Question Seeking some tuning advice

Hello,

I have a Prosimu 4-DOF simulator with 150 mm actuators. When I began using SimTools I tested the actuators at nearly their full length, but I quickly discovered this felt more like a carnival ride than a racing simulator. Accordingly, I knocked actuator limiting down to 50% and the platform feels fantastic now.

However, due to the actuator limiting, one thing I'm missing out a bit on is the feeling of elevation changes on various tracks. Ideally, I'd like to keep surge and sway locked at using only 50% of the actuators, while giving heave, pitch, and roll use of 95% of the actuators to get a better feel for changes in track elevation. I've tried lowering the % values of these forces by 50%, but things still don't feel right. Additionally, I clicked the "fit" buttons on the surge and sway forces and tried setting their "boundary" values at 50%, but as far as I can tell these adjustments have no effect on how the actuators behave.

Any suggestions on how to achieve what I'm shooting for would be hugely appreciated. Thank you.

It's the interplay and cancellation effect of roll/sway and pitch/surge which limits the capabilities of simulators like the Prosimu. You really need to fit Harness Tensioners and a G-Seat.

Funnily enough, I actually own one of your harness tensioning systems. It's fantastic and totally worth the price of admission. If you ever decide to offer a g-seat, it's definitely something I'd be interested in. Anyway, thank you for chiming in. I'm hoping that someone has an idea on how to winnow out the elevation effects, as surely there's a way to do so.

Hello Lasagna!

I own a prosimu 5dof with 150mm prs200 actuators so I assume same as you except for the traction loss. Would love to exchange notes. I have only tuned for iracing and project cars2. Not tried any flightsims yet. I’m at work for the next three days, but I’ll post my current tuning and axis values when I get home. What racing sims are you running?

A little disclaimer: I’m learning simtools and I’m not 100% certain on my own understanding and if I spread disinformation please let me know. Take it for what it is worth.

Come to think of it I believe I can give some settings from memory for pcars2 at least. In axis sway is at 25, surge 20, pitch 30, roll 30 and heave 20. I run axis limiting at 95. As you run at 50 limiting and your axis assignment at prosimu recommendation(?) this means you effectively cut your axis-assignments in half so in effect we are in roughly the same ballpark as far as axis utilisation is concerned. Did I get that right? As prosimu recommends 50 or 60 for pitch for example. Will there be a difference in feel when I set 30 in axis assignment with 95 limiting opposed to 60 with 50 limiting? I would assume it is two different paths to the same results?

If so with axis at 50 limiting you have excluded the actuators potential by 50% and with your current settings what you want to achieve is not possible.

If my assumptions are right you can try the following. Axis limiting at 95%. Then you tune pitch and roll by their own. Start at 60 for each. Recording telemetry in tuning center at the ring gives me a value of about 9 (perhaps a little higher) on both pitch and roll and as far as I can gather you need at least that to get the full telemetry on that track. My understanding is if you set 6 (degrees) on pitch and you are in an incline of 6 degrees, simtools cuts out any output from the game over 6. In other words your sim will not move further than the position it is in when the game telemetry reaches 6 and movement in the range +6 is clipped by simtools. Movement for pitch resumes again once simtools reach a value below 6. Could use some verification on this.

When pitch and roll feels ok, try them together. Still ok, move over to the axis that controls forces in G’s. Heave, surge and sway. I run these 20, 20 and 25 respectively, but with lower tuning center values. heave 1.22, surge max 1,3 and min 1,7 (for aggressive movement when putting the pedal down and more range when braking) and sway I have at 2.22 I believe. This gives short and snappy cues for the g-forces as to not mess up the feel of pitch and roll. If you have too high axis assignment value on surge for example it will feel like you drive uphill while accelerating on flat ground. If you are accelerating while going downhill worst case is these two directions of force cancel each other out and the sim remains in a horizontal position. Surge pulls the front up and the pitch pulls down. Result:flat.

Now for brevity’s sake, don’t bother with heave right now. When you get surge and sway right. Try them together. If you are happy, now try them with pitch and roll. Not happy? If so I believe you have met the physical limitation that SeatTime described and you have to make choice.

