Noise source... I suppose you could say that. The data that feeds it comes from a full engine simulation I wrote probably a year ago. (I.e., bore/stroke/compression ratio, cam timing/lift, intake/exhaust dimensions and configuration, head flow rates, etc., etc..) That obviously doesn't run anywhere near real time of course, but it creates a data file that powers the audio stuff that easily runs in real time now.
Nope, it's a single exhaust pipe. I.e., pure exhaust sound with no mechanical bits in there at all. There's a little bit of a high frequency squeak in that one sample (I should have used a better one). What you're hearing there are some harmonics that come into play due to sonic velocity, the pipe length, and so on, at certain rpms. I.e., you can get standing/travelling waves at certain frequencies (rpms).
Thanks. Yes, I was going for a racing V8 like Nascar in that particular sample.
For the most part, yes, that's indeed right. That should be the predominate thing you hear, especially at low speeds. At higher speeds you can get some resonances and things on top of it, but they're relatively minor probably for the most part.
Right, exactly.
The 7000HP nitro cars though are sort of a different breed as they appear to continue burning fuel in the exhaust system even after the exhaust valves have closed. I'm not 100% sure that that really happens though, so don't quote me on it :-p But if they are indeed doing so, that certainly changes the exhaust pressures too.
Scawen, if you want to try some engine vibration to see how that sounds it's pretty simple. If the wave form you are using is a function of crank angle, then you could have a 1D rotational model (torque/angular momentum/velocity only) alter the crank angle that you're looking up. I.e., have an offset to the real crank angle that itself varies as the engine rocks back and forth a couple of degrees in its mounts and use that to reference the sound instead. I tried this awhile back and it was fairly convincing. It's realistic too because that's what the real engine is doing. I.e., the crank speed is essentially changing from one instant to the next as the block vibrates. That might spruce up things just a tiny bit and is an easy addition.
Yes, I noticed this and like it too. The variations in sound with throttle changes are much nicer than before. Before (to me anyway) it seemed nonexistent pretty much. Now it's quite nice.
The whup whup sound are pressure variations in the exhaust. At the simplest level it's each cylinder "bang." At a more complex level it's the cylinder bang in combination with the dynamics caused by other cylinders also operating into the same pipe. I.e., the wave form can change with engine speed quite a bit in reality.
What causes the particular whup whup? Everything about the engine. Primarily what you hear is exhaust, so it's a combination of cylinder pressure at exhaust valve opening (along with any waves that might be travelling too/from the cylinder at that point), valve timing, exhaust dimensions and how they're linked together. I suspect the presence of shock waves is what gives really nasty engines like alcohol dragsters and so on the really radical crackling type of sound versus the more tame note you get from other engines.
Not sure if that answers your question properly...
As far as 2 simultaneous combustions are concerned, well, I'm not aware of any car engine that does this. Every cylinder in a Harley Davidson motorcycle engine goes at the same time (which is why they pop so loud and sound like they're running so slow), but I've never heard of this being done in a car.
What I mean is if you play a sine wave on top of it you can match it as though you were humming the same note as the engine. The exhaust on a real engine will "speak" at how ever many cycles per second (Hz, the frequency) the exhaust pulses dictate. A 4 stroke 6 cylinder engine has 6 exhaust pulses every 2 engine revolutions, or 3 per second.
The main pitch you hear should be:
Frequency = EngineRPM / 60 * NumberOfCylinders / n
n = 2 for a 4 stroke engine, or 1 for a 2 stroke.
I.e., taking the FZR 4 stroke for example with 6 cylinders at 7000 rpm:
Frequency = 7000 / 60 * 6 / 2
Frequency = 350 Hz
If I match a recording of this to a sine wave I get about 115Hz, about 1/3 of the right value.
Granted, if you have more than 1 cylinder firing at the same time, then that changes things :P
Just gave the new stuff a go here. The new sounds are quite welcome and an improvement imo. Good work I also like the new track selection screen and the auto updater.
One thing to note here, and I'm not sure if the guys are aware of this. The base frequency of the engine should match the number of exhaust pulses per second. When running the FZR for instance at 7000 rpm, it comes in at about 115Hz, which corresponds to less than 2900 rpm. I think it's always been this way in LFS, but maybe it's necessary if the quality at a much higher frequency (pitch) is too plain.
I like the low rpm sound more now. Got more of a whump whump whump to it.
I normally wouldn't bother responding to you, KidCodea, and have no plans on taking your suggestions or opinions there seriously. You've made a reputation for yourself as a downright nasty fella. I remember an encounter with you in the demo days pre-S1. Let's not go there though so the thread can stay on topic: About LFS.
