SimTools Presets ProSimu T1000 3DoF (SCN5) - Assetto Corsa

Experiment #2

_________________________________________

Experiment #1

I've been experimenting with a pretty different sort of axis configuration from what I've seen around. Figured it'd be worth sharing here. Would love feedback on the creative things others have tried.

I should preface this by saying that I probably spend more of my time playing with my settings than I do actually playing the games! (I guess my hobby is more the rig, than the driving :laugh

Right now, I've been feeling that most of the effects from raw telemetry are just too violent for me on the ProSimu. I've decided I would like something more subtle, so that I am not so overwhelmed by the feedback and can actually better understand what's going on.

Here's an example of my Assetto Corsa settings:

Some general things to note:

• Assetto Corsa's telemetry feed is *really* detailed - probably *too* detailed for even what I a high-end simulator can keep up with
• Multiple categories of telemetry will reflect some of the same inputs from the game. For example, rolling over a high kerb will show up in heave, surge, roll, pitch, and sway in some way. This means your rig is competing to reproduce this telemetry across several different axis assignments at once.

So... I've been using filters to help tone down each individual type of telemetry to do two things:

• Smooth out the input to remove micro-stuttering that just makes the rig oscillate constantly
• Refine each telemetry type to show mostly it's own effect - and to take out a lot of the cross-over effects that show up in multiple axes.

The net result is movement that feels much less violent and dramatic, but more detailed and realistic feeling (at least in my opinion).

Here are screenshots of my settings for Assetto Corsa:

Tuning Center Values:


Note that Surge values are asymmetric for snappy gear shifts.

Game Engine Settings:



Notes:

• I've decided not to use the Heave telemetry at all in Assetto corsa. Heave and Surge are both just too noisy, and all of the effects normally given by Heave are captured in the other telemetry types (So far as I can tell)
  
• You'll notice that my % values don't add to 100 per axis. This is because I use the 'boundary' value in filters if I want to limit the overall travel, and instead just use the general % value just to match each telemetry type relative to each other.

Surge Settings

• The smoothing value helps remove the micro-stuttering that's caused by the very spiky surge data.
• The deadzone helps to remove the very smallest motions, which again, are just too small to actually be accurately represented on a sim
• The boundary value prevents the surge motions from using more than 20% of the travel

Traction Loss Settings:

• Smoothing of 5% removes just the tiniest bit of notchiness in the traction lost axis
• DeadZone of 3 keeps the axis from stepping out until closer to the threshold of the slide, making the start of the slide feel snappier, and thus easier to feel as it begins

Sway Settings

• The 10% smoothing helps mute the notchiness of Sway on kerbing, leaving you just with Sway from turning g-forces
• The deadzone of 4% keeps the rig from constantly making micro-adjustments for distractingly small amounts of sway.

Pitch Settings

• 10% smoothing and 2% deadzone combined eliminates the micro movement for small pitch changes, focusing the axis on the larger, smoother changes in pitch

Roll Settings

• Very similar to the Pitch settings - focusing on smoothing out the motions to eliminate oscillation from constant micro changes in Roll

Heave Settings (I disable this)

• I don't use heave, but when I experimented with it, using smoothing helped with the stutter, and limiting the boundary to 2% of travel made sure it didn't overwhelm the other inputs too much

I hope this is helpful! These same principles can be applied to other game profiles.

I've definitely tried that both ways, but it doesn't seem like I'm getting the results I expect. I had also read on a few threads here that after the 1.3 update the 'Total to 100' guideline wasn't as relevant to the balance of overall forces. That might be completely wrong though, and I imagine I can trust you to know better than I would!

Right now what frustrates me really is that Deadzone in particular doesn't seem to be working the way I expect it to.

Firstly - I don't entirely understand why it's a 0-10 range, when the rest are 0-100.

Secondly - I really want to filter out the low speed oscillation of the Surge effects, but keep the gear change effects.

(So in other words, zero out small oscillations, but go ahead and move for the large motions over the threshold)

For example:
- Tuning center Surge: 0.5 to -1
- Surge filter: 10 <-- no apparent difference compared to 0?

I think I must be missing something...

Great writeup and will try some of these settings. I have the prosimu t1000 2dof

Great! Please report back what you find with your own experiments.

One thing I will note, I think the 'noisiest' telemetry feeds in AC are: Surge, Sway, and, to a lesser extent, Heave - so for anyone reading this who's like me and prefers smoother motion and less teeth-rattling, those are the things to reduce or even turn off.

Remember - a lot of the feel is duplicated across the other types of telemetry, so you aren't necessarily numbing out your rig entirely if you do that.

That makes sense @noorbeast - thanks for the clarification.

In other words, if there was a steep roll at 100% the servo would max out it's travel, and then if there was a steep pitch change, that additional change would be clipped.

But if the roll was only at 50%, there would still be 50% travel left to reflect the pitch change as well.

Bearing that in mind, I think a certain amount of clipping is probably acceptable for the kind of motion I like (I actually think it's okay for one telemetry to dominate the others under extreme change) - but I'll probably dial it back a bit for sure.

I appreciate your restraint

Yup -you speak the truth. And until a couple weeks ago I've also been using a total allocation of 100%, to be clear.

These settings are definitely experimental.

Strictly speaking, because of the boundary setting on surge, the total allocation is actually more like 210. And yes, that is potentially a pretty extreme loss of fidelity. I am more conservatively adjusting the proportion to a total of more like 160... still probably higher than your taste, but I think it's worth the experiment.

Here's why I think it's at least worth trying. The shape of the incoming telemetry data is very relevant to the experience of motion in the sim. If (for example) the extreme ranges of the inbound telemetry values only ever appears in a crash situation - i.e: the telemetry plot looks very spiky - then preserving those peaks is not very valuable anyway. But you still want a high resolution of movement for the non-peaky parts of the curve. (or at least... I do)

So the trade-off of using an increased axis allocation, but using boundary values to reserve some travel for other telemetry types, is an increased sensitivity to the majority of the movements that don't hit those extreme ranges. This would not be appropriate for every game or every telemetry type, but I think it is appropriate for some (particularly in Assetto Corsa).

... then again, my thinking may be bunk. Going to keep up the trial and error and see where I end up.

Next experiment

• Disabling Heave and Sway entirely. They are simply too noisy, and *most* of the same effects are also conveyed on the other telemetry channels anyway.
  
• Heavy filtering on Surge

This cleans up the low-speed noise and oscillation enough, to make it less overwhelming, which makes me comfortable to use unfiltered Roll, Pitch, and Traction Loss.

• Total Axis 1a allocation: 180* raw, but with Surge limited to Boundary: 20, it's actually closer to 100 in terms of clipping behavior.
  
• Total Axis 2a allocation: 180* raw, but with Surge limited to Boundary: 20, it's actually closer to 100 in terms of clipping behavior.
• Total Axis 3a allocation: 100

• As far as I can tell, this is the best balance between eliminating low speed oscillation and yet still feeling fairly responsive higher value Surge spikes.
  
• Not using Deadzone any more - isolated the engine to just Surge and tested 0 - 0.5 - all the way up to 10... never found it to make a significant impact.

Had a lovely evening driving the LA Canyons in the Porsche 718 RS 60 Spyder with this setup.