My sim consists of a Logitech G923 wheel and pedals, gear lever and simple desk and gaming seat that I have to secure to the wardrobe with a piano stool and a guitar speaker for it not to move (it still does a little when I push the brake). I suppose it could fall a little short of F1 teams' expectations...
@@docmain999griptape under a diningchair was my setup for years, although a cockpit and better wheelbase and all that sort of stuff adds alot to your quality of life.. it doesnt make you a quicker driver
My room mate blows next to me when there's wind and spits when it's raining. He also kicks me depending on the crash. Much more real phyisics than those "top" simulators.
The other team is McLaren. The names of Ferrari and McLaren show up in the SEO results in the short description of SAE International's website when they covered Dynisma as a startup in 2022. I wonder if McLaren decided later to keep their supplier info private, while the search engines still have that information and the webpage is still up.
I m very perplexed because ALL teams missed the ‘porpoising’ effect and coming from a mech eng with 40 years exp. in CFD / STR analysis because the repeated bouncing up and down should be simple to solve I think !
Several of the Apollo astronauts were Marine or Navy pilots so they were already trained for high stress environments but of course still needed to practice working the specific controls for the Apollo flight plan so plenty of simulators of all kinds were created, truly amazing for 1960s and 70s 😎
@@philippemichelvidori7248 it's not really just bouncing up and down mate! It's a rather complex problem that I'm not exactly competent to explain to you having only a basic understanding, but what I can tell you is that fluid dynamics get complicated really quickly and when you have to consider flow on all sides, in the venturi tunnel, around the wing, etc. and on top of that distinguish between turbulent and laminar flow... it gets utterly mad.
@@philippemichelvidori7248 The problem with porpoising is that the windtunnel models are static and can't move up or down with regards to downforce, it's specifically to prevent the model from hitting the belt they run on as it would likely explode and create massive damage to both the wind tunnel and the model. So they're stuck relying on CFD simulations which are very performance intensive, you can either run a very accurate simulation on a small section of the car or run a more complete but inaccurate simulation on the whole car, or somewhere in between but never fully both. Porpoising is also a very complex feedback loop making it especially hard to simulate accurately even without the issues i mentioned before.
@philippemichelvidori7248 porpoising IS easy to resolve, you just run the car stiffer and higher. But if you also want to be fast, you have a particularly complex problem.
@unklesuga1644 thats why I am saying that. Cadillac knew for awhile they were gonna be 11th team. As soon as Micheal Andretti stepped down that was what F1 wanted.
@blade0667 i do actually forget about their team🤣 you're right. IF Ferrari is using that Simulator and Ferrari are supplying Cadillac's engines it makes sense
Interesting video, but I would have liked it go way more into detail. “Simple” things like how they manage to keep latency low, make use of synchronization between visuals and movement of the platform, perhaps even sound (you didn’t mention that at all). But more importantly, you sidestepped *how* teams translate a part that only exists on the digital drawing boards to be “fitted” onto the virtual car and then influences the way the simulator behaves. To me, that is the most interesting aspect of this workflow.
>“Simple” things like how they manage to keep latency low, I agree, mostly because of how ridiculous the 3-5ms number is, not just for something of this scale, but in general for electronic input delay lol.
This Synchronisation is by far the easier part. Other F1 sims litterly run on a heavily guarded and Custom version of rfactor 2 some real gt drivers especially GT3 and 2 use assetto cors. So getting something that is ,,faster" probably isn't the hardest part considering the price. Also how fast motors be it servo, stepper or also electronically controlled hydronics can be controlled/adjusted has developed quite a bit in the past 10-15 years. I think the biggest part is the price and what racing series has the money for it. The "cheaper" ones (still probably over 300k) do mostly run on less customized software and something closer to heavily modded at home sim and have parts made by different engineers and developers from different companies. Connecting all these pieces in hardware as in software in a less customized sim with less budget introduces the most amount of latency most likely. This seems to be the by far biggest difference and what also makes the most difference for the divers, your average person obviously wouldn't notice these tiny delays nearly as much as a driver used to the real thing and with such reaction time.
"you sidestepped how teams translate a part that only exists on the digital drawing boards to be “fitted” onto the virtual car" I build a vehicle model in MSC ADAMS or Carsim with the new parts in it . This then gets uploaded to the simulator . The model is driven by steering wheel angle, throttle/brake position, gear selector and so on, and then outputs the vehicle state to the simulator.
@@naplesbluesrt i do have a proper rig but i can't use it as i kept stubbing my toe when i got out of bed due to how close my desk is to my bed and just how big the rig is lol
You can't 1-1 simulate the g-forces the inner ear feels in a sustained way, but you can do it to some parts of the body. Eg your skin won't know the difference between being pressed into the side or your seat in a corner vs your seat pressing into your skin in a corner. Some sim rigs use this trick to simulate the sustained Gs with movable panels in the seat. Don't think it's 100%, but it does sound like 1 more level of immersion where you could quite literally be driving by the seat of your pants ;)
You can! Just make the sim rig actually move, and to sustain the g force, it should move miles away with wheels... Ah crap I just described actually driving the racecar
@@fookapi that's with compressed air, right? not sure if that would have the response time they are looking for. but yeah, there are other ways to simulate Gs without moving the head about
@@alex8373 I’m imaging a full body suit that can apply pressure to different parts of the body to simulate forces. So unlike a traditional g-suit (which I wore as a military pilot), it shouldn’t just compress the legs. Just an idea and yes, it probably wouldn’t be fast enough.
Do these simulators have custom software for graphics and track specs? Did they laser scan all the tracks themselves and create their own sim software or are they using 3rd party data? Super curious!
@ that’s what I’m wondering. Like if you’re 2mill in a sim. Is that “enough money spent, let’s use iRacing”. Or is it, “only our own personal laser scans of tracks meets the standards for our super expensive sim”
Yeah I'm curious about the economics of it all. 8 million isn't that much for software development really, especially when a lot of that cost is actually the hardware. Amazon probably spends 200-400k on average per year per software developer, maybe more when considering costs of management etc.. iRacing probably has a much bigger budget? So I'm surprised these exist at all.
