My Motion AGAINST Motion. (And Some Ideas For Improvement) (But Mostly Rambling)
Vložit
- čas přidán 7. 04. 2021
- What are some common issues with race sim motion systems? And how could it be (slightly) improved? Purely coming from a realism angle, not so much from an immersion point of view. Probably use 1.5x playback speed so I look less dense. ;-)
- Hry
When we worked with one of the F1 teams back in 2008 they were trying to find a motion system and they said that a motion platform that gives the wrong cues is worse than using a static system, pretty much what you are showing here
This is why I buy Heusinkveld equipment, dude is my kind of guy.
his attitude speaks volumes. positive, always additive and informative.
His argument was a tad flawed though. Read my comment.
@@Gorilla_Jones nah I don't think so
Totally agree with your comments regarding the screens not moving with the platform. We built a system to compensate for this later on using a hard mounted TrackIR, which is documented on the channel. But a solution 100% should be included with the platform for the price. Something we will be elaborating on when we do the full review later on.
Also note this was a "first drive" video before we dialled things in. Since then we've dialled down the motion quite a lot.
In the case of the NLR Traction Plus system, you do have to tell it your head's relative position.
The biggest issue I have with traction loss is that it cant react fast enough to realistically simulate the car snapping back into line after a slide.
That's a great system, really liked the idea and video.
Might be able to dial it back up when you invert those axis' !
@@willdarling1 it feels ridiculous with the axis inverted. I totally get what Niels is saying but in practice it doesn’t work, at least not with the gear I have.
@@boostedmedia That's because it's wrong no matter what you do... If you invert, you get incorrect yaw but correct lateral accel. The regular axis orientation means incorrect lateral accel but correct yaw. "Correct" meaning in the right direction. The solution would be to redesign the software/hardware to do it correctly, which is mostly Niels' point in this video.
@@jacksonp.928 yep. He’s spot on about your head being the centre of motion. I just don’t see motion being “realistic” without sustained geforce. The goal for a consumer accessible system should be to give cues your brain can use to better interpret what the car is doing, and as Niels alluded to, you can train your brain to associate just about anything as long as it’s consistent.
Please more singing ☕
The only aspect of "motion" I found to be good is for suspension compression / road texture and the rig giving a sense of being in a suspend mass up down. With the rig tuned so the motion is really minimal. Not really any use for driving info but a ton of immersion , mind you so are transducers really anything that removes the totally static rig feeling.
I quite like seatbelt tension on rigs but not sure if I'd just get anoyed with it after a month or so.
No motion rig I'm aware of is fast enough to do yaw g force motions in a way that's not distracting has no reset, unless the thing being simulated is a road car / lorry. And it takes up a warehouse.
I think a DD wheel with AC's FFB or some specifc cars in specifc Sims is the only device that gives the seat of pants feel in an abstract sense that's anagouse whalst also fast enough and detailed enough to drive in a similar way to how a driver would drive from bum feel in real life.
Trust you to find any excuse to bring up AC FFB! :)
Motion helps a lot with VR for people sensitive to motion sickness. Seat belt tensioner mods are absolutely worth it and it actually feels weird to drive without one now. Simple seat mover solution tuned the right way and without an excessive range of motion is a relatively inexpensive way to get a lot of additional feedback about what's happening to the car.
Very interesting take. It all makes sense from a physics standpoint. I have developed a non-conventional motion system that uses translation for transient accelerations, and the inertial cues are all spot-on. No inverse cues whatsoever.
This is a complex topic. Thanks for the video/thoughts. This is well covered in this brief segment!
@19:33 The word is JERK, derivative of acceleration... its a key concept we try to include in all our home or commercial systems. Since you don't have 'unlimited' stroke and/or degrees of freedom and you should not try to replicate something you fundamentally have no control over. Do what you can and do it well. Don't do it all, as its not possible and has greater compromises. (ex. consistency)
@19:55 HEAVE is also very important, but also few systems out there can provide the vertical jerk required to sell the road textures, most will top out or saturate with heave, or keep heave under 1G to prevent chassis lift.
@22:33 SATURATION is also very prevalent in most motion systems. This is a good point and inherent due to most algorithms just doing basic linear translations.
The canned vs. real effect is also to be considered. As many systems use canned 'effects' to sell immersion, ex. is skid vibrations, often these are unrealistic to real drivers but good enough for most enthusiasts and it does get the point across.
To sum it up: most motion systems rely on changing the driver's orientation to generate sustained Gs, but do so at the cost of disregarding accelerational Gs. Any motion system which fails to take account of both forces will provide inherently flawed feedback to the driver.
Motion systems should not be attempting to replicate G forces.
I definitely get the feeling that motion systems have been the dancing bear in simracing. We’re too excited to have it to scrutinize whether it’s working well. Time to get serious.
I agree 100%, I'm very sensitive looking out to those incorrect things, I can't ignore them and that's why most motion rigs I've tried felt totally wrong to me and I rather not use them even if I had the money. But if there is something out there someday which does it right and is not too expensive I would totally want one.
Btw. I remember meeting you at the Expo 3 years ago and you've asked a motion rig booth to invert their settings to feel correct :)
VR + great motion compensation + large well tuned movement + gseat + active belts = most singing a different tune
This makes perfect sense. I did 30 minutes in a 2dof down in Clearwater FL and it made very little sense to me how it was moving. This made it clear to me why
Maybe the only SIM racing youtuber I have seen that actually understanding physics and motion. Very informative
So basically motion systems are simulating the 'motion' of the car, when they should be simulating the 'forces' on the driver, which are different things?
I think you are very correct!