That choice is if you want to prioritise longer movements (more range and detail) for g-forces (surge and sway) or DOFs that affect the inclination of the car (pitch and roll). My current priorities is long movements (high tuning center values) for pitch and roll and shorter and faster movements for G’s (low-ish tuning center values). From my own testing this is the best result I have achieved. My conclusion is that sway in particular gives a lot of usefull cues in mid-corner, but pitch and roll is increasing the immersion factor and if the track is relatively flat you can have both to a degree, but you will run into challenges on the ring. As you have a SeatTime tensioner (I do too, but haven’t put it in the mix yet) I would run sway on that for longer g-forces, a little sway on the prosimu to give a cue at the beginning of turn in and at the end of the turn, low tuning center settings for surge for aggressive, but short movement for breaking and accelerating and higher tuning center for pitch and roll to get nice range and linearity on those.

As I’ve said I will post my settings (so you don’t have to decipher this post) when I get back from work and you can try those

I’ve never thought about using axis limiting while tuning. Would you mind posting your settings for me to try? Have you tried changing speed and acceleration in the interface settings or do you run the recommended starting settings given by prosimu 160/160?

I’ve found that pcars2 in general is easier to get defined and crisp movement as opposed to iracing which feels more soft.

That said. Getting good readable settings for pitch is difficult for me. It’s a very gradual movement that is hard for me to detect. When I run my RL car I’ve tried to analyse what makes me notice difference in pitch and I’ve come to realise that for me it’s more about how the car moves around me than the feeling on my body. Imagine that your head is mounted on a camera gyro always trying to stay level. When the car pitches up you tilt your head forward to keep it level. This in turn the dashboard moves up in your vision. It’s hard to describe, but imagine someone lowering or raising a wall you can see the top edge of in front of you.

Most sims have a setting that lets you set how much your viewpoint is locked to the horizon. In project cars 2 this setting I believe is called world movement. At 100 this means your head movement follows the car and as you pitch and roll (I use vr) nothing changes. The car doesn’t move around you. If you lower this value to 0 your viewpoint is locked to the horizon and you will see the car move around you. I believe using this setting you can get what you see more in line with what you feel in the sim and this might increase the feel of pitch on your sim. To much lock to horizon and I get unwell in vr and the movement of the sim also interferes, but I haven’t tried fine tuning this setting. perhaps 50 is a good compromise?

This link explains lock to horizon better than I can.

I am by no means expert in SimTools, and my system is different (ProSimu 3dof with Dyadic actuators), but I would imagine you may get better results by:

* Reverting the limit to minimum recommended by ProSimu (in my case, 85%) IMHO, it is better to limit travel by manipulating Min/Max values, because you get more range headroom.
* Make sure DOFs you want to be more pronounced have enough Axis assigned.
* Manipulate Min/Max values ranges to adjust how much weight each DOF gets. I multiply Min/Max value by some multiplier to reduce/increase travel. In your case, I'd say multiply Surge by higher value than the one you use for Pitch, so that Pitch gets to full travel before Surge.

This might sound confusing, but basically smaller Min/Max range makes actuators reach full travel earlier, whereas larger range will require more force to be reported by the game for actuator to travel. By using different multipliers, you could prioritize forces.

What I did personally, I collected Min/Max for an average car, then I multiplied those values till forces are not too strong at 0s in game manager profile (Really wish SimTools had master multiplier in Game Manager). I settled on multiplier of 2.5 for all DOFs Min/Max values. The great benefit is that when you crash or hit something, it is only then you get a strong blow, because using this approach leaves some headroom for extreme situations. Then, I just use Game manager profile sliders to amplify/reduce individual forces for each vehicle/track.

I'll just ask some more as this has not been clear to me. So the max/min tells at what output from the game where full use of a given axis should be used. Tuning center setting for pitch set at 3 means that as the game gives out a value of 3 I'm at full travel of the actuator. If I set a tuning center value of 6 and the game gives a value of 3 I will be at 50 % travel? What I can't wrap my head is how this relates to the speed of the actuator. I guess it might just be better for me to accept this as a "just how it is" kind of thing.