Deggis wrote in this thread about my system and how it sounded quite synthetic. I was merely pointing out that it has advanced since those earlier demos. Clearly he was interested so I thought it was logical to pop up the latest. I surely was not trying to show up LFS or anything like that.
There is a lot of talk in the LFS forums about sound. My point was that it is indeed possible to get a raspy, obnoxious type of uncorked 700HP+ V8 sound in real time synthesis. Obviously that sound is not right for a mellower engine running through mufflers
Regarding my behavior in the LFS forums: If the mods or developers would like me to tone it down or stop doing this or that, all they need to do is ask via a PM and I'll gladly comply out of professional courtesy. There seem to be quite a few people around here interested though, so I'll continue posting unless asked otherwise.
Some people will like it, some won't. I can live with that
Not sure what you mean there by "what u never did." You seem to be aware of plans I have that I am not. That's remarkably insightful
Oh, it's been longer than that, come on now
I'm more flattered by your need to attack me than anything else. Congrats, you're the first person to do so here that I can recall I might be offended if it were coming from anyone else.
Now let's avoid a pissing match here and let folks discuss what's important instead, shall we?
Lots to do yet on it. It's fed by only a single data file from 1 engine rpm and throttle setting at the moment. Should get much better once the rest of the exhaust system is modelled and proper variations for throttle and volumetric efficiency are in.
Due to my not currently having working pedals, I haven't driven any sims in a month probably and haven't tried LFS with the new sounds.
Big time. I'd have paid more that for it in a heartbeat Figure it's been a few pennies on the hour of entertainment, not counting all the demo playing I did before S1 release.
Thanks Actually, Doug is handling the presentation since he does this sort of thing all the time and I never have (plus I will probably have stage fright to a large extent ). I might answer a few questions on data reduction and the VRC end of things if anyone asks, but the majority of it is Doug's deal. I'm sort of along for the ride on this one, really.
I don't think so, no. Part of the reason it's oral only is the proprietary nature of the data. Yeah, even with the RC tires we're being secretive about a lot of it as one of the car manufacturers is involved and is interested in the data.
No recordings of any kind can be made at the conference including pictures and audio. From what I'm told this promotes a more open dialogue between attendees.
SAE is Society of Automotive Engineers. That link is the schedule for a presentation we'll be doing at the Motorsports Engineering Conference in Detroit on Dec 5th. It's a presentation on the tire research we did for Virtual RC Racing 4. Hopefully I'll get to talk to a few other tire geeks too. Can't be a bad place to make friends. Renault F1's Symonds usually attends this. Bobby Rahal is the keynote speaker. Should be interesting
For any still interested in purely kart stuff. Here's Matt Wall, karting Champion in Australia, Japan, and damn near it in the Austrian series, taking a spin at the show.
Thanks for the kind words and encouragement, guys. Means a lot to me
The engine model takes a couple seconds to produce a full torque curve and so on at one mixture setting, so I'm not sure we'll have it adjustable on the fly. Real karters do this though so it's important to have, I feel. Perhaps I can find a way to pull it off. Would be cool to have
I hope we can get Rotax karts into the sim. There is at least one kart manufacturer interested in working with us, but I don't know who that is. Hopefully we can get a lot of manufacturers on board so we have plenty of brands included. I don't think it will be too difficult. This isn't a big dollar series like Nascar or F1 and the like. With VRC we had no trouble getting people involved for free. Literally everybody we asked got on board with no requests for money. I imagine the karting world would be similar.
The dev team consists of three of us. They were originally using Ogre for the physics engine, but I talked them into trying mine instead and now they're hooked So yes, I'm the sole physics guy. Zach is a kart racer himself and quite knowledgable in the dynamics end of things after years of setting up and racing his own kart, so together we ought to have it running quite like the real thing. They've arranged for me to take a spin or two in Jim Cassi's 28HP Rotax kart as well so I can know for sure how the real thing really runs. Also, the 2 stroke engine simulation I wrote for Virtual RC Racing 4 will be included as well so you can tinker with the engine a bit too by trying different exhaust pipes, fuel mixtures, etc.. It's very important in real karting so we want to include that. Actually, they originally wrote me with some questions on engines when I proposed they use the engine sim. Then figured, "hey, you want the physics engine too?" Off we went...