Saved me 28 mins
In real car driving, drivers' input throw the car and his body toward the given direction and drivers body push against driver seats. Like you turn right from a straight , car is rolling counter-clockwisely your body left was against the driving seat. You feel pressure on your left body when turning right.
However in simulators based on simulating car tilting, the counter-clockwised tilting movement will give you the pressure on you right body rather than left side in the real world, so it's opposite.
The inverted force! that's why the motion platform I tried a few years ago didn't make any sense to my brain. Thanks Neil!
I have three SFX150 actuators on my rig and a belt tensioner. I never try to replicate gforce on my rig, it’s just unrealistic. I use pitch and roll only for elevation changes on track which works well when tuned. Heave on these types of rigs are amazing like you said, and when you tune heave, pitch and roll to correctly replicate the cars orientation not suspension movement on the track and over bumps you really feel connection of the car to the track. When you have heave tune well you can feel the difference in downforce when in a high downforce car.
Great video and as a motion user I am not at all insulted, rather intriqued by it :) I tend to setup my motion to mostly get feeling of the road texture (heave) and utilize the pitch/roll for the car orientation on the road. This way the transitions are not that violent and I still get an immersive experience with some extra information about the track / orientation of the car without getting those initial inverted forces. After all you do feel the lean when the car is going downhill/uphill and when having steep sloping/banking of track.
This is exactly how I set up my motion too.
Good point. Motion can be setup relative to the environment, the car or the driver. You see some sims struggle with this on banked high speed tracks like Daytona Oval. When the car banks (the environment) and turns (the car) and experiences G-forces (oversteer), the forces will cancel each other out. You have to determine, based on your setup, if you want the chassis to lean into the corner or lean with the track. In such cases focusing on what you can do well, like heave and banks would be best than saturating g-forces of the turns.
Brain implant that tricks your human gyroscope + VR.
Or fill the balance-organs with a ferro metalic liquid and then change the angle with an electromagnet looking like headphones.
btw There was a Japanese "science" video where they controlled a person by flipping his balance by using some sort of headworn vibration device that vibrated the balance organs, i`ve been looking for it but cant find it.
@@seculi7757 Rather then "spend 100K on overpriced motion rig" this is the true test of how seriously people take simracing :)
I like the idea of motion rigs but this really got me thinking... so thanks.
My only actual experience of motion rigs is trying a system that only had a few cm of vertical travel on 3 points (front left, front right and rear). I was convinced that I would come out thinking it was silly and didn't do much for the experience but was pleasantly surprised by how much it added to the feel of what was going on with the car and especially the road surface. Just that tiny little kick up and down to simulate bumps, vibrations from kerbs, etc. goes a very long way. Emphasis on vertical motion is probably the way to go, as you say.
Overall, the most important thing must be to focus upon the movement of the rig, rather than what position to move it to. The only actual forces involved in any rig anyone could ever dream of having at home are always going to be 1g of acceleration vertically, plus whatever acceleration you add to that by starting and stopping various motions. One of the keys is that starting quickly and stopping slowly can trick the senses of the occupant to only "feel" the initial kick and not the stop. However, that can only ever give you short little kicks and never any sustained forces. As demonstrated by my own experience described above, that could be enough.
The neat thing about rotation, which is probably why everyone is trying to make it work, is the resulting tilt angle. Tilting at 45 degrees should trick our senses into feeling 1g of horizontal acceleration, because the perceived horizontal and vertical forces would be equal (although at ~0.71g rather than 1g). From your analysis the problem with rotations is that in order to make them happen a rotational acceleration is required, which becomes a force in the opposite direction of what you are trying to achieve. For example a rig in a state of leaning forward would give you a slight sense of being pulled forward in your seat, great. But the forward rotating motion required to get to that position would cause an initial acceleration resulting in a force in the opposite direction, as shown by many of your examples. So rotations could only be used if they are done slow enough so that the acceleration (and deceleration) is negligible. As you say then, probably suited for simulating airplanes but not cars.
But this got me thinking further, as you talked about where the center of rotation is or should be. In general, available rigs have the center of rotation at the base below the occupant resulting in this inverted kick of the head whenever a rotation is initiated. I'm not 100% convinced the focus should be fully on getting the correct sensation at the center of the occupants head, though. A lot of the feel of driving a real car is more at the hips and spine I think. Obviously it's really a combination of what you feel throughout the entire body, but in any case. What if the point of rotation was above the head? In such a case leaning the rig forward would also move the entire body of the occupant rearwards, which should make the forces align so the initial kick no longer is inverted. The kick would also be larger at the hips than the head, because it is further from the center of rotation. It would probably require a much larger rig for the same range of motion, but with the forces aligned in the correct direction maybe a smaller range would still give a better sensation?
Wicked smart. People should be thank you every single day. 🙏
I'm a big fan of your work. As a mechanical engineer I'm also always looking for the most accurate physics in about everything.
Regarding these motion rigs and your very clear explanation, I'd say one could decompose the forces in two domains: transient and steady(ish) states. So acting in current motion rigs, the software could be tuned to respond in opposite directions to high and low frequency accelerations. This way fast transients and steadier forces could all be somehow represented (perhaps with a "hole" in between). The trick would be finding this separation frequency and the transfer function as a whole. I'd be very interested in developing this concept (and some others) if I had the opportunity.
Keep up with the good job! And keep throwing us spreadsheets.
Thanks for pointing this out! For braking, on one hand it make sense for cockpit to pitch forward because the suspension of the car compresses as the inertia of car's body will move forward but in the absence of surge you don't get the effect of the head's inertia forward which probably the dominant sensation. Torso and head move forward together, and force of seat against back decreases, and force against seat belt increases.
