An Interview with Reiza Studios’ Renato Simioni - Big hint at new content!

This kind of custom FFB tweaking stuff really has got to go IMO. There is no need for that, you can just do a physically accurate FFB. Yet gMotor sims insist on tons of sliders and settings that the end user should have no access to, and should not exist to begin with. Would probably standardize the feel and impressions of all the content very much to just physically solve those.

Neils on Wheels said:

Is there a way to get the FFB to load up in the corners? I can't seem to find that in any of the custom or default files?

Click to expand...

Hmmm... well to me (low end wheel) it feels a little too on/off switch like. Definite feeling of weight in a corner (and also loss of grip when over doing it), but i wouldnt say i feel a significant difference in the weight changing from light weight to heavy....just no weight versus weight (hence the "switch" comment). Yet you read of other people talking about feeling the rubber flex in the tyres etc...i assume that might be high end wheels only (or people talking nonsense !!)

Kyuubeey said:

This kind of custom FFB tweaking stuff really has got to go IMO. There is no need for that, you can just do a physically accurate FFB. Yet gMotor sims insist on tons of sliders and settings that the end user should have no access to, and should not exist to begin with. Would probably standardize the feel and impressions of all the content very much to just physically solve those.

Click to expand...

Not everybody wants that though. For example cars that have power assisted steering, I wouldn't want that accurately felt in my sim wheel. Why? because I'm missing other forces that would be felt in a car & for me ffb is important for immersion. Even those that claim they want their wheel to mirror exactly how it would feel in the car in reality have it setup so they are getting way more feedback.

Neils on Wheels said:

Not everybody wants that though. For example cars that have power assisted steering, I wouldn't want that accurately felt in my sim wheel. Why? because I'm missing other forces that would be felt in a car & for me ffb is important for immersion. Even those that claim they want their wheel to mirror exactly how it would feel in the car in reality have it setup so they are getting way more feedback.

Click to expand...

If you can, you should try a sim with PS. Personally for me the lack of significant bumps increased confidence tremendously, and the lack of SAT dropoff allows much more accurate driving. SAT dropoff doesn't provide any useful information that can be correlated to tire grip, so it's basically a distraction IMO.

Last part's not true. The main issue with PS feel in sims atm is that generally it's difficult to filter it little enough to let the big bumps through like in racecar PS, but to do it without any additional issues resulting from it, oscillation for example. There is too little feedback, not too much.

Last part is complete speculation on your part, and it's true for some casual consumers, but not true for anyone serious.

Kyuubeey said:

This kind of custom FFB tweaking stuff really has got to go IMO. There is no need for that, you can just do a physically accurate FFB. Yet gMotor sims insist on tons of sliders and settings that the end user should have no access to, and should not exist to begin with. Would probably standardize the feel and impressions of all the content very much to just physically solve those.

Click to expand...

Reality with FFB is that in a real car most feedback comes from you ass, and not the steeringwheel
You need to compensate somewhere to get a idea what the car is doing.
Not every steeringwheel is the same.
A Logitech of 1,5 NM or Servodrive of 35NM, there is quite some difference.

Richard Hessels said:

Reality with FFB is that in a real car most feedback comes from you ass, and not the steeringwheel
You need to compensate somewhere to get a idea what the car is doing.
Not every steeringwheel is the same.
A Logitech of 1,5 NM or Servodrive of 35NM, there is quite some difference.

Click to expand...

Firsthand and 2ndhand experience with driver training and telemetry correlation would make me inclined to disagree. It's not true for anything except braking, there the main sensory method is indeed feeling the LonG. For lateral and yaw? Not really, it's very much so visual and linear slip region based from what I can tell.

Richard Hessels said:

Reality with FFB is that in a real car most feedback comes from you ass, and not the steeringwheel
You need to compensate somewhere to get a idea what the car is doing.
Not every steeringwheel is the same.
A Logitech of 1,5 NM or Servodrive of 35NM, there is quite some difference.

Click to expand...

Fully agreed. The one-size-must-fit-all approach of ACC is a total failure. It may feel good on some high-end DD wheels, but is a total dud for lower-end wheel users.

wolftree said:

Fully agreed. The one-size-must-fit-all approach of ACC is a total failure. It may feel good on some high-end DD wheels, but is a total dud for lower-end wheel users.