To give a concrete example for what I am struggling with is when I like the feeling I get with a certain tuning center value, but the output value of the game is higher than that. Lets take pitch as an example. After tuning pitch I've found a tuning center value of 6 which feels responsive and good after running a medium car on imola. When I then take the same car on the ring where pitch can reach around 9 or 10 I experience that pitch axis goes dead when reaching 6 and it feels disconnected because there should be motion from the 6-10 range. I then put a tuning center value to 10. Then I get the complete range from the game, but motion feels less responsive and to soft. How do I get the feeling of a 6 with the range of a 10 is what I'm asking I suppose?

Will increasing speed and acceleration in the interface settings be the way to go or is it possible to solve this by manipulating the axis assignment percentage in some way?

I need more time to deeply think about your problem, but I think to achieve what you want you expand Min/Max range and increase Axis allocation for the desired DOF. To take your example, suppose orgininal min/max is:
[-6;6] and Axis allocation [50%]
if you change to:
[-12;12] and Axis allocation to [100%]

your sensation for forces up to 6 should be similar, but yet you would have more range for higher forces.

I would stay away from speed adjustments, in fact I'd even refrain from Axis allocation changes, I'd rather keep this stuff in ProSimu recommended values, as they're pretty good, and instead work on Min/Max and DOF sliders in Game Manager.

And once again, I could be wrong, but this is my understanding so far.

I believe you are right and I've tried this for a short time today and it seems to be the way to solve my spesific problem. Higher tc values combined with increased axis value for the given dof.

From what I gather how prosimu intend you to tune simtools is different to what the FAQ here on xsimulator tells you to do. FAQ states: Keep axis assignement to total of 100 % to make sure to avoid clipping because the actuator runs out of travel. You might run into clipping of the telemetry depending on how you set the tc values. In a racing sim mostly related to roll and pitch as I have tried to tune this way and for the g-forces I get good motion even though I have chosen to set those values lower than the output from the game in certain cars. For g-forces it works for me because I find the initial low-G movements (from 0 to 2), but of course will loose the g-forces present when driving a formula 1 car in a tight bend (G's higher than 2 and further) for example. I will attach my settings that I'm fairly happy with after following the intended way of tuning posted by Noorbeast.

The way I've come to understand prosimu tunes their simulators is with high axis values and then you have to tune your tuning center values way up from the recording you do while tuning center is gathering telemetry or else it will feel like a riding a wild bull. I guess that is why you @The Iron Wolf use a multiplier of 2.5 for your tuning center values right? As far as I can tell you might risk reaching clipping of the actuators travel and that way loose information from the sim, but if your tuning center values are high it will mean that a given DOF will never use the full length of the axis assignment value.

Example: Tuning center value is 6 degrees for roll. Axis assignment is at 25 %. If the actuators total range is 10 cm that would be 2,5 cm movement available to that DOF. Game outputs a 6 which is full travel. So actuator will move 2,5 cm.
Next example: Tuning center value is 12 degrees for roll and axis assignment is at 50 %. What you have done is doubling these two values which means that total travel is 5 cm, and as the game outputs a 6 which is half of 12 it will only using half of the 50 % assigned to that axis meaning actual actuator travel will be 2,5 cm, but with headroom for additional telemetry being conveyed through the actutator if the game outputs values above 6 and 12. The risk here is that you run out of actuator travel capacity. Now does this mean that in each of these cases the actuator will respond with the same speed and force?

@Lasagna Smoothie Sorry if this comes across as a high jacking of your thread. I hope you can get some meaningfull info out of it, atleast that was my intention as well as clarifying things for my self, unless I'm the only one wondering these things. I have posted my settings for tuning with interface settings at 160/160 as prosimu recommends for project cars 2, but have tuned according to the FAQ here on xsimulator to the best of my abilities. As I've said I believe this way of tuning is different from what prosimu intends. If you have that game try those settings out and please tell me (us?) what you think

I am glad it helped. I am perfectly aware of ~100% axis alloc requirement, however I am yet to observe any clipping. This is one of the things I couldn't understand in SimTools yet, but I am still learning. I observed Virtual Axis output test and I never saw any axis maxxing out in racing condition at 60% alloc per DOF. If I adjust axis to ~100% sum, I receive very fast and very limited motion,that is not my preference.