Excellent input on this topic! Haven’t really seen many videos that explain this as well as you do. I think a 6DOF (has independent actuators for moving around and along the axis) reverse motion system that has very fast long travel actuators combined with something like BeamNG’s “inertia strength” setting in the driver cam would be the best attempt based on the points you have made. I can see why some people would think their rig has to move the same way the car is, but that is obviously not the case because the rig is not moving at a constant speed, so no circular motion forces can be felt! Hope you have more vids this like this in the future!
fantastic explanations. I built a sfx100, but as a karter and long time sim racer, it just feels wrong. so I sold it. Thanks for discussing the proper reasons why I always just felt it was wrong :). if the motion developers spent some time in karts they would walk away and see the flaws as well, karts are (esp 125cc single speed and shifters) perfect as they have no suspension, so pretty much exhibit all the forces you speak of with out pitch and roll. cheers niels, nice work
Christmas list is getting short ,,, great video
I just bought the Ultimate Pedals and Sequential and couldn’t be more amped to get in literally a couple days...Thanks Niels! Wow the service!
Ultimates are unreal, so good. I suggest you calibrate the brake to start after the spring has been depressed, the spring compresses before the stoppers as it's meant to simulate the gap between pressing the brake and the pads touching the disk, you can customise how much travel the spring has too, and honestly I'd just jump straight in with the green stoppers included in the box instead of the softer ones that are already on the brake.
Enjoy, they are so good :D
@@illiteratethug3305 Thanks for the tip!
I've been saying this for ages.
I want to build my own motion rig, but instead of just making it tilt with the cars suggested suspension movement on track, I want to make the seat move slightly in the opposite direction to simulate body inertia within the car.
For this you need two opposite movements, one with rig and one with seat.
For example when braking hard the rig will pitch forward, but just a split second before that happens, pretty much the very instant you touch the brake, the seat will pull backwards. Therefore pulling your head forwards instead of pushing it backwards as you have pointed out.
Same as simulating oversteer or cornering roll......Create yaw around the front wheels, or lean on the loaded suspension with the rig, but just as you start to turn the wheel, tilt the seat slightly in the opposite direction to simulate your body movement within the seat. This way you can sustain simulated G-forces because you would in effect be leaning against the tilt of the rig for the continuation of the corner.
It may take a lot of witchcraft of which I know nothing (I'm just a lowly carpenter)......but I'm determined to do it.
My thoughts exactly Niels. Good video. Scott, Sala Motorsport.
The biggest challenge is always creating the actuators transfer functions from the physics correctly
I second this motion.
Good video, thanx Niels. Your thoughts are inline with mine, my pretty expensive motion actuators are fitted to the rig, but I am not sure I am going to finish wiring it up even, as I was pretty disappointed when testing the systems on display at SimExpo during 2016. Cues are in the wrong direction and very muddled.
I will probably just design and use my own quad belt-tensioner system, so at least I won’t let the servo-drives go to waste. That combined with a pneumatic pillow in the back and good vibration transducer under the seat, will probably be better than anything else, tbh.
Keep up the good work though. Remind me to sent you over a nice bottle of Aussie red ;)
Cheers,
Beano
This is the reason why big end expensive motion platforms mounting a tripod on another platform that can move around on the ground. You can not simulate lateral forces only with pitching your seat.
You are my favorite physics teacher@Niels
Love your videos. Wish i could afford your equipment!
Thank you!
You describe it well after 14min and it's something I thought about on the Expo when looking at a giant right, it would need the actuaturs hanging from the roof instead of the bottom and act a bit like a pendulum, another thing which could work is if the lower platform could make huge horizontal movements to the back when braking while the upper platform tilts to the front.
When your car pitches forward due to braking you DO get your head pushed down into your shoulders in a very small way...because the nose of the car drops and the rear end (may) come up. So...nothing wrong there. Should the seat move backward to drop the head forward? Yes, agreed.
Motion simulators, IMO, are all about giving the body and mind initial impressions--shocks--which ideally correspond with what we would experience in the real world. After each shock they should--quietly and gently--retreat back to the static position in order to deliver the next shock...and not much else. Except at NASA, they generally can't anyway, as there is only so much you can do with even an expensive motion platform. Until we can all afford our own "Vomit Comets" anyway.
Good vids. Always interesting to hear your thoughts. Keep them coming.
Great video as always. A term that I like more regarding sim racing is convincing over realistic.
Another thing most cockpits don't do right is heave on bumps, crests or landing after a jump because they simulate the cars suspension. Most go down when you land after a jump but they should go up to give you the feeling of the impact.
How is it possible that manufacturers of these things don't understand this?!
I love the double parenthesis as a title!
A Gseat can offset the initial opposite tilt of the platform to provide the initial proper side push and the tilt can help with sustained G's
I've been on the fence about investing into motion.. you made you really interesting points.
It's worth the investment if you do it with the right expectations and approach. Not all motion rigs are created equal and there are an infinite number of settings combinations in the software. You can tune out most of the exaggerated effects that Niels is talking about so that you just get short, sharp cues about what's happening. Motion rigs can transfer kerbs and bumps and road surfaces with remarkable fidelity, and there are simple seat belt tensioner mods you can make to dramatically improve immersion. Niels completely omits VR and how motion rigs can help people who suffer from proprioception issues (motion sickness) to overcome those.
i tried full motion, seat mover, buttkicker only, on screen and in vr... best setup (for my experience) is VR + seatmover (nlrv3) + buttkicker + windsimulator.... it worked by far the best for me in terms of realistic feeling... the seatmover in VR just triggers the right queues in my mind to make me feel what i would kind of feel in real life...
