Cars Honda NSX 3.6.5

[Kyuubeey]

show me the law, access to the

files is educational use and personal use only distrobution is illegal XP know the law

Well, if he's releasing a pack, he's distributing the files, so...

 

[DanWeidlich]

I've got my audio set up exactly as recommended on page 14, the car idles so quietly it's virtually inaudible until ~3,000 rpm. The entire low end of the exhaust note is missing, even with the volume cranked so as to blow every other sound in the car out of the water.

If I knew fmod better than I do, I'd offer to help narrow it down. Here's my suggestion--at RPMs below 5,000, it needs more volume, and more bass. The gain just isn't there until halfway up the tachometer. Beyond 5,000, it's fine as is.

My next door neighbor drove an NSX for years. I've heard it idle both inside and out, and this ain't it.

 

[Kyuubeey]

I've got my audio set up exactly as recommended on page 14, the car idles so quietly it's virtually inaudible until ~3,000 rpm. The entire low end of the exhaust note is missing, even with the volume cranked so as to blow every other sound in the car out of the water.

If I knew fmod better than I do, I'd offer to help narrow it down. Here's my suggestion--at RPMs below 5,000, it needs more volume, and more bass. The gain just isn't there until halfway up the tachometer. Beyond 5,000, it's fine as is.

My next door neighbor drove an NSX for years. I've heard it idle both inside and out, and this ain't it.

From what I was told, it was tested by people with a surround setup with bass. I can't remember if Zygrene himself did or just some users.

Perhaps you're missing out on some bass via headphones? That's how you feel/hear most roadcar idles IRL.

 

[gecco]

At idle and low RPMs, a cultivated petrol engine is almost inaudible. At least, that's my real life impression as well. I find the sound of the NSX to be quite believable therefore. After all, back in its day, the NSXs outstanding feature was to be the first really refined supercar, that was also pleasant to drive to the shops, unlike the Ferraris of that era.

Not otherwise judging the sound, I'm just not an expert at all, just referring to the quiet idle, I think its about right.

 

[Kyuubeey]

At idle and low RPMs, a cultivated petrol engine is almost inaudible. At least, that's my real life impression as well. I find the sound of the NSX to be quite believable therefore. After all, back in its day, the NSXs outstanding feature was to be the first really refined supercar, that was also pleasant to drive to the shops, unlike the Ferraris of that era.

Not otherwise judging the sound, I'm just not an expert at all, just referring to the quiet idle, I think its about right.

IIRC the actual sound and volume is in fact a bit stronger so it can be heard from headphones (Was that right, Modek?) but I think you just needed some good surround bass to feel it closer to IRL.

EDIT: I too had said how quiet the idle is and asked if it's right, and this is about what I was told IIRC. I use headphones.

 

[aphidgod]

My PC is primarily a work computer, and my work is in the music and audio field. (I design and manufacture analog synthesizer modules.) So my machine is part of my studio, and I route its sound through a properly set-up studio monitor rig - without a sub. Still, I don't have any issues with the sound at idle. Plenty there.

 

[DanWeidlich]

1. I don't use headphones, I've got a speaker system with a subwoofer. It's just this car, every other car I've got whether KS, or mod, or KS w/ Fonsecker has an audible idle. So it's definitely the way the sound was setup on this one.
2. Having ridden in one, I'll put it this way, it's a comfortably quiet rumble, but it's definitely a rumble.
3. That's odd that you can hear it on yours, and I can't on mine. I've never heard of a mod where a studio setup with headphones is necessary just to hear the car at idle lol. I even tried turning my sub to the max, making my desk shake when other cars idle, but almost no exhaust note for this one whatsoever.

I know how much work you guys have put into this, and I don't mean to be nitpicking. I still think it needs a bit of tinkering.

 

[alekabul]

Guys, am I the only one noticing how weak the springs are? In data for Acura NSX there is value for front of 15.353 N/mm instead of 29.75 N/mm which is the lowest stock for the car IRL. At the rear data is set at 25.694 N/mm instead of 38.5 N/mm. This results in too floaty behavior which is very unrealistic for a legendary sports car - bottoming out at every bump and even if steering harder. For me while playing offline is easy to edit and retune the data, but please do an update of the mod for everybody to use. Thank you for the work you`ve done so far!