My sim consists of a camping table as the desk, with my G29 wheel secured with a chopping block to keep it in place. For my pedals it’s under the camping table for blocks of wood so it doesn’t slide around on my carpet under braking. My gear shifter is on a container with every Harry Potter book staked on eachother as the stand with a flat wood plank and the top with the gear shifter secured to that. It also has wood counterweight blocks.
Fantastic video! Masterly created and on a very interesting topic at the heart of F1 development. My F1 sim consists of a MX5 and while it also does not achieve the same G-forces it does replicate without any delay many of the minute and sometimes major feelings like the rear end rotating mid corner and stepping out.
Great video!! :) amazing to see this high end systems! The speed of this system is very high, but some more budget friendly setups like Qubic S25 comes quite close in terms of speed and latency. But at an fraction of the cost. Would love to see an video were you are testing the most high end systems around 100k . Or give us your thought about it. Best regards
As a professional in simracing I've tested and installed many systems. The best of all is the one I still use since 12 years, it also have such frequency filtering no latency, it is affordable you can see on my channel. In combinaison with dlp projectors it has real 5ms latency from my hands to the feel and picture I never feel that alive feeling in any other motion thing
You guys gave got to add some haptics to that simulator of yours. It makes such a difference to the immersion. Vibrations make up at least half of the immersion of a full motion rig and they are a good 80% of the useful information. A butt kicker is like $350, or a kit by slip angle thats much better as it provides 4 corner independent vibrations is like $600.
Sustained g-force is possbile. AMST Desdemona offers 3 g sustained - being a centrifuge. It probably can't achieve the necesary latencies but all I want to say is, you can simulate sustained g-forces.
That was my idea, then the car can rotate 360 ( if needed) to simulate g force. Then the car itself can be on rails itself and go up down the arm spinning, like a camera zoom steeper motors. To simulate another g force potential. While it can also increase / decrease g force as needed ( physics). While the main spinning motor can be slowed and speed up to also simulate more. The guy said you only need to get a percentage of the way there, so 2.5 g feels like 6 in sim ? 😮... I want to try it please !!
Really good video - explains all the fundamentals at play in the best motorsport simulators. 3 to 5 milliseconds latency - this is very good, but not the best available in F1 - agree totally that it is one of the critical things, but this sim is not the best for this particular metric. Am certain that they are working to improve this further. Driver reaction time is about 20 milliseconds or so, thus the latency is excellent compared to older driver in the loop simulators. Agree that the slight remaining latency is the main cause of drivers feeling sick - the world normally reacts faster in a real race car. Some really good and honest info about how the lower G-force is used to give a driver the most realistic feeling one can have. Loved it!
The limitation on testing the car on track is one of the stupidest ever. It forced teams to spend literally millions on building and running the simulators. Nice cost reductions there 😂
It was quite sensible before the cost cap, now i 100% agree with you - if you'd prefer to smash around silveratone for 3 days rather than use a sim you ahould be able to. Prior to the ban teams had whole 2nd teams for testing, extra cars, trucks, pit crews aletc. And Ferrari had (still have) their own track. Main issue now is engine allowance - i dont think many teams would want to add miles to their engines. But they could use a surrogate engine for aero testing. Bring it back!
I wonder if you could simulate gforces in a big wheel with a rotating capsule, if you turn it a bit upside down and accelerate a little over 1G you could counter act the earths g force. The question would be how hard is it to increase Gs fast, can you go from 1 to 6 in fractions of a second. Maybe with a clutch or moving the capsule outward on tge wheel, or both.
I always think that the best way a simulator could traduce g forces could be that the whole unit would be like a ball, where the pilot gets in, and inside of it , it would have like a cockpit, and with enough distance, rounded screens like the sphere in las vegas, but small size. Then, that ball-like unit would have the ability to rotate in any direction with a mechanic structure in the outside, but when youre sitting in the cockpit, with the seatbelts on, the horizon in the screens would be alingned properly in your field of view, so, when breaking, the ball viewed from the outside would rotate forward in the range of motion between 0 and 90 degrees depending on the strengh of the break, but viewed from the inside the pilot would have the screens with him so he wouldnt notice he is inclined towards the floor but he would only feel the weight of gravity as g forces on the break. And that same idea for the lateral g forces they feel when turning on a corner, or to feel speed G forces when accelerating inclining the ball backwards having the pilot facing the sky (without him noticing cause hes inside the screens sphere). Instead of having a big screen attached to the floor and even if you have a simulated cockpit that moves and has some forward or lateral inclinations it wouldnt ever be able to incline much, without it loosing realism because the screen is separate from it and it would break the feeling of a real car-view behavior. I got this idea from a f-16 or f-18 flight simulator I tried once, but it was different than this idea, it would be like 2 cockpits attached in front of the other, then you would have like an axis between them, so youwould get into the cockpit and you would only see like an inside of an airplane cockpit and like a screen where you would see like the airplane view or something like that. And then it would "take off" and you would begin to feel the G force of a real f-16 but viewed from the outside, the 2 cockpits would be spinning like a fan or a blender, and the cockpit would be on each blenders blade looking towards the axis. That way you have the g forces only in one direction, you would only be able to feel "speed" G forces. Not breaking or lateral G forces. Thats why I thought about the other Idea. I dont know much about simulators but it has always been interesting to me that Ive never seen one with the same idea I have in mind, not even in f1 simulators. Maybe thats a bad idea but it makes so much sense to me that it would be way more realistic for the pilot.
You'd still be limited to roughly 1G, much less than the forces in a real car, and you might get some unwanted effects or cues to the driver due to having to rotate forwards, backwards or sideways to apply the g forces
Not able to apply high enough forces, will move the pilot in a manner inconsistent with the vehicle, and a hugely complex video problem (to create something less real).