Hey Niels :-) Your analysis is spot on!!! I am also looking at the motion-sim scene from a more professional angle and I agree: 95% of hobby motion sims generate more false cues than correct ones. The software to do it right is available however and 6DOF systems can be built at resonable cost. The limiting factor is mainly the inability of builders to understand the interdependence of the different DOFs driven by a system of highpass and lowpass filters. I would try to put it somewhat euphemistically: There is a lot of unfullfilled potential in DIY-motion-cueing :-D Contact me, if you want to discuss deeper.
You my man are spitting truth. Love this.
Great video Niels. As is often the case people get carried away with cool things at the expense of common sense. When i used a motion rig at the Hungarian Grand Prix a couple of years ago had similar thoughts to you. I loved it but in terms of truely similating anything other suspension movement, most systems don't do the job and cost a fortune as well. I've ordered a Geko GS-Cobra seat as i believe this will simulate sustanded G forces on the body through turns much better at a 10th of the cost. I'm not sure about braking and acceleration yet until i experience it for real. Perhaps a belt tensioner would be a good addition. The icing on the cake would be the suspension motion but the cost of it means i will have to do without for now. Perhaps some sort of Giro based system allowing 360 degrees of motion would be better for a motion rig.
that was one of your best videos and explanation i thought i was the only one who notice the inverted quest and the 0 sustain forces i had it always in my mind that's why never invest in them how ever i have tried to a friends of mine his sfx100 and the vertical movement on rally is pretty okay and fairly accurate ..i don't i am still very skeptical though
Hi Niels, what about G Seats?
I love mine and hope he's engineering one:)
Tuning your motion setup correctly is essential. Iv owned a motion system for a number of years and can’t race without it now. The combination of the FFB from the wheel and the emulated G forces from the seat allows me to be more immersed and consistent especially in long races. I couldn’t go back to a static seat setup.
B..b..but you’re not allowed to have fun it’s must be 100% realistic at All times /s
@@dunnyzed6953 The fun is in getting as close to realism as you can. Hence the term “simulation”
@@tradesman1000 I was being sarcastic. Hence the "/s"
@@dunnyzed6953 my bad. 🥴
@@tradesman1000 Haha it's all good :)
I was thinking this also for some years now (well not in all your details!!) , watching YT vids about motion platforms. Now I am not thinking I am stupid anymore!
So you basically want an omnidirectional swinging ship. Though you probably don't want to go upside-down :P
So maybe put the sim in a bowl and move it up the edges.
well, this was really instructive,also i fell less bad for not having one of those motion rigs that i can't afford.
Here is the solution: Replace the fluid in the player's ears with magnetic fluid. Headphones can be developed that have a ring of electromagnets around each earpiece. The sim can send signals to activate the correct magnets, pulling the fluid, and tricking the player's sense of balance into making them feel like they're being tilted. Advantage: no longer requires big space-consuming actuators and big frame etc. Disadvantage: probably toxic and or corrosive, real life balance probably ruined. Some sim racers would do it anyway.
Interesting. When I was thinking about this my conclusion was that you would actually want the center of any rotation above your head so that you can even translate the moment of the movement itself to the force to get the initial kick at least since gravity itself can't possibly emulate that. That would lead to system more like a swing. I still don't understand why pivoting around a head would be better than that but I definitely agree in general.
Couldn’t agree more Niels! And on top of that latency and speed is still an issue. In a real race car moves like breaking, hitting a kurb or heavy bump are near instant.
which also changes with the amount of mass on the simulator adding further latency and inconsistent response.
In short stick to tactile feedback, so you can disconnect the suspension feedback (curb rumble, etc) from the steering wheel's FFB to the tactile transducers, which will give you more accurate and precise feedback through the steering wheel.
Hallo Niels, ich teile deine Meinung und Ausführungen in mehreren Punkten !
Viele Motion-Simulatoren für Rennsimulation sind leider nicht wirklich zielführend von der Konstruktion und Einstellung !
Ich beschäftige mich nun mittlerweile über 6 Jahre mit diesem Thema und habe > 30 Motion-Prototypen für Rennsimulation konstruiert und gebaut. Meine eigene praktische Fahrerfahrung mit Rallye und Rennfahrzeugen haben mir dabei sehr geholfen.
Oftmals haben mich meine physikalischen Ansätze und kinematischen Umsetzungen in die falsche Richtung geführt.
Mann sollte sehr genau differenzieren, möchte ich Fahrzeugbewegungen oder Fahrerbewegungen abbilden oder antrigern.
Ich selber bin der Simulator-Typ des Antrigerns.
Daher konstruiere und baue ich mittlerweile meine Simulatoren fokusiert auf meinen renn-geschulten Popometer bzw. Gleichgewichtsorgan. Hierbei bevorzuge ich kurze und eindeutige Impulse als Informationsgeber, die ich so mittlerweile sehr gut in Punkto Fahrzeugbewegung und Fahrzustand interpretieren kann.
Daher bin ich Verfechter von Motion-Modulen, mit klarer und eindeutiger Motion-Informations-Zuordnung, wie z.B. mein Sitz-Traction-Loss, Air-GForce in den Sitzwangen oder Längsverfahrschlitten für Surge.
Ich bevorzuge alle Informationen wie schon erwähnt, als Impuls und nicht als reale Körper-Beschleunigungen und oder Bewegungen.
Die Motion-Auslegung für VR-Brillen sollte nach meiner Erfahrung eine Andere sein, als die für einen Bildschirm.
Motion und Bildschirminformation sollten möglichst synchron laufen.
Das FFB vom Lenkrad kann und sollte beides nach Möglichkeit unterstützen.
