowledge makes pose templates for at least a radial and an inline aircraft engine type?

In order to properly address the fact that flaps may be single or multi step type, I had to 'abuse' ROLLOVER. I haven't tested on many aircraft, but on the P-51 you can stop at any of the five steps and rumble starts and stops synchronously. For the same reason I had to integrate a ROLLOVER in gear effect filtering. For some reason, I assume it's rather a bug than a feature, on some aircraft the gear output value (usually from 0 (up) to 1 (down) and reverse walks only 2/3 of the way and stops at either .333 or .667.

WRT the jolt of the gear arriving in down position (or starting from there), it looks I need once more advice of @pmvcda . I'd like to have some distinct jolt at the up position, too. How can I trigger a heave strike right before the 1-to-0-transition ends? I won't go into details what filters I tried (including the "Add(crop(value, min1, max1), crop(value, min2, max2))" you once mentioned), this time with a SUB , but to no avail.

I also tried the REMAP filter with the intention of 'eliminating' a certain range by remapping it to ZERO, but I didn't succeed either. Is there a bug in REMAP or do I not understand it correctly?

Lastly, I noticed new ELLIPTIC, BESSEL and AWAY filters, for which I didn't find (yet) an explanation. Especially AWAY sounds fancy, but I couldn't make out its effect.

@J-1775 Unfortunately as the gear % is affected by speed in drag (as it's intention was primarily to allow drag effect) it's hard to get a completely reliable result as full extension is a moving target and not always 1.0

If a non speed-modified version of gear extension was exposed it would be more reliably triggered.

That said, the methodology I think shows promise is below - this was based on Fly PT's advice elsewhere, I've yet to experiment much though.

EMAHP(CROP(CUBIC(ADD(VALUE;-0.500);0;100.000000;0.000000;-20.000000);0;5);100)

Hopefully screenshots attached.

Basically if you consider the gear a value between 0 and 1 you are making only 0-0.X and 0.Y-1.0 result in changing value (in this case a quick jump between 5 at either extreme and 0 in the middle), and shunting the graph over to the middle by "ADD" -0.5.... Whether its best with HP or LP filter I've not yet tried.

Thanks Zeeflyboy, will give that a shot!
EDIT: I had my two MoverRecs from Spitfire and P-51 run quickly thru your filtering and saw quite the results I wanted. Actually I have two spikes succeding quickly each other, the gaps being differently wide on each of the four events (each AC, start and returning events). I'm sure I can tune that!
And btw, the Spitfire had asynchronous gear ─ so maybe I should rather widen the gap instead of eliminating it!

Great. I’d have a play with a LP at the top to alter the thunkyness too, and you can try altering the “b” of the cubic function to change how steep the walls are.

Let me know if you end up finding improvements during your tweaks.

Hopefully @pmvcda can expose a pure gear position in the source for MSFS/IL2/DCS to make it a little easier to trigger correctly.

EDIT: I guess I found a workaround! It's only the input fields that are restricted to integers. If I type in the values manually floats are accepted!

@pmvcda: I have a problem with LSPLINE filter. When I 'draw' a simple spline (like the one attached) and test it with a triangle loop the graph comes up with the expected results ('barcode lines'). But only when the amplitude is set to 100. And because LSPLINE input values must be integers I can't scale down the spline to be within the range of my actual source (which is within 0-to-1). My natural instinct was to compensate with Gain, but that wouldn't work.
I think LSPLINE and CSPLINE are interesting and useful filters that are not only very intuitive , but in many cases may replace a whole chain of other filters. I appreciate if you'd look into this, thanks!

I can add decimal places easily for the next build.
Those filters are not what I wanted to.
The idea was to have them dynamic, I mean the values could be filters, and the result could be filtered also.
But this type of interface is problematic, and we have access only to the value we are filtering.
In Mover 4.0 this is solved already, and can be really powerful.

Hail to Mover 4.0!

Don't know if this is the right place for it, it's an observation I made wrt transition from ground to air in the new IL-2 Source module.
Short story: there's a number of aircraft that don't transit to air: P-47D-22/28, Bf-109 E/F2, G2, G4, G6, G14 (list may not be exhaustive).
Long story: While the Bf-109s just never transition, the P-47D actually flash yellow while taking off and transit to AIR, but after a few seconds airborne they return to 'Ground' and stay there.
Can anyone confirm?
P.S. This time it's NOT the REPLAY function!

Thanks, I will look at the code.

I start in the air, and with bf109 and spitfire I have air while in the air. P47 shows ground.

I have the same problem.

Thanks again Zeeflyboy. I was on the way of republishing my IL-2 profile including your suggestion, when I stumbled across the LSPLINE filter. Though its use in v3.x is indeed very limited (allowing no other filter before or after) it seemed like sufficient for my purpose. LSPLINE and CSPLINE are extremly intuitive to handle and to understand. And the drag'n'drop interface eases the creation of a filter even more. This is the 'compromise' jolt version 'for the lazy' I came up with. Works with 2/3 and 3/3 gear range, but gives an additional (unmotivated) jolt with the latter...
LSPLINE(0;0;0.01;3;0.04;3;0.04;0;0.34;0;0.34;3;0.38;3;0.38;0;0.96;0;0.96;3;0.99;3;1;0)

I'm tired of looking for the reason, but DCS still doesn't want to connect. I guess I'm stupid.