 

[Kyuubeey]

Hey,

I've already went over the suspension with approximate motion ratios.

It's a bit foolish to claim the springrate is wrong when they even tell you to input the wheelrate in the hints.

Spring rate x (Lever Ratio x Mechanical Advantage) = Wheel rate

Or,

Spring rate x VR^2.

For example 30000 x 0.800^2 = 19200

EDIT: Also I corrected the bumpstops already as well, so it will have more proper cornering roll and bump bottoming behavior. I agree the current public version bottoms easily, rolls too much etc.

 

[aphidgod]

It's very important to be right before attempting to correct others.

 

[alekabul]

Hey,

I've already went over the suspension with approximate motion ratios.

It's a bit foolish to claim the springrate is wrong when they even tell you to input the wheelrate in the hints.

Spring rate x (Lever Ratio x Mechanical Advantage) = Wheel rate

Or,

Spring rate x VR^2.

For example 30000 x 0.800^2 = 19200

EDIT: Also I corrected the bumpstops already as well, so it will have more proper cornering roll and bump bottoming behavior. I agree the current public version bottoms easily, rolls too much etc.

Even using your formula it`s considerably weaker than the real suspension. About the hints from Kunos - one example is "car polar inertia. Calculated from the car dimensions. Just enter the generic width,height" and they never enter the real dimensions, faking the inertia to be smaller and the model - faster. So I tend to use what works best and not what looks good on paper because I like to feel the real car in my hands and don`t care for laptimes. In this case I only opened the data because of the weak springs, corrected them with the real values from the pic above and adjusted rods accordingly. Cars ride much better without being stiff as a scateboard which already tells me if I`m right or not. On the subject of stuntdriving a NSX without a rollcage take a look at this video - I don`t see a floating and diving "loosers Cadillac" in there, do you?
Oh, and if laptimes are important below is a cap of the lapboard at Ahvenisto - only "NSX-R" is not edited for spring rates. Will leave it for diversity I guess And what`s the deal with the different ai.ini for the non-R cars? Different values and even "HEADER"? Adds to the diversity too.

 

[alekabul]

There is a strange problem with the sound - all of the cars sound the same both in session and replay, onboard and outside. Sound is of the car the player drives. Happens for the first time and only with this mod. If I change the sound of the car I`m driving, all other 3 sound the same - Acura NSX-R.

 

[Kyuubeey]

@alekabul

If you were knowledgeable and knew what you were talking about, you would know the NSX is designed to impact onto fairly soft and progressive bumpstops during cornering, in order to control body-roll, suspension compression and handling. I have implemented those. I doubt they are exactly correct as I didn't go and stroke and rate test the unit myself, but based on various criteria, it's far better than just stiffening the springs blindly.

Most JDM cars of this era ended up something like 1.3hz at the wheel and used bumpstops to control attitude and handling. S13, R32, FD3S, JZA80, NSX, NA5 and others which I am yet to confirm. Then there are exceptions like the R33 V-spec and NSX-R which directly put in harder springs adding up to the high 1hz or low 2hz range.

Please, as an advice to everyone in general, there can be so much about a subject that you can't know just looking at it. I really don't want to spend all day here correcting every single thing you post and have you just argue against it. We're handling it, have already handled it and just stay tuned for the update. We've gone over the car so many times already and I did more than just update the suspension as well.

 

[aphidgod]

I tend to use what works best and not what looks good on paper .

Well, this is all we need to know about that, then. I really don't think Assetto Corsa is for you.

 

[Kyuubeey]

Well, this is all we need to know about that, then. I really don't think Assetto Corsa is for you.

Translation: Human perception is extremely subjective and you can reach multiple extremely wrong end results that all feel right, thus a theoretic approach should be taken first and narrow it down. Hell, just look at some cars I've made and their earlier versions.