Es bleibt Simulation und meine persönliche Herausforderung ist, diese so real wie möglich auf mich wirken zu lassen.
Manchmal ist weniger sogar mehr !
Google Übersetzer:
Hello Niels, I share your opinion and remarks on several points!
Many motion simulators for racing simulation are unfortunately not really effective in terms of design and setting!
I have been dealing with this topic for over 6 years now and have designed and built> 30 motion prototypes for racing simulation. My own practical driving experience with rallye- and racing cars helped me a lot.
Often my physical approaches and kinematic implementations have led me in the wrong direction.
You should differentiate very precisely if I want to depict or antagonize vehicle movements or driver movements. I myself am the simulator type of driving.
That's why I now design and build my simulators focused on my race-trained popometer or balance organ.
I prefer short and clear impulses as information providers, which I can now interpret very well in terms of vehicle movement and driving condition.
That is why I am an advocate of motion modules with clear and unambiguous motion information assignment, such as my seat traction loss, Air-GForce in the seat cheeks or longitudinal slide for Surge.
As already mentioned, I prefer all information as an impulse and not as real body accelerations and or movements.
In my experience, the motion design for VR glasses should be different from that for a screen.
Motion and screen information should run as synchronously as possible.
The FFB from the steering wheel can and should support both as far as possible.
It remains simulation and my personal challenge is to make it as real as possible to me.
Sometimes less is even more!
Niels is right of course, he always is. But there's one aspect he didn't touch upon, and that's just the thing that makes it believable while driving. Your head isn't rigidly attached to your body and it's the realtion between head and body that dictates to a big degree what your mind perceives to be the angle of motion.
Your vestibular organ is inside your head and works directly with your neck- and eye muscles as an autonomous gyroscope. What I think might be important in making you fill in / augment the motion cues is the muscular tension used to keep your head up straight. So when you turn right, the body leans left and your right side neck muscles work to keep your head straight. I think this muscle tension to keep your head level is perceived as needed to overcome the lateral force on your head from "cornering", aka G.
I have the GPL-lean as it is called, leaning your head into the corner to alleviate stress from the g-forces on your head. I do it on a static rig too, and in my road car and on my pushbike and even when doing the shopping trolley drift around the supermarket. It autonomously happens in my D-Box sim and, even if it's 0.05G it's enough to hint at a cornering force acting on my head. Same for braking/accerating.
This is all uneducated ramblings I know, but there is something that makes it believable when my rig dips under braking and heaves when hitting the throttle. There is one big pre-condition to all this and it's VR. Your eyes MUST be in the same loop as the rest of the senses. I've tried many motion rigs now and every single one that doesn't have VR is a waste of time and money and feels like a carnival ride. VR is key.
great video Niels!
Super interesting video thank you. 😊
Motion seems difficult to fit with fixed screens. Either screens should be on the rig. Or VR should be used.
I think rotational DOF should be treated differently for they can be completely replicated to match the track, eau rouge in Spa for example.
Translational DOF (surge sway heave) are different. Heave can most surely be well replicated.
Surge can’t, but I do think low latency can inform well the driver. Coupled with a precise belt tensioner, it may help for trail braking.
Sway is definitely another thing, I believe the double traction loss system should be used with the driver above the rear end.
The front end leading the corner giving yaw and under steer informations.
The rear end being used only for oversteering.
That makes it incompatible with huge triple screens like boosted media will has.
Tough area 😅
As one physics nerd to another ..you are right on the money.
I've been in the sim game for 20 years... I thought I was alone thinking these thoughts
Ultimately, I'm not getting in a race car any time soon so I'm not super concerned, however I will be watching this video again when I eventually get a SFX100 style setup - it's in my budget for next year. Planning to aim at vertical motion & pitch without too much roll (likely none after watching this), add in a static mounted belt (dumb tensioner system), and leave my transducers & wind sim going. That with VR should be pretty fun - can't wait to be racing Formula Vee's in a sim rig that cost more than the actual car :D
Here’s an idea to deal with the problem of saturation. Take your entire platform and stick it on wheels with motors. Simulating acceleration then becomes simple - you speed the whole thing up. Similarly, braking is reduced to slowing the whole thing down. To simulate cornering forces, you could add some kind of way to steer the front wheels. Once you have all that, all you need is a large enough space to run the thing in. A larger space would remove the risk of ever inning out of range of motion.
More seriously, a lot of these aren’t really any more sophisticated than the arcade rigs that Sega in particular used to make in the 80’s. The motion versions of afterburner, space harrier, and power drift were fun, and all they really did was pitch you about the way a lot of these rigs do. I never did get to try the really fancy rug that Sega came out with in the early 90’s. I think the system was called R360 or something like that; it literally weighed a tonne, and could literally flip you upside down, but it was so expensive that none of the arcades I grew up around could ever afford it.
And there I was contemplating how I explain to the other half, why I think I want to drop about ten grand on a dbox system! Thanks Niels. Crisis averted. 👍
I always raised the point that accelerations are being wrongly represented in consumer based simulators, but every time I’m being hit as a hater or jealous person 🤷🏻♂️
well i just left a long comment saying that vr goggles already give same immersion as motion rigs and with way less money, looking forward to all the responses :D
I have thought about the belt tensioners before, but it's really tough not to strangle the user with it. The advantage of real belts is that; even when they're tight, they create extra space in the back of the driver when under tension. Belt tensioners remove this space. You would say, add some actuators to the seat to give back the space, but that takes the tension of the belts again. Essentially, the strangling is exactly how belt tensioners create sustained force...