@@alan_davis why would it be inconsistent with the vehicle? Understanding that max g forces that can be aplied is 90 degrees rotation in any direction, would be set to the max g forces on any given moment, lets say a fast curve it would be inclined to the opposite side, then if breaking it would spin forward depending on the strength of the break pressure, If the pilot is attached to the cockpit with the seatbelts, and the whole thing can spin fast enough whith the screens either attatched to it or vr inside the helmet. It could feel more realistic than what they have (even if it cant reproduce the same g forces, it would simulate g forces to a degree) opposed to the actual setups which only move a little bit, and dont simulate any g force
@@philipk4475 Unwanted effects like what? In a real f1 g forces are way stronger, that means that when youre going through a corner you feel like if the the car is throwing you to the opposite side of the corner. The same as when breaking, you would feel that you have to hold your head from going towards the steering wheel, so, if youre in the sim I described you could have a less serious g force simulation bcause it would only be gravity, but the horizon in your field of view would still be where it should, the spinning would only be for traducing simulated g forces in a softer manner than real f1 g forces but still having the same vibe inside the cockpit instead of having everything attached to the ground and having subtle movements. I guess is not easy to imagine if you dont completely understand the idea.
@@Supersonic_racing Pressure on the skin is one thing. Feeling acceleration in the inner ear is another important thing - which can't be influenced by pressure suits.
So they derive the acceleration once more to get jolt, that is the rate of change of acceleration. Using a tilting sim would allow for simulating a weak but rather constant centripetal force, but it would produce some other unwanted forces too. It's better suited for aviation sims, where constant g-loading builds up rather slowly, so the required reaction speed is much slower, apart from the little bumps you'd feel from aerodynamic turbulence.
The rate of change of acceleration is Jerk, not jolt. And the rate of change of jerk is snap, then crackle and then pop. Don't say physicists don't have a sense of humour
@@PaulHickford Whatever order of magnitude we look at, at some point we reach a degree of amalgamation where bones behave like play-doh, planets behave like liquids and space-time behaves like a fabric of information. Isn't the universe fascinating! - some ER surgeon pondering about his last job, probably
The consumer/console games don’t do a good job of decoupling the front and rear axles, but I’d imagine from the remarks about the feel of the tear end under braking, these are beyond an order of magnitude better in that regard. So cool. Would like to know if this company is supplying the NASCAR teams as well.
I wonder if parallelization could be used to pre-compute all the predicted values that can happen on the next frame or frames... I mean, if the value of something is '5' in this frame it won't be 50 000 the next frame. It will be 6 or 8. Add 16k predictions matrix and you can pick the solution from a storage. Kind of what our brains do all the time... Computing something takes time but if we can solve all possible predictions of what is going to happen next then you only need to execute.
Interesting that platform pitch and roll aren't used as surrogates for acceleration as they are in flight simulators on Stewart platforms. One reason is because the visual is not on the platform and would have little screen space left to compensate for the tilt, particularly in roll, and another is that these types of cues produce undesirable rotation rates in the inner ear, ever so more with the required onset of the acceleration, leading to motion sickness. With flight sims, braking is the worst offender, and difficult to tune the filters to achieving a compromise. Surprised 3ms latency (and they probably meant throughput) is required, as the human animal has a mean 1/3 sec time constant - I would have thought 30ms would suffice in these high gain applications. Nice work.
Hey scott can i ask you something? is the spanish track an AI? because its sound just like your voice, if is a real person well congrats because its so good
@@stevenhaggar132 Been giving it a bit of thought lately. I reckon three wires, two diagonally out the front and one going straight back. I would mount that to a bearing mounted ring so you can turn your head and mount the ring to a helmet with velcro. I figure you want a system that you can break free off in case of a fault in the system, no responsibility taken for broken necks or whiplash... :) Edit: Scratch the whole ring and bearing idea. Just mount the wires to the top and centre of the helmet, that should allow enough freedom to turn.
@@stevenhaggar132 Thanks for the vote of confidence, I emailed Dynisma and passed on the idea about 10 minutes ago.. Not sure if we'll ever see this for consumer grade simulators but if you are spending mega bucks it wouldn't be hard to do.
I found the part where the gentleman explained the “high pass filter” interesting. They use the “jerk” (technical term) to impart the feeling of changes in acceleration and deceleration. The drivers’ imagination does the rest! Cool.
I kinda expected the bigger rigs to use some sort of 6 axis system to use gravity to simulate G-force, though i suppose when the cars can pull 6G maybe that is less useful.
"The driver is stationary when the sim thinks it's going 320-kmh, and in order to replicate a braking situation, you have to accelerate him backwards." Or...you know...tilt him (and the video screen) somewhat face down relative to gravity. This wouldn't give you 4 or 5 G of course, but at least it's capable of providing the perception of force for extended periods of time. Got a long carousel to get around? Tilt everything to the outside.
It’s definitely nit-picky but the colour reproduction from the projectors seems awful. Assuming it would not add too much latency, replicating this with either a modular panel type screen or large curved displays with the ‘optical bezel removers’ would be very cool.
These are actual simulators, everything else you guys use at home is just a video game. The iracing rfactor crowd gets an aneurysm everytime you bring this up lol
The simulator platform knows where it is at all times. It knows this because it knows where it isn't. By subtracting where it is from where it isn't, or where it isn't from where it is (whichever is Greater), it obtains a difference, or deviation. The motion control subsystem uses deviations to generate corrective commands to drive the platform from a position where it is to a position where it isn't, and arriving at a position where it wasn't, it now is. Consequently, the position where it is, is now the position that it wasn't, and it follows that the position that it was, is now the position that it isn't.