Click to expand...

That's why I don't talk about ACC, but I talk about AC + CSP which does all of the FFB stuff correctly. I'm not sure if ACC is doing it right. They are not the same product. I'm not talking about them interchangeably.

To begin with, ACC is not a "one-size-fits-all" approach; they have different gyro stuff for DD wheel users. CSP implemented a physically accurate gyro that works very nicely on just about any wheel. It's not the same thing and I wouldn't interchange them.

It'd maybe be a good idea to try a relatively accurate manual rack car in AC1 that is also using all the relevant Cphys features. The only car I know of right now that fills that criteria and is available to the public is the NSX by Some1, which I happened to make physics for. At the risk of appearing as a self-promoting AC shill, do try it out, maybe you will understand what I mean.

Either way my opinion is that "gamey" FFB effects that are not completely physically based is a step backwards, and it's a very bad idea to let the end-user control your physically based effects individually. It's almost always catastrophic.

Kyuubeey said:

Either way my opinion is that "gamey" FFB effects that are not completely physically based is a step backwards, and it's a very bad idea to let the end-user control your physically based effects individually. It's almost always catastrophic.

Click to expand...

They are not "gamey" FFB effects, they are just enhancing the real physical forces with different weights (just like in your beloved AC) because low-end wheels cannot bring out those small forces naturally without some magnification. And what you call "catastrophic" is actually a life-saver to anyone not owning a DD wheel because it actually lets us experience useful information instead of the usual combination of road-noise rattling and vague centering force that is the FFB signal in ACC and default AC.

wolftree said:

They are not "gamey" FFB effects, they are just enhancing the real physical forces with different weights (just like in your beloved AC) because low-end wheels cannot bring out those small forces naturally without some magnification. And what you call "catastrophic" is actually a life-saver to anyone not owning a DD wheel because it actually lets us experience useful information instead of the usual combination of road-noise rattling and vague centering force that is the FFB signal in ACC and default AC.

Click to expand...

It is "gamey" because it messes with the steering torque curve and produces sensations that don't line up to what the kinematics suggest. I'm not sure what is more useful information in a simulator than having steering forces line up with the kinematics. I'm not even on a good wheel to begin with and I never felt a need to magnify anything.

You really do not want to start messing with that and adjusting the ratio of the inputs, it easily leads to a situation where incongruent sensations are produced through the FFB. I don't really care how useful they are or how much of an advantage I can get out of them, I would just want it to line up with what it should be in terms of curve shape and leave only adjusting the global gain down to the user. Perhaps you care more about FFB as a "driving aid" so to speak.

Unfortunately real manual rack steering sensations are characterized by a centering force and vibrations. I'm not sure what exactly else you expect. There is some SAT dropoff as well, but that is a (realistic) misleading effect that you definitely don't want to blow out of proportion. Vanilla AC has quite strong SAT dropoff and it's only for the detriment of it. There is no benefit to allowing the end-user to adjust things like that because they will more likely than not end up at a bad value.

I desire the best for Ams2 but i see that if there is not a strong online the effort is futile.
And Reiza never has had a very populated multiplayer.
If they dont solve that they never will be relevant versus the other titles.
Yep i know that i am not gonna make happy the people of this forum.
.....And this last sim still falls behind ams1 in terms of ffb.
Just an opinion, dont get so crazy.

Kyuubeey said:

Firsthand and 2ndhand experience with driver training and telemetry correlation would make me inclined to disagree. It's not true for anything except braking, there the main sensory method is indeed feeling the LonG. For lateral and yaw? Not really, it's very much so visual and linear slip region based from what I can tell.

Click to expand...

But some FFB feels more immersive, to me that is a more realistic experience. How is it wrong to prefer that over a literal representation of wheel behaviour, its personal preference? You are just projecting that your own preferences are the correct & only way.

Kyuubeey said:

It is "gamey" because it messes with the steering torque curve and produces sensations that don't line up to what the kinematics suggest. I'm not sure what is more useful information in a simulator than having steering forces line up with the kinematics. I'm not even on a good wheel to begin with and I never felt a need to magnify anything.