Wauw, verhelderend Niels, heb genoten van je uitleg. Speelde met het idee om een SFX100 te maken, daar denk ik nu nog wel een keer over na. Ik heb ooit met een buttkicker gereden en de schakelmomenten voelen door je stoel was al een enorme toevoeging voor mij! Hoop dat Heusinkveld nog eens iets moois uitbrengt.
Waar ik ook benieuwd naar ben is het VR gedeelte, wat als je een VR bril hebt die zich beweegt op basis van INSIDE OUT registratie..... dan zou je als een gek door je cockpit gegooid worden.
Yep, very much my situation as well, except I already have most hardware fitted to my rig and busy wiring the PT Actuator monster actuators. But I will probably not complete it and sell it, rather just add one buttkicker under the seat. Also agreed with the VR aspect, my Reverb G2 is already battling with tracking in stationary mode, will drive me insane in motion :(
PS: You can reply in Nederlands, ek kan dit lees/verstaan., no dramas.
Boosted media did make a solution to the screen not moving when the platform moves
and RF2 solve it natively. working perfectly with NLR systems
I own a motion system... I approve this message! :) Super fun, not the most "realistic". Less is more with motion!
Agreed.., 🤔 I second your position regarding this debate.., my good man..!!! 👍😉👏
But the point is: is it worth the investment? I'm thinking of purchasing the Next level v3
@@joseluismaravall1429 No, sim gear like DD wheels or hydraulic pedals or motion is really an "investment" they are all toys. Neil said motion is fun and great for immersion which is true. But do not buy it if you are wanting lap time improvements or a "realistic" experience. Also Neil is one opinion, just check video of people who have sim rigs :)
@@joseluismaravall1429 I have this one. Small foot print, super quiet and really adds to the immersion, for VR anyway. And they a great with coming out with software updates.
@@CombatSantaLTD I'd play using VR, so I appreciate your feedback. Guess I'll try it
Curbs, Shifting an so on you can make with Body Shakers. I have 4 shakers , one for every wheel. You cannot make crests with them , so a combination of 4 post rig, and seat belt tensioner would be a good idea
Thanks for posting this. I've been saying these same things about motion rigs as well, and will reply with this video the next time somebody @s me with a video of some motion rig and tells me "you need this". Biggest flaw IMO is trying to simulate translational forces by using rotation.
In Flypt Mover software you can adjust the center of rotation. You can also adjust the yaw center of rotation. It works really well with a 6dof hexapod.
The two big problems with any type of yaw or slide is the range and the recentering movement that has to take place every time it moves out of origin. Other than being laggy, once it reaches the end if its range, it will stop while the car might not be done sliding, which is killing. And then, while the car is going straight again, you totally notice the traction loss returning to zero. That feels like it's turning while it's going straight ! Completely unusable.
With that said.., maybe a 6DOF Hexapod with a (360° 7DOF) underneath..??? 👍💯🤔
@@normanrobinson1932 No 😁
I have a GS5 G Seat. Not perfect but it helps with immersion for sustained g forces and for bumps. I like how it signals lateral traction loss very clearly.. They have a compatible belt tensioner that augments braking forces and cornering a bit too; I have not bought it yet but have read some good reviews on it. I have tried a dbox and it replicates vertical suspension travel nicely but it is pretty expensive for that feature even with the recent price drops of some 40% from historical pricing.
The GS-5 is getting a lot better with the G-Belt and without I did some tricks to make it better. Be sure the 'maximum G-force to consider' is lower than 2G for more effects. It's a bit hidden under every force-tab and 6G is default in case you've missed it.
With a passive belt-system I switched the bottom- and backplate-function because it feels IMO better getting pushed into the belts while braking and not while accelerating. Certainly not ideal and the reason why only with a belt-tensioner this system feels complete for what it does.
I think you might consider a "PT Actuator 6DOF" setup in combination with the GS5 G-Seat..!!! 😎 That would definitely satisfy all of the attributes he's stated as required in the video above..!!! 👍💯😉
I have a D-box system and i set it up to almost only the heave function (up and down) because this motion is just like your ffb it is always there.
I do no use Roll (left to right) i agree with Niels it adds nothing.
But i use pitch (front and back) only because i attached my seatbelt to a lower frame that does not move with the rig, so i can feel the difference between braking and trailbraking and has long has i brake i feel this force and this works really great.
But to really trick your mind you have to use VR for me it does not work on screens and for sure not on screens that do not move with the motion.
Does it make you faster, for me not it just adds to the fun and stay’s a very fun gimmick.
When i really race i do not use it.
I agree that I’d rather have that slide forward and backward motion and the up and down g-force motion over the side to side motions. What systems are out there that i should look at for this type of movement. Im looking to build my first motion rig so I’m exploring all options. Thank you. Informative video.
Visibly you've never heard about simxperience. I have the GS-5 motion seat and the G-belt and it works wonders in doing sustained g-forces in a very dynamic (high frequency if neaded) way. Please check that out. I love it...
very interesting - thank you. I have a DOF reality. I find the best results for immersion is with it turned WAY down. Any lateral sway as you describe just feels all wrong ( i used to drive rally cars and it feels anti intuitive HORRIBLE ) . Trying to simulate lateral simply G doesn't work and feels weird if you know what a car really feels like ( i never understood why till this video ) . just turn it Off is best by far. Where the immersion does work for me is in using it in moderation for gear change haptics , accelerating / breaking and effects like rumble strips, If you just do this is is much more immersive ( slower probably but hey who cares ) cant really comment on vertical G as my system isn't built for it
If you feel the motion is lagging, I would add some transducers (if you haven't done it already) to add some initial feedback and also pseudo-hieve-effects.