I'm a bit confused. For the inner ear breaking and pitching forward feels exactly the same. Can't you just simulate g forces by pitching/rolling? Sure you can only simulate 1g of cornering/braking acceleration, but this could be compensatet for with additional weight the driver wears. Instead of the huge screen you would obviously need vr googles or somthing comparabel to fool the driver into experincing only the g forces without a relation to their surrounding. The only downside would be the constant weight pulling driver down. But this way you could have sustained g forces throughout longer corners/braking sections. This exact same principal is used for flight simulators. Maybe I'm just overseeing something, but this would be a logical way for me simulate sustained g forces. Maybe you can help me see my flaw in logic here?
The inner ear knows nothing about the additional weight on the body. For the ear, 1 g is still the maximum sustained force by pitching - leaving 0 g felt towards the butt. If you want sustained g-forces even above 1g that the ear can sense, you need a centrifuge (see AMST Desdemona for example).
Wouldn't it be interesting if one day simulators were a part of a teenager's driving training as they went through the process of learning to drive safely before getting their permits and driver's license. In 2022, 2,883 teenagers 13-19 years old died in car crashes in the USA. And in 2020, 227,000 teenagers were injured in motor vehicle crashes. Saving one life might be worth considering the prospects.
Hi Scott, excellent video. I have a basic motion simulator from Pro Simu. It has 2 actuators in the front to simulate heave, roll, bumps ect and a traction loss actuator at the rear. It would be great to know how these 4000 euro sims stack up to the ultra pro ones you tested.
Does anyone kind of miss the days when racing was just a fly by the seat of your pants operation with the "this is Jerry he's our guy he says it will work so do it" philosophy? The technology is impressive, amazing, mindblowing and really cool but it lacks a certain something that slapping together a whole car in an old shed in a few months has.
Hello Driver61; I have a serious question also😅; did they also develop the software, like the tracks, realism, the surroundings, ? Or do they use it from existing sims like AC? And if they make it themselves; why do they nail it so perfectly, and companiies like kunos or iracing cannot..? And why don’t they sell the software part also to those companies; so we all can buy it? It has been payed for already so they make double profit.. Again, just talking about physics and surroundings, not the motion hatdware etc
My sim consists of a Logitech G923 wheel and pedals, gear lever and simple desk and gaming seat that I have to secure to the wardrobe with a piano stool and a guitar speaker for it not to move (it still does a little when I push the brake). I suppose it could fall a little short of F1 teams' expectations...
you could still perfectly learn to drive, and if serious enough, win races of drift events in real life! try it
I bet this pro sim is still more quiet than g923 :D Mine sounds like a drill
i thought i was the only one who put my amp behind my chair😭
And you have fun! That’s 🎉🎉
@@docmain999griptape under a diningchair was my setup for years, although a cockpit and better wheelbase and all that sort of stuff adds alot to your quality of life.. it doesnt make you a quicker driver
My room mate blows next to me when there's wind and spits when it's raining. He also kicks me depending on the crash. Much more real phyisics than those "top" simulators.
Peak Immersive experience
What about snow? Or rallying in wet mud?
I hope you shower him with “champagne” after a win
Play rally games with him and see how he simulates mud
@ 😂
Wait. Is this the british UA-camr/ Racing driver that holds the Lap record at brands hatch???
Correct sir.
Why would you write racing driver and lap with capital letters but British and Brands Hatch in lowercase?
@@jansz1589priorities, my Friend. also i think it Causes issues in the truly british. i kNow it buggEd me.
Still holding onto that one.... just.
how did you know that?! he never mentions it 🤣
The other team is McLaren. The names of Ferrari and McLaren show up in the SEO results in the short description of SAE International's website when they covered Dynisma as a startup in 2022. I wonder if McLaren decided later to keep their supplier info private, while the search engines still have that information and the webpage is still up.
AWESOME! Great video guys :)
When you guys collabing?!
I agree Will
People seem perplexed that theres so much trust in simulators but NASA used various analog simulators for the Apollo missions in the 60's.
I m very perplexed because ALL teams missed the ‘porpoising’ effect and coming from a mech eng with 40 years exp. in CFD / STR analysis because the repeated bouncing up and down should be simple to solve I think !
Several of the Apollo astronauts were Marine or Navy pilots so they were already trained for high stress environments but of course still needed to practice working the specific controls for the Apollo flight plan so plenty of simulators of all kinds were created, truly amazing for 1960s and 70s 😎
@@philippemichelvidori7248 it's not really just bouncing up and down mate! It's a rather complex problem that I'm not exactly competent to explain to you having only a basic understanding, but what I can tell you is that fluid dynamics get complicated really quickly and when you have to consider flow on all sides, in the venturi tunnel, around the wing, etc. and on top of that distinguish between turbulent and laminar flow... it gets utterly mad.
@@philippemichelvidori7248 The problem with porpoising is that the windtunnel models are static and can't move up or down with regards to downforce, it's specifically to prevent the model from hitting the belt they run on as it would likely explode and create massive damage to both the wind tunnel and the model. So they're stuck relying on CFD simulations which are very performance intensive, you can either run a very accurate simulation on a small section of the car or run a more complete but inaccurate simulation on the whole car, or somewhere in between but never fully both. Porpoising is also a very complex feedback loop making it especially hard to simulate accurately even without the issues i mentioned before.
@philippemichelvidori7248 porpoising IS easy to resolve, you just run the car stiffer and higher.
But if you also want to be fast, you have a particularly complex problem.
That high-pass filter trick on the acceleration is really cool. So simple!
The other team that he cannot mention is Cadillac. By the way
This was probably made before the Cadillac announcement
@unklesuga1644 thats why I am saying that. Cadillac knew for awhile they were gonna be 11th team. As soon as Micheal Andretti stepped down that was what F1 wanted.
Or Audi
@blade0667 i do actually forget about their team🤣 you're right. IF Ferrari is using that Simulator and Ferrari are supplying Cadillac's engines it makes sense
it has to be Haas
0:33 Kubica mentioned RAAAHHHH
Interesting video, but I would have liked it go way more into detail. “Simple” things like how they manage to keep latency low, make use of synchronization between visuals and movement of the platform, perhaps even sound (you didn’t mention that at all). But more importantly, you sidestepped *how* teams translate a part that only exists on the digital drawing boards to be “fitted” onto the virtual car and then influences the way the simulator behaves. To me, that is the most interesting aspect of this workflow.