I use 2.5 multiplier because if I use captured Min/Max ranges I get:
* High motion in average scenarios
* No headroom for extreme forces, crash, off road etc.
* I also run into problem of adjustability of Game Manager profile. +-50% is not enough for drastically different vehicles - Karts vs F1. Here are my math notes for myself to explain the problem, and benefit of expanded range:

Expansion of Min/Max works, because:
[-1, 1] -> Min adjusted [-1.5, 1.5] Max Adjusted [-0.5, 0.5]
[-1.5, 1.5] (scaled by 1.5) -> Min adjusted [-2.25, 2.25] Max Adjusted [-0.75, 0.75]
[-2.0, 2.0] (scaled by 2.0) -> Min adjusted [-3.0, 3.0] Max Adjusted [-1.0, 1.0]
[-3.0, 3.0] (scaled by 3.0) -> Min adjusted [-4.5, 4.5] Max Adjusted [-1.5, 1.5]

I do plan to experiment with ~100% allocation in the future.

Hehe well then we will do a switch. Going to do a fresh tuning session with prosimu recommendations for axis and use higher tuning center values when I get home from work tomorrow. I think trying both “schools” when tuning is helpful to understand how simtools work and get to know your simulator rig. In the end what matters is the result and that any compromise we might choose to do with regards to tuning is a result of a conscious decision knowing the pro’s and con’s. It’s easier for me now to think about what I want to achieve and know what to adjust to reach that goal. This conversation has been enlightening for me at least, so thank you!

Yes, I really appreciate this communication as well, because this stuff is a lot to learn and I always find it better to talk to people rather than not I also appreciate you partially confirming my observations so far, means I am not entirely crazy.

Please share your new insights as well. One interesting area is how Axis limiting and allocation impact the calculated Virtual Axis in the Output Testing. In particular, suppose we limit Axis X to 50%. Does that mean that Virtual Axis slider will never go past +-50%? Another area is how to detect clipping? Does clipping mean that Virtual Axis reaches 100%? Or it can never go above allocated amount + limiter? A lot to figure out still

Hi Petro,

Thanks for chiming in. I apologize for the slow reply. That's quite the comprehensive response you've provided. I can see you've given this topic some serious thought. To answer your questions:

-The racing sims I've currently running are Project Cars 2 and Assetto Corsa. I've also played around a bit with Assetto Corsa Competizioine. but I'll probably wait for that to get out of early access.

-Like you, I've assumed that axis limiting and force percentage limiting are essentially different routes to the same destination. So if you limit the axis to 50% then this is roughly equivalent to changing surge/roll force percentage from 60% to 30%. However, we could be totally off the mark in this regard. The SimTools user manual simply states that force percentage is "the amount of selected force that will be combined into the final output for this axis." Regardless, I've been playing around with these values and I believe I'm presently in a good place in terms of limiting surge/sway to more reasonable levels.

-I have made some modifications to the speed and acceleration of the actuators. These values were recommended by ProSimu to another user and I decided to adopt them as well. I'm not sure if they make much of a difference from the base values, but I've been happy with them.

-Thank you very much for your very detailed explanations and suggestions. I really appreciate it. I also really appreciate you linking that video on viewpoint/horizon locking. I'll certainly take that into consider when using VR.

Here's some screen captures of my settings for Project Cars 2. I set all of my actuator limiting back at 95% and by messing around with the "force %" for sway and surge as well as the "main level %" for PC2 I believe I've found a good compromise between solid motion and feeling too much like a roller coaster. Please feel free to try them out if you're so inclined. Thanks again!

Hi Iron Wolf,

Thank you very much for those excellent suggestions. I took them into consideration in modifying my motion profiles and I feel I'm in a great place motion-wise at the moment. Thanks again!

Just had a moment to read this more carefully, and it sounds like you understood ideas behind my reasoning and we are on the same page.

Regarding speed, I would assume it should be the same. In both cases of your example requested end travel is 2.5cm, and I would assume speed does not depend on the scale used, all that matters is how fast force value game send changes. Scaled or not speed of change should be the same. Of course, only SimTools dev can clarify this for sure

I am using axis allocation as recommended by ProSimu at the moment, so there's no reason to share that, but attaching my current configuration for rFactor 2. I'll probably be dropping my multiplier to ~2.2. I don't use global smoothing (Main level 100%), but I suspect I'll use a bit of per-DOF smoothing down the road, most notably I suspect surge needs a bit of smoothing.