They have the interest of a tire manufacturer and at least one kart manufacturer, whose engineers have agreed to give us data. I'm not sure exactly what we can get from that (hopefully tire data at least), but it should improve things. Right now it's a hoot to drive. Better to drive than anything else in the karting market by a long shot according to those who've run it. But that goes without saying
As far as demos are concerned, I honestly have no idea when that will happen. I'm pretty sure we'll have a demo of some sort eventually. Can't imagine not having one. Got to get people hooked somehow
Doing a really mean engine model like Euro did is not really possible in real time. I wrote a couple of engine simulations like he described, one of which is a two stroke model for the next Virtual RC Racing. While that one can run in just a couple of seconds, my latest four stroke model takes about five minutes to generate a complete torque curve at only one throttle setting. It does something like ten to sixteen engine revolutions per rpm point, and I was using maybe twelve to fifteen points. So it essentially took five minutes for the engine to spin around 120-200 times or so. Pretty far off from real time obviously.
The old engine audio samples I did a couple years ago is actually a very simple engine simulation running in real time though. All eight cylinders are modelled, but it's too simplistic to generate anything useful other than audio.
In short, look up tables are just fine provided you can preprocess all that stuff during development if it's something that can take a long time, like engine simulation. The answer you get for "how much torque is there right now" is the same either way. You just have to make sure there are enough data points. Filling thousands of cells wouldn't be done manually, but rather by the simulation. There's just not a simple little formula that will model an engine remotely accurately, so a much more involved approach is needed and with today's CPU power, look up tables are a must.
In general that's probably true, even though it's not always the case. I suspect the cords have a much greater influence on cornering stiffness then the sidewalls do. Changing cord angles has a massive impact on the stiffness of the belts in different directions, much more so than anything you can probably do to the sidewall.
All else being equal though (cord angles, etc.), a low profile tire will indeed generally have a higher cornering stiffness as you said.
Formula one tires have this as well. The area between the sidewall and tread is so flexible that you can stomp the tires nearly flat. Paul Haney visited an F1 tire manufacturing facility and was surprised to see a bunch of guys doing that. That's just how they stored them.
Anyway, from what I read, the reason for that is likely for allowing the tread to remain flatter and in better contact with the road even when there is undesirable camber present, like under straight line braking when you have a few degrees camber because a passive suspension system just can't give perfect camber control in all situations. So it's a bit of a crutch. That would make sense with kart tires too as you guys run a lot of caster which is going to make some pretty serious camber at high steering angles, and not all of it is probably desirable.
Initial turn in bite, if you're thinking of the same thing I am from the 'behind the wheel' perspective, is likely the initial cornering stiffness, unless you're talking about steering to a large slip angle and then activating your butt-meter. The higher that initial cornering stiffness is the more turn in bite you'll feel. Granted, two tires could still peak at the same slip angle, but the shapes of the curves before that could still be quite different. This is probably what you're feeling as the difference between the two.
Generally speaking that's probably true. However, the internal construction of the tire has a huge influence over the cornering stiffness. One could build two tires of the same tread width, one low profile and one high, and depending on what was done with the cord angle(s) and materials, winding details, and so on, either one could have a higher cornering stiffness.
I know a guy that worked with the test data from the Firestone tread separation case years ago for the US government, and what he described happened to the cornering stiffness when you removed the tread was quite startling. The peak force didn't come in until an unbelievably high slip angle and the initial cornering stiffness was only 1/10th what it otherwise would be with the tread (and maybe the first couple layers of cords, I'm not sure exactly what they removed). So it's safe to say that the internal structure and build have a much larger influence on cornering stiffness than the actual sidewall height.
That being said though, I'd suspect that if you tested a whole bunch of tires you'd probably find that in general, the high profile ones have lower cornering stiffness just as you're saying. I've seen a comparison on a couple hundred bias/radial/cross ply cornering stiffnesses that were laid out in bell type curves on the same graph. The construction definitely had a larger impact on cornering stiffness than the dimensions did. That was probably by design to a large extent, so it's tough to say for sure just from that.
One way to render a track efficiently is to break it into sections along the length of it. I.e., there might be 20 sectors from the start line back to the finish line. When you're in sector 5 you render the stuff in sectors 4 and 6 (ahead and behind you). This is very efficient for a sim like LFS where the circuits are all closed and there's no branching off. If this is how it works, I can see why it might not be possible to take a right on a road on sector 5. You don't ever proceed to sector 6 or 4...
I don't know how it works in LFS, but maybe it's something along those lines, which could make it not really possible.
That being said, I'd have a ball with such a "track" too. Would be fun