Yeah, coz they’re really going to divulge their competitive advantage 🤦♂️
>“Simple” things like how they manage to keep latency low,
I agree, mostly because of how ridiculous the 3-5ms number is, not just for something of this scale, but in general for electronic input delay lol.
This Synchronisation is by far the easier part. Other F1 sims litterly run on a heavily guarded and Custom version of rfactor 2 some real gt drivers especially GT3 and 2 use assetto cors.
So getting something that is ,,faster" probably isn't the hardest part considering the price.
Also how fast motors be it servo, stepper or also electronically controlled hydronics can be controlled/adjusted has developed quite a bit in the past 10-15 years.
I think the biggest part is the price and what racing series has the money for it.
The "cheaper" ones (still probably over 300k) do mostly run on less customized software and something closer to heavily modded at home sim and have parts made by different engineers and developers from different companies.
Connecting all these pieces in hardware as in software in a less customized sim with less budget introduces the most amount of latency most likely.
This seems to be the by far biggest difference and what also makes the most difference for the divers, your average person obviously wouldn't notice these tiny delays nearly as much as a driver used to the real thing and with such reaction time.
"you sidestepped how teams translate a part that only exists on the digital drawing boards to be “fitted” onto the virtual car" I build a vehicle model in MSC ADAMS or Carsim with the new parts in it . This then gets uploaded to the simulator . The model is driven by steering wheel angle, throttle/brake position, gear selector and so on, and then outputs the vehicle state to the simulator.
better then my sim rig thats for sure
God could you imagine the grief you'd get from the wife? I have a hard enough time convincing mine to deal with my simple desk + chair setup
@@naplesbluesrt i do have a proper rig but i can't use it as i kept stubbing my toe when i got out of bed due to how close my desk is to my bed and just how big the rig is lol
@@The_Rally_Forge I covered the edges of my rig with neoprene pads lol
I bet Mazepin could find a way to make one of these spin around, or even on his best day crash and break the whole sim.
What is he doing nowadays?
@@Moonfaster fighting with the Russian army in Ukraine.
Maz once crashed a bike-hometrainer...
Funnily enough he was the only driver on the 2021 grid to not use a simulator
(Haas are broke)
you can replace this with Lance Stroll and you'll have a more accurate example
0:34 POLAND POWER! 💪💪🇵🇱🇵🇱
wruci
Best competent video about pro simulators available public.
Thank you for showing this. Has been a bit of a mystery as to how this all works and how they can be so reliant on it. Really cool to see this.
You can't 1-1 simulate the g-forces the inner ear feels in a sustained way, but you can do it to some parts of the body. Eg your skin won't know the difference between being pressed into the side or your seat in a corner vs your seat pressing into your skin in a corner. Some sim rigs use this trick to simulate the sustained Gs with movable panels in the seat. Don't think it's 100%, but it does sound like 1 more level of immersion where you could quite literally be driving by the seat of your pants ;)
You can! Just make the sim rig actually move, and to sustain the g force, it should move miles away with wheels... Ah crap I just described actually driving the racecar
@@kerimca98 lol
A g-suit could simulate the sensations of g-force and the brain would interpret that in a similar way
@@fookapi that's with compressed air, right? not sure if that would have the response time they are looking for. but yeah, there are other ways to simulate Gs without moving the head about
@@alex8373 I’m imaging a full body suit that can apply pressure to different parts of the body to simulate forces. So unlike a traditional g-suit (which I wore as a military pilot), it shouldn’t just compress the legs. Just an idea and yes, it probably wouldn’t be fast enough.
I was waiting for someone like you to try a real F1 factory team simulator!! I’m so excited to watch it
I was literally asking someone about this yesterday, and you go and make a video just for me to answer my curiosity. You are so generous! :)
bot
@danieldorn9989 bleep bloop
Do these simulators have custom software for graphics and track specs? Did they laser scan all the tracks themselves and create their own sim software or are they using 3rd party data? Super curious!
We will never know
It's all straight from Forza Motorsport.
If I’m not mistaken some of the guys from Iracing started working on a project that was based off their models on Iracing
@ that’s what I’m wondering. Like if you’re 2mill in a sim. Is that “enough money spent, let’s use iRacing”. Or is it, “only our own personal laser scans of tracks meets the standards for our super expensive sim”
Yeah I'm curious about the economics of it all. 8 million isn't that much for software development really, especially when a lot of that cost is actually the hardware. Amazon probably spends 200-400k on average per year per software developer, maybe more when considering costs of management etc.. iRacing probably has a much bigger budget? So I'm surprised these exist at all.
My sim consists of a camping table as the desk, with my G29 wheel secured with a chopping block to keep it in place. For my pedals it’s under the camping table for blocks of wood so it doesn’t slide around on my carpet under braking. My gear shifter is on a container with every Harry Potter book staked on eachother as the stand with a flat wood plank and the top with the gear shifter secured to that. It also has wood counterweight blocks.
Fantastic video! Masterly created and on a very interesting topic at the heart of F1 development. My F1 sim consists of a MX5 and while it also does not achieve the same G-forces it does replicate without any delay many of the minute and sometimes major feelings like the rear end rotating mid corner and stepping out.
This thing is absolutely wild, props to Dynisma.
You should get a motion rig for the office! They’re pricey obviously, but being able to add the motion aspect into a home sim setup is pretty amazing.
Great video!! :) amazing to see this high end systems!
The speed of this system is very high, but some more budget friendly setups like Qubic S25 comes quite close in terms of speed and latency. But at an fraction of the cost.