You really do not want to start messing with that and adjusting the ratio of the inputs, it easily leads to a situation where incongruent sensations are produced through the FFB. I don't really care how useful they are or how much of an advantage I can get out of them, I would just want it to line up with what it should be in terms of curve shape and leave only adjusting the global gain down to the user. Perhaps you care more about FFB as a "driving aid" so to speak.

Unfortunately real manual rack steering sensations are characterized by a centering force and vibrations. I'm not sure what exactly else you expect. There is some SAT dropoff as well, but that is a (realistic) misleading effect that you definitely don't want to blow out of proportion. Vanilla AC has quite strong SAT dropoff and it's only for the detriment of it. There is no benefit to allowing the end-user to adjust things like that because they will more likely than not end up at a bad value.

Click to expand...

FWIW, even F1 drivers have what amounts to steering wheel feedback preferences. The latest I saw was Robert Kubica remarking that this year's Alfa Romeo doesn't give him the feedback he needs. But Kimi and Alonso have also needed certain adjustments to get a subjectively better feel in recent times.

Anyway, my orthogonal argument is that my old trusty G27 is known to have a not-entirely-linear response to a linear torque signal, meaning that I typically benefit from a little signal compression as the motors will otherwise not react to weak torque signals around TDC.

That said, I agree that the SAT signal can feel counter-intuitive and should mostly be left alone. But for people who want to get better nuances from Fy (side forces) on low-end wheels like my G27, it may be beneficial to boost the Fy slightly (via multiplier slider) for helping to gauge steady-state cornering forces without having to resort to creating a physically longer lever (trail) by increasing the caster angle (with all the kinematics changes that entails).

This is -- as I understand it at least -- essentially what a power steering system does, except it actually _scales down_ forces (including Fy) so the slope of the torque response curve is more flat in order to avoid overwhelming the driver with torque under simultaneous heavy braking and cornering conditions.

I do tend to agree with your pure approach in terms of making sure that there is good correlation between actual kinematics and the signal being delivered to the driver's hands. Arguably, this (along with low latency) is _the_ most important thing in a driver-in-the-loop-sim used for training for the real thing (simulating surface asperities and vibrations via buttkickers is arguably also a nice-to-have tool in that particular toolbox).

Great interview! really enjoyed watching it. Some really interesting points and discussions, he seems like a top bloke!

If he reads this, congrats on making a great sim and i love the continued support and development for it. Its exciting to think how good it will be in the coming years.

My father has, for years, refused to ever modify EQ settings on his sound system. His rationale? "They intended it this way, why should I play music producer?" My counter is that the producer, in his sound booth, doesn't have the same hardware you have at home. He doesn't have the same speakers, the same room size, the same echo, the same amp, so EQ is a way to compensate for your setup's possible quirks and make it sound more like the producer intended.

I think the FFB situation is similar. We're driving on a variety of wheel bases, on a variety of steering wheels, on a variety of pedal sets, on a variety of seating positions, and with rare exceptions, a fully stationary body with no G-force being applied to it, and with 0% FFB being delivered to the parts of the body that aren't your hands. In real life, you get FB from your pedals, from the back of your thighs, from your back.

My point is, FFB is necessarily extremely abstracted, it's very detached from the actual experience; giving only the forces the driver feels through the steering column IRL would be hard enough but I suppose possible, but once you introduce information about other parts of the car, it becomes a question of dosage and creative license.

Amazing how some people get old before their time, they only know how to complain. The fact is that AMS 2 is at a great level of development and with a lot of quality content to come, long live Reiza !!

I have mixed feelings about AMS2.

It does the weather thing absolutely brilliantly - really genre leading, it looks pretty good too. Although it is a bit cartoony looking at times (Imo of course - and it may well get tweaked in time).

The AI needs work, they just barge into me like I'm not even there still, although they have got better ill admit.

The stock FFB implementation is complete crap - there is no other word for it - I found a fantastic custom FFB file which I've tweaked a little, and now it feels almost as good as anything else I drive (ACC, R3E, rF2).

However, there is something wrong with the cars. I don't understand the engineering enough to say exactly what it is, but, its as if the front isn't quite connected to the road.

You've only got to drive the Porsche RSR around and watch the front end bobble up and down like a bloody kangaroo to see what i mean.

There's just something that doesn't inspire confidence in the car.

I've never played Project Cars 1/2/3 so have no biases there, or nothing to compare it against other than AC, rF2, ACC and R3E.