Interesting video! Could you do a video about transducers or bass shakers?
i own a motion platform v3. and i agree totally. breaking, acccellerating, corner forces and so on feel awkward. i turned it almost off. whats very cool is things like bumps, curbs, collisions, elevation changes... so i still think for that stuff the plattform is really really cool...
Yup fully agree that is really immersive and doesn't 'mess'with you in a potentially incorrect way.
maybe try reverse the roll and pitch in motion post-processing section? it helps a lot
Interesting. Thanks Niels.
Can't believe you never included mine !!
edit: im glad my sim would pass all of your ermm ....observations, i have COR way out the front so you do get traction loss movement only on the back. ohhh man dont even know why im bothering lol
I cant either Craig lol
I was hoping to see your rig show up. Yours is ideal in many of the concepts that he's speaking of.
I hadn't seen it!
@@NielsHeusinkveld well you have now at least haha
This has made think, would a solid solution be a dual axis linear system?
Imagine a platform with 2 layers. The top layer is the forward/backward one that you showed regarding braking.
That top layer is then mounted on top of a traction loss system similar to Boosted Media's, but with the settings not inverted.
I guess a third layer would be the 3 or 4 post setup for vertical forces.
Either way, it looks to be well outside the range of most consumers. Doesn't solve the sustained G issue though.
Sim racing garage has it pretty close with the 4 corners acting as suspension and the surge in conjunction with the seatbelts. The one down side being the incorrect que when going into a corner because you are leaning into the turn.
So it's important for motion software to allow user to put in the location of their actuators on their rig and where their head is relative to those actuators so the algorithm can translate the motion of the virtual car chassis and driver location in virtual car to be properly simulated on the motion system.
When looking at a motion system, it’s important to visualise the cockpit ( or seat ) and the pilot in space. No other reference points. Moving or not moving does not matter at that point, we don’t feel motion anyway, only forces. A motion platform cannot replicate the acceleration and movement of the cockpit, since it is constrained. What you can replicate, to a degree, are the forces felt by the pilot. Therefore, a motion platform should not a priori move like the cockpit, but simulate the forces the pilot feels. For example, when turning to the right, in space, the cockpit would go in a orbit, to the right, but the pilot has no forces acting on him, at first, he only goes where the cockpit takes him. The pilot would be ejected from the cockpit, where it not for the restraints, the seat, the belt, … So what does he feel? Well, to the pilot it feels like he is pushed to the left side of the cockpit, that’s where physics wants him to escape. In fact, should a magical force throw the pilot to the left, while the cockpit just does nothing, he would feel the same thing. He would not be able to differentiate the cockpit in an orbit to the right, or him being pushed into the left part of the seat. Now we do, on earth, have this magical force, it’s called gravity. And that is why a motion platform should throw the pilot to the left, i.e. roll left, when the cockpit turns or orbits to the right. This way, you can simulate sustained acceleration, because as long as the cockpit is in this orbit, there is acceleration. I have simplified a little, since a roll to the right also needs to throw the pilot to the left, for the same reasons, but an orbit and a roll are not the same, therefore, the platform would have to throw in a bit of extra here and there. For example, for the orbit to the left, you would also pitch the platform a touch forward, because the pilot’s back is pulled off the back of the seat. It really bothers me that motion platform experts and creators are sometimes quite bad a physics, and resort to reasoning that is unscientific.
This is exactly my problem with sim racing - WHICH I LOVE (for the record). I have experience racing in reality (amateur), and when I'm in the sim, I'm completely oblivious to my grip and the "seat of the pants" feeling of the vehicle stepping out on me. I know that a good bit of this, is the lack of resolution in my steering wheel force feedback - that is the subtle movements of a tire on the edge of grip up front or the subtle hint of the vehicle stepping out in the rear.
But what my biggest issue is that I don't get the vital "seat of the pants" feeling that I drive by. I can't push 10/10ths in the sim because all of the information that the car is feeding me is disconnected in the sim. For example - just like the subtle movement of the wheel when the back steps out, I get a feeling in my rear when the rear breaks loose and I can immediately react with counter steering when I feel it push out or let off the throttle/brake to bring it back in line. In the sim, I'm always late and end up having to dial more countersteering in or I've already binned it - and it's usually always too late.
I hope getting a better wheel will help a good bit because I haven't sat in a motion sim yet that imitates the physical cues I'm expecting. The last thing I need is the improper cues they're giving to confuse me even more.
Good assessment.
Motion compensation and picture go with you :D its cheap and work very well
In practice tilting the motion platform forward on braking doesn't impact your back or head near as much as the pressure you feel in the belts, your hands, and feet on the pedals. The braking effect can be very, very good. You are over estimating the head/ neck compression effect.
amazing analysis, thanks for posting this!
Great video, and as a fake D-Box (e.g. SFX100) user, I agree: Sustained g-forces cannot be replicated realistically...but I hope that's obvious to anyone considering a motion system. G-Seats are better at simulating sustained g-forces, but since our body's primary g-force measurement tools are situated in our head and neck, g-seats are still a partial solution at best. Also, even well designed g-seats still have the inherent issue of conflicting cues (e.g. most put pressure, via paddles/air bladders/etc, on the opposite side of the direction of the turn, but this actually moves your body in the seat ever so slightly to the inside of the turn). There is no realistic solution, but the "most realistic" would be a system that attaches to your head and pulls your head in the appropriate direction, which you then have to strain your neck muscles against to counter-act (e.g. pulls head forward while braking, rearward while accelerating, right while turning left, and left while turning right). Of course, this would need to be compatible with VR as well. This paired with a SFX100/DBox for vertical motion cues would be pretty special. Add some transducers and maybe wind simulation (for open-cockpit enthusiasts), and I think it would at least be somewhat consistent with real driving physics. Of course, all forces would be vastly diminished as compared to real-life forces. Not sure why I'm rambling to a bunch of folks that will probably never read this whole message, but if it helps even one person, it was worth it. At the end of the day, spend your money and time however you want, and take solace in knowing you can train your brain to learn even totally unrealistic motion cues...as long as they are consistently and predictably unrealistic of course.