Would love to see an video were you are testing the most high end systems around 100k . Or give us your thought about it. Best regards
As a professional in simracing I've tested and installed many systems. The best of all is the one I still use since 12 years, it also have such frequency filtering no latency, it is affordable you can see on my channel. In combinaison with dlp projectors it has real 5ms latency from my hands to the feel and picture
I never feel that alive feeling in any other motion thing
You guys gave got to add some haptics to that simulator of yours. It makes such a difference to the immersion. Vibrations make up at least half of the immersion of a full motion rig and they are a good 80% of the useful information. A butt kicker is like $350, or a kit by slip angle thats much better as it provides 4 corner independent vibrations is like $600.
Sustained g-force is possbile. AMST Desdemona offers 3 g sustained - being a centrifuge. It probably can't achieve the necesary latencies but all I want to say is, you can simulate sustained g-forces.
14:20 if your FFB is set correctly you can feel exactly what the rear is doing and how much weight is on each tyre in Assetto Corsa.
dont screw up their 3 million project right away.. not funny..
In theory easy to simulate g-forces. But the entire thing inside a centrifuge, like used for g simulation for fighter pilots.
That was my idea, then the car can rotate 360 ( if needed) to simulate g force. Then the car itself can be on rails itself and go up down the arm spinning, like a camera zoom steeper motors. To simulate another g force potential. While it can also increase / decrease g force as needed ( physics). While the main spinning motor can be slowed and speed up to also simulate more. The guy said you only need to get a percentage of the way there, so 2.5 g feels like 6 in sim ? 😮... I want to try it please !!
0:21 prolly caddilac or audi lmao
Really good video - explains all the fundamentals at play in the best motorsport simulators. 3 to 5 milliseconds latency - this is very good, but not the best available in F1 - agree totally that it is one of the critical things, but this sim is not the best for this particular metric. Am certain that they are working to improve this further. Driver reaction time is about 20 milliseconds or so, thus the latency is excellent compared to older driver in the loop simulators. Agree that the slight remaining latency is the main cause of drivers feeling sick - the world normally reacts faster in a real race car. Some really good and honest info about how the lower G-force is used to give a driver the most realistic feeling one can have. Loved it!
This puts my moza setup to shame…. 😭 the only thing I have that they don’t is a fan on the base to simulate wind 😂.
and a seatbelt tensioner.......
I've long been curious about F1 teams' simulators, so this was really great.
Really nice video! So interesting to have a focus on such a topic, with a good mix of facts, interviews and visuals. Keep it up!
I really enjoy being educated about these subjects. Thank You!
The limitation on testing the car on track is one of the stupidest ever. It forced teams to spend literally millions on building and running the simulators. Nice cost reductions there 😂
It was quite sensible before the cost cap, now i 100% agree with you - if you'd prefer to smash around silveratone for 3 days rather than use a sim you ahould be able to.
Prior to the ban teams had whole 2nd teams for testing, extra cars, trucks, pit crews aletc. And Ferrari had (still have) their own track.
Main issue now is engine allowance - i dont think many teams would want to add miles to their engines. But they could use a surrogate engine for aero testing.
Bring it back!
I wonder if you could simulate gforces in a big wheel with a rotating capsule, if you turn it a bit upside down and accelerate a little over 1G you could counter act the earths g force. The question would be how hard is it to increase Gs fast, can you go from 1 to 6 in fractions of a second. Maybe with a clutch or moving the capsule outward on tge wheel, or both.
An active harness and G seat helps with feeling sustained forces.
Qubic belt tensioner!!
@@michaelhuyer1479 Yep, have one. 😁
@@seattime4075 next year i will get one for my V3 from motion system EU👍
Maintenance, replacement parts and the electric bill add up too….!
I always think that the best way a simulator could traduce g forces could be that the whole unit would be like a ball, where the pilot gets in, and inside of it , it would have like a cockpit, and with enough distance, rounded screens like the sphere in las vegas, but small size. Then, that ball-like unit would have the ability to rotate in any direction with a mechanic structure in the outside, but when youre sitting in the cockpit, with the seatbelts on, the horizon in the screens would be alingned properly in your field of view, so, when breaking, the ball viewed from the outside would rotate forward in the range of motion between 0 and 90 degrees depending on the strengh of the break, but viewed from the inside the pilot would have the screens with him so he wouldnt notice he is inclined towards the floor but he would only feel the weight of gravity as g forces on the break. And that same idea for the lateral g forces they feel when turning on a corner, or to feel speed G forces when accelerating inclining the ball backwards having the pilot facing the sky (without him noticing cause hes inside the screens sphere).
Instead of having a big screen attached to the floor and even if you have a simulated cockpit that moves and has some forward or lateral inclinations it wouldnt ever be able to incline much, without it loosing realism because the screen is separate from it and it would break the feeling of a real car-view behavior.
I got this idea from a f-16 or f-18 flight simulator I tried once, but it was different than this idea, it would be like 2 cockpits attached in front of the other, then you would have like an axis between them, so youwould get into the cockpit and you would only see like an inside of an airplane cockpit and like a screen where you would see like the airplane view or something like that. And then it would "take off" and you would begin to feel the G force of a real f-16 but viewed from the outside, the 2 cockpits would be spinning like a fan or a blender, and the cockpit would be on each blenders blade looking towards the axis. That way you have the g forces only in one direction, you would only be able to feel "speed" G forces. Not breaking or lateral G forces. Thats why I thought about the other Idea. I dont know much about simulators but it has always been interesting to me that Ive never seen one with the same idea I have in mind, not even in f1 simulators. Maybe thats a bad idea but it makes so much sense to me that it would be way more realistic for the pilot.
This! I agree totaly. Something similar is in fun parks where you go inside "simulator" for a rollercoaster ride or similar
You'd still be limited to roughly 1G, much less than the forces in a real car, and you might get some unwanted effects or cues to the driver due to having to rotate forwards, backwards or sideways to apply the g forces
Not able to apply high enough forces, will move the pilot in a manner inconsistent with the vehicle, and a hugely complex video problem (to create something less real).