@Lasagna Smoothie Hello and happy new year! I've gotten some time to tune a bit today and I've tried your posted settings. One thing I noticed is that with main level at 94% I get this lagging feeling. One of the reasons I choose to increase speed and acceleration in interface settings is to get more crisp, aggressive and defined motion. Especially for small bumps which I think helps the feeling of acceleration. The problem I have with choosing to aggressive speed and acceleration is that I in some cases get actuator lift off. The prs200 actuators are powerfull and haven't found a way to use them to their full potential with higher speed and acceleration without the one or more actuators lifting from the ground in certain cases. It breaks immersion and I wonder how healthy this is to the actuator in the long run. Is this the reason you choose to run main level at 94%? I would suggest you try to lower the speed and acceleration to a level where you can run it without reducing main level % and still avoid actuator lift off. You might experience more detail in your motion than you currently do. I have lately wondered if I should tune simtools to a formula 1 car on the ring. I tend to prefer as aggressive tunings as the sim can handle without actuator lift off and perhaps tuning to the extreme is easier than doing it with a medium car. If not just to identify limits because if that work it works for the most demanding (of the rig) car/track combos. Its easier for me to manipulate the tc values directly because I find it easier to relate to those values. When I start tuning in game manager my calculation skills on the fly and lousy short term memory stumbles a bit and I loose track of where I am tuning wise.

@The Iron Wolf Happy new year to you as well! I just want to say that I agree with your understanding of the the things we discussed here regarding simtools axis values and tuning center values. If it came across as I was disagreeing or trying to lecture you on how to tune your rig put that down to language and wording issues on my part. English is not my first language. I've tried the high axis and high tc value and I get good motion with that and can't seem to notice any clipping present. Today I've done a lot of tuning and I find that after a while I get de-sensitized to my tunings. It kind of feels like a fall into a tuning-black hole. Going to take a break and come back to my current setup tomorrow. I find that helps sometimes.

Will post settings after letting them stew for a day or two

@noorbeast Thank you for sprinkles of simtools wisdom and happy new year!

Petro

What simulator are you optimizing for, Petro?
Do not worry about our communication, I grew up in Europe, I know our mentality and I also know not everyone reads docs (but I do), so reminders of recommended settings are annoying, but understandable

I'll also post evolution of my understanding of SimTools on this thread, so that we all can learn from each other. But in general, I am truly amazed by both T1000 3DOF and SimTools, exceeded all of my expectations.

Hi Petro,

Thank you very much for your feedback and insight. Also, no need to apologize for hijacking the thread as you put it. I'm simply happy to provide a foundation for discussion.

To answer your question, yes, part of the reason I set the main level lower was to reduce the frequency of actuator lift off. It's noisy, immersion breaking, and as you mentioned, probably not great for the long term health of the actuators. Unfortunately, I don't think it can be completely eliminated without drastically reigning in the platform as even on much lower settings you're still likely to get some actuator jumping during a collision or off road mishap. It would be nice if the actuators simply screwed into the platform so this sort of thing could be avoided, but there's likely some sort of mechanical issue precluding this.

Regarding feeling a sense of delay between the on screen action and the platform responding, I can't say I really feel that myself. To me it seems pretty close to one to one experience even at 94%, however, we all feel things differently and some people are much more sensitive to any perceived delays. Having said this, to be honest, I don't really want a perfect 1:1 experience. When in a a real vehicle there's always a slight delay as road conditions are transferred through the car's suspension and even the most powerful engines don't instantly alter a vehicle's speed. With the main level at 100% the motion just feels too mechanical and clinical to me - I'm far more aware that I'm sitting on a motion platform. I much prefer a slightly softer motion profile that at least to me seems to better mimic the suspension and power transfer that occurs in a real car - the added benefit as you mentioned being the platform is less inclined to jump off its moorings.

I having said all that, there's certainly something to be said for ultra crisp, sharp motions, and I can certainly understand the appeal. Are you presently using ProSimu's default settings for actuator speed and acceleration? Are you also still using the PC2 values you posted above? I'm certainly interested in giving your settings a try. There's certainly a chance they'll win me over. Once again, thanks for your input.