I want AMS2 to do well, I bought the season pass, so I'll check in on it for the next year, but it doesn't get played all that much at the moment.

ermo said:

FWIW, even F1 drivers have what amounts to steering wheel feedback preferences. The latest I saw was Robert Kubica remarking that this year's Alfa Romeo doesn't give him the feedback he needs. But Kimi and Alonso have also needed certain adjustments to get a subjectively better feel in recent times.

Anyway, my orthogonal argument is that my old trusty G27 is known to have a not-entirely-linear response to a linear torque signal, meaning that I typically benefit from a little signal compression as the motors will otherwise not react to weak torque signals around TDC.

That said, I agree that the SAT signal can feel counter-intuitive and should mostly be left alone. But for people who want to get better nuances from Fy (side forces) on low-end wheels like my G27, it may be beneficial to boost the Fy slightly (via multiplier slider) for helping to gauge steady-state cornering forces without having to resort to creating a physically longer lever (trail) by increasing the caster angle (with all the kinematics changes that entails).

This is -- as I understand it at least -- essentially what a power steering system does, except it actually _scales down_ forces (including Fy) so the slope of the torque response curve is more flat in order to avoid overwhelming the driver with torque under simultaneous heavy braking and cornering conditions.

I do tend to agree with your pure approach in terms of making sure that there is good correlation between actual kinematics and the signal being delivered to the driver's hands. Arguably, this (along with low latency) is _the_ most important thing in a driver-in-the-loop-sim used for training for the real thing (simulating surface asperities and vibrations via buttkickers is arguably also a nice-to-have tool in that particular toolbox).

Click to expand...

A proper range compression postprocessing can be okay although I haven't ever really messed with one. Unnecessary for anything but low-end wheels but it might be a (proper) way of doing what wolftree mentioned. Scaling the different output effects is NOT the way IMO. In the end you do need some kind of gain effects due to differing hardware and user needs. Stereo added one to AC1 recently via CSP that replaces a car model line which was IIRC just a gamma on the load portion of the FFB. It was terrible, but this is supposedly good.

SAT should be left alone (or up to the car model maker) as long as it's reasonable. FWIW it's not very reasonable in vanilla AC so an argument can be made for user-adjustable SAT; or just have the default(s) per-car/product be reasonable. Tires do have all varying SAT curves IRL but there is a range where they are reasonable, and I don't believe in letting the end-user choose which end of the range the tire on the car belongs to. The responsibility of that decision should belong to someone with more insight or data.

In a way my argument boils down to "an endlessly adjustable car is always in need of adjustment". There is no way whatsoever the end-user consistently knows what is a good FFB setup that will produce an enjoyable simulation. If there was, all of the DD wheel users would just set correct friction, reasonable damping, precisely accurate gain per car and leave it alone. Yet they just endlessly tweak for years on-end searching for some imaginary ideal of FFB.

I know my quite strict idea can feel old-fashioned or misguided to some users, but I've been dealing with this stuff for about a decade. You don't come to any other conclusions once you learn enough and have heard thousands of people's worries and claims about FFB.

Powersteering systems don't scale any individual forces though, that's the thing. It's a boost assist that has input/output curve to it and some filtering. Basically they clamp the torque although there are nuances and indeed some people prefer one feel over the other. Racecar PS systems filter a bit less and you can feel some bumps and stuff through, while roadcar PS might be 100% unreactive. Most casual sim users would hate the feel but it immediately connected with me so it is what it is.

couri said:

Amazing how some people get old before their time, they only know how to complain. The fact is that AMS 2 is at a great level of development and with a lot of quality content to come, long live Reiza !!

Click to expand...

Leave your thank you for the complainers, even with them most games have a unbelievable number of grotesque issues. The fanboy team has nothing to do with all those improvements. Their pink glasses and the brown noses only delay the development.

A man with a Harmonica said:

Leave your thank you for the complainers, even with them most games have a unbelievable number of grotesque issues. The fanboy team has nothing to do with all those improvements. Their pink glasses and the brown noses only delay the development.

Click to expand...

So you are saying, that games are just progressing, because people keep complaining. This is implying, that game devs have no clue, what they're doing. I would reflect over that attitude for a moment...
Some "grotesque" issues aren't that easy to fix.