@@alan_davis why would it be inconsistent with the vehicle? Understanding that max g forces that can be aplied is 90 degrees rotation in any direction, would be set to the max g forces on any given moment, lets say a fast curve it would be inclined to the opposite side, then if breaking it would spin forward depending on the strength of the break pressure, If the pilot is attached to the cockpit with the seatbelts, and the whole thing can spin fast enough whith the screens either attatched to it or vr inside the helmet. It could feel more realistic than what they have (even if it cant reproduce the same g forces, it would simulate g forces to a degree) opposed to the actual setups which only move a little bit, and dont simulate any g force
@@philipk4475 Unwanted effects like what? In a real f1 g forces are way stronger, that means that when youre going through a corner you feel like if the the car is throwing you to the opposite side of the corner. The same as when breaking, you would feel that you have to hold your head from going towards the steering wheel, so, if youre in the sim I described you could have a less serious g force simulation bcause it would only be gravity, but the horizon in your field of view would still be where it should, the spinning would only be for traducing simulated g forces in a softer manner than real f1 g forces but still having the same vibe inside the cockpit instead of having everything attached to the ground and having subtle movements. I guess is not easy to imagine if you dont completely understand the idea.
Porpoising must have been like the day of reckoning for simulation departments on every F1 team
9:47 Or a human centrifuge with a swiveling basket and rapid spinup/spindown could do it. :)
Pressure suits could replicate the sensation. This tech is already developped
@@Supersonic_racing Pressure on the skin is one thing. Feeling acceleration in the inner ear is another important thing - which can't be influenced by pressure suits.
You are kind of right. It was already built for aerospace. See AMST Desdemona.
bang on
So they derive the acceleration once more to get jolt, that is the rate of change of acceleration. Using a tilting sim would allow for simulating a weak but rather constant centripetal force, but it would produce some other unwanted forces too. It's better suited for aviation sims, where constant g-loading builds up rather slowly, so the required reaction speed is much slower, apart from the little bumps you'd feel from aerodynamic turbulence.
The rate of change of acceleration is Jerk, not jolt.
And the rate of change of jerk is snap, then crackle and then pop. Don't say physicists don't have a sense of humour
@@PaulHickford Whatever order of magnitude we look at, at some point we reach a degree of amalgamation where bones behave like play-doh, planets behave like liquids and space-time behaves like a fabric of information. Isn't the universe fascinating!
- some ER surgeon pondering about his last job, probably
The consumer/console games don’t do a good job of decoupling the front and rear axles, but I’d imagine from the remarks about the feel of the tear end under braking, these are beyond an order of magnitude better in that regard. So cool. Would like to know if this company is supplying the NASCAR teams as well.
I wonder if parallelization could be used to pre-compute all the predicted values that can happen on the next frame or frames... I mean, if the value of something is '5' in this frame it won't be 50 000 the next frame. It will be 6 or 8. Add 16k predictions matrix and you can pick the solution from a storage. Kind of what our brains do all the time... Computing something takes time but if we can solve all possible predictions of what is going to happen next then you only need to execute.
Congratulations, this video is awesome!!!!! But I ask: Why don't they use VRs?
there is a road car sim in the background i think that would be very interisting to!
This was a brilliant watch, your work is incredibly informative and well put together. Thank you Scott.
Interesting that platform pitch and roll aren't used as surrogates for acceleration as they are in flight simulators on Stewart platforms. One reason is because the visual is not on the platform and would have little screen space left to compensate for the tilt, particularly in roll, and another is that these types of cues produce undesirable rotation rates in the inner ear, ever so more with the required onset of the acceleration, leading to motion sickness. With flight sims, braking is the worst offender, and difficult to tune the filters to achieving a compromise. Surprised 3ms latency (and they probably meant throughput) is required, as the human animal has a mean 1/3 sec time constant - I would have thought 30ms would suffice in these high gain applications. Nice work.
0:20 Apex GP ;)?
he is not saying no so you might be on to something
No.
Hey scott can i ask you something? is the spanish track an AI? because its sound just like your voice, if is a real person well congrats because its so good
If I had 20 million a year , I’d have this next to the underground swimming pool .
I think they could add wires to the helmet to drag your head around to simulate g=force.
I like the way you think
@@stevenhaggar132 Been giving it a bit of thought lately. I reckon three wires, two diagonally out the front and one going straight back. I would mount that to a bearing mounted ring so you can turn your head and mount the ring to a helmet with velcro. I figure you want a system that you can break free off in case of a fault in the system, no responsibility taken for broken necks or whiplash... :)
Edit: Scratch the whole ring and bearing idea. Just mount the wires to the top and centre of the helmet, that should allow enough freedom to turn.
I like it
@@stevenhaggar132 Thanks for the vote of confidence, I emailed Dynisma and passed on the idea about 10 minutes ago.. Not sure if we'll ever see this for consumer grade simulators but if you are spending mega bucks it wouldn't be hard to do.
If only Nick Wirth had a simulator this good back in 2009, with his "CFD only" built F1 car
Surprised no mention of belt tension to help simulate g force, wonder if this rig has that as well
I found the part where the gentleman explained the “high pass filter” interesting. They use the “jerk” (technical term) to impart the feeling of changes in acceleration and deceleration. The drivers’ imagination does the rest! Cool.
Great interesting video...do other race series teams use these as well.. like IMSA/WEC endurance racing teams??
Yes. Peugeot just put out a video of Theo Pourchaire in their simulator the other day. Even F2/F3 teams have simulators (though not as advanced).
I kinda expected the bigger rigs to use some sort of 6 axis system to use gravity to simulate G-force, though i suppose when the cars can pull 6G maybe that is less useful.
2:24 - this looks like trackmania but professional.
When is the Dynisma sim/game coming out!
mmh, i felt bad spening 800$ on my PC.
2:27
Hamilton in Singapore 2018, "Hold my tea"
"The driver is stationary when the sim thinks it's going 320-kmh, and in order to replicate a braking situation, you have to accelerate him backwards."