I've had a couple of different motion solutions and I currently run with my rig static and intend to do so from here on out. My racing benefits from a static rig most definitely and while the motion adds a little bit of interest to your driving at the beginning, I find that I enjoy chasing lap times and competing so much that I don't need the platform at all. I have a decent tactile setup and some nice high end gear re steering and pedals so I am very happy. I feel that VR shouldn't be lumped in with motion as I find VR absolutely helps you and portrays exactly what you'd expect in real life. The FOV can be debatable and the image quality varies but we're at a time now where the HP headsets are delivering close to lower res monitors in terms of display. I feel it's pretty hard to do VR wrong where motion is expensive, messy, noisy, easy to be overbearing and unrealistic, can be detrimental to lap times and at time unreliable during important situations such as online races. I thought that motion was the holy grail but I've since changed my tune on that sentiment. I'd easily prefer a thorough tactile set up with some sort of active seat belt tension system OR a G seat solution. I'm sure they produce their own sets of challenges to provide realistic and authentic translation of physics and effects but I'd definitely advise to research those first.
Niels, I believe that actuators do level out during sustained G's!
It’s probably not correct, but damn it’s fun and with VR mega fun. I have the SFX 100 system and love it 😍.
Rotation should be not inverted.
The center of rotation should be under the seat.
If the rug rotated to the right the sway and yaw is correct.
So rotation should not be inverted. It messes with your inner ear !
Will should invite you to Australia, and you could try set up his system 😂 great video!! 🎉
He's more than welcome any time. Not paying for his 2 weeks quarantine though. LOL. I don't disagree with any of his points though, for the record. The examples he used were from our "First Drive" video. We haven't done the full review of the traction system yet.
@@boostedmedia 😊👍 yepp, he allso pointed out, what you said was correct. And yes, guess Australia wouldnt let him inn easily atm.. 🙈 (btw things looking a bit brighter for you now, than many many other places, well done!)
Yes, went down this rabbit hole many years ago, there are solutions.
Fighting fire with fire..let me just start by saying..."Hold On to Your Breaks"!
Now more seriously, first i am a BIG Fan of Niels, both as Sim enthusiast, as Sim related to business/technical point of view! Although being a Fanatec Affiliated Brand, for pedals and other hardware solutions, many times i use in my equipment and choose for the ones of my customers, HE products as first choice!
Many things pointed by Niels, are not new, and us that work for years, know them...what happens is that, Motion will probably be the Last things in Simulation to be achieved right!
And many theoretical things that are right and logical, sometimes are the opposite when we implement them the material world...for example, those who have experience with Motion in Simulation, know that if we adapt many of the things stated in the video (that AGAIN are true) will be even stranger than the feeling we have when implementing the wrong way!
The physics laws 1st will limit the technology just by itself, then with the generalization of SIm products, including Motion, the more options, the more users, the more setups/configs, the more yo will have good and bad examples!...but generalizations are always a bad Turn to take in approaching any subject!
I had in my equipment, thousands of casual users, but also pro Sim and Pro Real drivers, and even for these ones, they have a tough time adapting, specially to braking!...but that doesn't mean that many manufacturers out there aren't trying their best to deliver us the closest to real and better experience possible!...it's juts that there is a limit factor of what is possible, for a certain stage of current technology!
So Key word in SImulation, also always got to be...adaptation!
The same way i have to tell a Real driver that altough is great in braking the real deal, he has to adapt to Sim Pedals, the true is also for motion...one has to be realistic, give the Ilusion the best way possible of Reality...but always adapt!
This will always be a very subjective debate, based in many objective facts...and that's the main thing we can take out of this endless Topic (a bit like game physics), no mater how great a technical solution is, it always be in the end a matter of subjective personal opinion and in the end, SImulation and Gimmicks of the Real counterpart!
Sure, i'm defending my dame and line of business, but i'm sure Niels will ans also defends his in relation to the software/hardware he helps develop...and that is the natural thing to do!
Again Motion, always be the most difficult part of Simulation to achieve, but that doesn't mean that it's not possible with a GOOD PRODUCT, PROPERLY SETUPED, to achieve a GREAT Experience and add another GREAT Dimension to Simulation, without compromising a lot realism and performance...but like i said, with compromise and adaptation..like in everything in Simluation..and Hell with all the rest of our life!
But then again, it can be only a problem of personal perception, or what i suspect, the type/quality of the Wine Niels is drinking...and in that department as a Portuguese and also small wine producer, maybe i can help in mixing that ultimate Secret ingredient!😅
Keep Up the Great Work and always welcomed here in Portugal!
just cannot wait for a h shifter and smart control for ultimate
Niels, the SimX G seat is great for sustained corner G and so is their belt tensioner for braking simulation ;) Key with these two products being is that you configure them to do just those two things as out of the box the profiles just have them doing too much.
I should try some of that stuff at some point!
@@NielsHeusinkveld It great. I have the GS-5 and G-belt. Plug and play. Motion is for elevation changes only. Tactile for effects and VR motion compensation. The sum of the parts focusing on their strengths as you have mentioned in the video. Dont ask a motion platform to try and replicate G-forces. Your kidding yourself and moving away from a real car.
So should there be different profiles for VR users vs monitor users? In VR, the g-forces would be most important. With monitors, you would move relative to the monitors (inverted motion).