 

[alekabul]

@alekabul

If you were knowledgeable and knew what you were talking about, you would know the NSX is designed to impact onto fairly soft and progressive bumpstops during cornering, in order to control body-roll, suspension compression and handling. I have implemented those. I doubt they are exactly correct as I didn't go and stroke and rate test the unit myself, but based on various criteria, it's far better than just stiffening the springs blindly.

Most JDM cars of this era ended up something like 1.3hz at the wheel and used bumpstops to control attitude and handling. S13, R32, FD3S, JZA80, NSX, NA5 and others which I am yet to confirm. Then there are exceptions like the R33 V-spec and NSX-R which directly put in harder springs adding up to the high 1hz or low 2hz range.

Please, as an advice to everyone in general, there can be so much about a subject that you can't know just looking at it. I really don't want to spend all day here correcting every single thing you post and have you just argue against it. We're handling it, have already handled it and just stay tuned for the update. We've gone over the car so many times already and I did more than just update the suspension as well.

Maybe trying to learn is my thing and being knowledgeable is yours, but for NSX and suspension there is a lot of sources for information - here for example. I know what you mean about it`s bumpstops and appreciate the effort, I also know the difference between the low and high natural frequency and the effects of bounce and pitch associated with the change of the spring rate without change in dampers... Anyway, I`m only trying to help the development with some real life suspension data, not to waste anybody`s time here. Peace!

 

[mclarenf1papa]

@ Arch don’t use VR and MR interchangeably. They’re the inverse of each other.

The formula for wheelrate is as-follows:

WR = SpringRate * Average_motion_ratio * Instantaneous_motion_ratio

Instantaneous MR is otherwise known as mechanical advantage. Avg MR is *not* otherwise known as lever ratio (which doesn’t actually make sense in this context).

It’s simplified/approximated to WR = SpringRate * MR^2 in typical documentation.

Which, in the case of the NSX, makes the mod car implemented with a very reasonable ~0.7 front motion ratio. Inputting the spring rates directly, as you’ve done, @alekabul, is incorrect.

 

[Kyuubeey]

@ Arch don’t use VR and MR interchangeably. They’re the inverse of each other.

The formula for wheelrate is as-follows:

WR = SpringRate * Average_motion_ratio * Instantaneous_motion_ratio

Instantaneous MR is otherwise known as mechanical advantage. Avg MR is *not* otherwise known as lever ratio (which doesn’t actually make sense in this context).

It’s simplified/approximated to WR = SpringRate * MR^2 in typical documentation.

Which, in the case of the NSX, makes the mod car implemented with a very reasonable ~0.7 front motion ratio. Inputting the spring rates directly, as you’ve done, @alekabul, is incorrect.

Thanks, I'll write it down. My terminology is pretty bad. I didn't remember the instant vs avg MR terminology so I used "lever ratio" because I remember it being in some publications. It's incorrect in this context?

I didn't mean to imply VR and MR are the same thing though. MR is >1 usually, VR <1 usually. So if you see an "MR" of 1.5 or something, no, you probably don't need to 1.5^2.



I ended up with 0.715 and 0.815 VR IIRC. Rounding out what David ended up with.

 

[Stereo]

Average is not really in the publications cause they don't convert everything to wheel rates, it's used to get the actual compression of the spring due to the integral of motion ratios from zero spring pressure to current position. The more sensible thing to do if you're fully simulating the suspension is just do the entire spring kinetics so you know its actual compression directly, then only the instant motion ratio matters.

 

[alekabul]

From the work you`ve done before and the credit that you receive from many modders I am sure that theory is no problem for you, but did ever crossed your mind that Kunos in all their wisdom implemented the correct formula behind the scene and left for us in *.ini files the easy possibility to enter the correct real life data? In that case the formula if applied twice will give the resulting weak suspension. Out of curiosity only
Kyuubeey how did you arrive at the exact number for the motion ratio for NSX smaller than 1? And what`s wrong with this formula apart from the practical difficulties of measuring damper angle from vertical and calculating installation ratio separately for front and rear wheels?