Or...you know...tilt him (and the video screen) somewhat face down relative to gravity. This wouldn't give you 4 or 5 G of course, but at least it's capable of providing the perception of force for extended periods of time. Got a long carousel to get around? Tilt everything to the outside.
0:36 that is definitely Robert Kubica Haha
Is the cost of the simulator included in the team's cost cap? If a team cannot afford the best simulator, is that an unfair advantage?
Why "unfair"? If they can't afford the best race engineers, is that also "unfair"?
one of the very few video's that I actually liked.
a part from Pete Windsor, nobody can touch Scott's content quality on motorsports on yt
It’s definitely nit-picky but the colour reproduction from the projectors seems awful. Assuming it would not add too much latency, replicating this with either a modular panel type screen or large curved displays with the ‘optical bezel removers’ would be very cool.
yeah, that's what I noticed too. All that money just for the most Washed Out image you've ever seen.😂
These are actual simulators, everything else you guys use at home is just a video game. The iracing rfactor crowd gets an aneurysm everytime you bring this up lol
Awesome video!
i had all 3 logitech wheels, my favourite one was the g25 amazing wheel! sad that it only does 900° considering dritfting
all the people ive talked to drift on 900. and im.... trying to. why couldnt you?
That is some jackson storm technology right there.
Virtual Spa versus the Real one? Cool concept to test on any track.
The simulator platform knows where it is at all times. It knows this because it knows where it isn't.
By subtracting where it is from where it isn't, or where it isn't from where it is (whichever is Greater), it obtains a difference, or deviation.
The motion control subsystem uses deviations to generate corrective commands to drive the platform from a position where it is to a position where it isn't, and arriving at a position where it wasn't, it now is.
Consequently, the position where it is, is now the position that it wasn't, and it follows that the position that it was, is now the position that it isn't.
The day they invent a super fan in which they can manipulate the density of the air, it will be possible to simulate G-force in racing simulators.
Exceptional performance is usually an exponential cost.
Just let AI Drive, don't need to pay millions to real drivers 😂😂😂😂😂
Actually... They already have this... 😅 It's first year was a disaster but it's going to get fun... Would be cool to see AI vs. Human.
I'm a bit confused. For the inner ear breaking and pitching forward feels exactly the same. Can't you just simulate g forces by pitching/rolling? Sure you can only simulate 1g of cornering/braking acceleration, but this could be compensatet for with additional weight the driver wears. Instead of the huge screen you would obviously need vr googles or somthing comparabel to fool the driver into experincing only the g forces without a relation to their surrounding. The only downside would be the constant weight pulling driver down. But this way you could have sustained g forces throughout longer corners/braking sections. This exact same principal is used for flight simulators. Maybe I'm just overseeing something, but this would be a logical way for me simulate sustained g forces. Maybe you can help me see my flaw in logic here?
The inner ear knows nothing about the additional weight on the body. For the ear, 1 g is still the maximum sustained force by pitching - leaving 0 g felt towards the butt. If you want sustained g-forces even above 1g that the ear can sense, you need a centrifuge (see AMST Desdemona for example).
All that tech an 1/2 the time the new component doesn’t “perform as expected “ when it comes to race day
Wouldn't it be interesting if one day simulators were a part of a teenager's driving training as they went through the process of learning to drive safely before getting their permits and driver's license. In 2022, 2,883 teenagers 13-19 years old died in car crashes in the USA. And in 2020, 227,000 teenagers were injured in motor vehicle crashes. Saving one life might be worth considering the prospects.
I don’t know about today! But when I was a kid back in the ‘70’s they had simulators as part of Drivers Ed!
A neura link type helmet to stimulate g forces in different sensory parts of the body.
Hi Scott, excellent video. I have a basic motion simulator from Pro Simu. It has 2 actuators in the front to simulate heave, roll, bumps ect and a traction loss actuator at the rear. It would be great to know how these 4000 euro sims stack up to the ultra pro ones you tested.
I have aTraction loss and sway Motion Sim,i drive in VR.Its fantastic with drift cars around corners,super real Feeling,my Sim cost about 15000 Euro
in those bigger setups , longer travels, at the end u feel more "weight"..
Is audio fed to the driver, engine noise, vibration etc?
0:35 GIGAKUBICAAAAA \o/
Great content, appreciate it!
I drive right past this company everyday to work! Never knew what they did. Its in a tiny village in the middle of nowhere.
Somewhat makes on track testing restrictions a bit redundant.
Does anyone kind of miss the days when racing was just a fly by the seat of your pants operation with the "this is Jerry he's our guy he says it will work so do it" philosophy? The technology is impressive, amazing, mindblowing and really cool but it lacks a certain something that slapping together a whole car in an old shed in a few months has.
I wonder if "pogo-ing" story of f1 cars a while ago weren't because of problems/limitations in that simulating to manufacturing workflow
Spa looks exaclty like assetto corsa's model. Are they using a modded version of the game? Or just the track scan?
Wow. That is amazing tech. I need one
Hello Driver61; I have a serious question also😅; did they also develop the software, like the tracks, realism, the surroundings, ? Or do they use it from existing sims like AC? And if they make it themselves; why do they nail it so perfectly, and companiies like kunos or iracing cannot..? And why don’t they sell the software part also to those companies; so we all can buy it? It has been payed for already so they make double profit.. Again, just talking about physics and surroundings, not the motion hatdware etc
So did the yellow platform on the thumbnail make a difference?
I met a guy who made some of the first racing simulators for McLaren in 2007
If if the simulation becomes so realistic that they start holding the races using it and no one noticed
Am left wondering who the £2M one is aimed at as the F1 teams would want the £8M one.
100 quid says he sat down and said "Fire up brands hatch thanks"
I swear he explained that turning the wheel and braking is detected, kinda would expect that for a 2.5 million dollar sim
would be interesting to know about their software they use
Which software are using? Assets Corsa or something native?
I think I know what I want for Christmas...
Does F1 or the FIA limit sim driving time? Or can teams and drivers just use the sim as much as they feel they need?