The online racing simulator
What really makes an engine sound
(14 posts, started )
What really makes an engine sound
Dear LFSers,
Here are a few questions about the sound of an engine. In particular the sound of a V10 Formula 1 engine.

I'm a composer and for a music piece I'm writing I want to show the way I sometimes hear an engine: as music. I want to look at an engine in a musical way.

There are 2 reasons I'm picking a V10 engine: nostalgia (do I really have to explain this one?) and a more musical reason. Every sound is made up of multiple different sine waves: harmonics. Every wave can be seen as a musical note. In a V10 there are 2 waves which are very prominent. The fundamental (the bottom note), which I think is basicly the rpm (18000 rpm = 300 rotations per second, or Hz). The other one, appearantly, is an octave and a third above that one. If the bottom. Note is 300 Hz, or just above a D4 the other note is just above an F#5, around 750 Hz. But why?

This blog explains a few things.
http://formula1guitar.blogspot ... years-ago-new-engine.html
From what I understand, a cylinder makes an explosion every other time it makes 2 rotations. So, if the fundamental is 300 Hz, you get 300 / 2 * 10 = 1500 Hz. But that's twice the frequency of what you hear. Why? How does he arrive at that other note, 750 Hz in my example? And, if it does make 1 rotation every other time a cylinder fires, do you hear that note (150Hz, a D3) too? And what exactly ARE those sounds?

Am I right when I say that that 1500Hz is made up of 1500 explosions per second? What is the sound of one rotation? How does a single explosion from a cylinder sound? What makes that explosion? The fuel combusting, right? So those 2 notes aren't pure sine waves, they are much more complex. They are made up of explosions, each with their own set of harmonics.

And then there are the other sounds. In what way does the exhaust change the sound? Does it resonate, depending on the length and width of the pipe? How can I calculate that? Which other sounds are there? How does the gear sound? Where does that sound come from? Where does that backfire sound come from? Why does it backfire? Of course, in those days they had traction control. What makes that sound?

I realize this is all quite technical and goes quite deep. I'd like to create a full understanding of the sound. Maybe even to recreate the sound. I tried to fiddle around with LFS' sounds. With that I think I got a fundamental, then the 1st harmonic an octave up and then a second harmonic another octave and a third on top of that.

Who can help me with this? Is there anyone who can explain a few things about the way engines work, or about synthesizing a car engine?

Some other intersting things I found so far:
http://www.youtube.com/watch?v=Iyur9neCmPs
http://youtu.be/DzhFRpJHAwQ (he also explains a bit about the major 3rd note).

I would love to hear it, if anyone knows something about this. Thanks!
Well engine sound is also created by air intake and exhaust as the opening and closure of the valves creates some waving effect on intake and exhaust. These waves are also used for performance purpose as they enable motorist to increase the efficiency of the engine at particular engine speed. The length of the intake/exhaust pipes are important to set the rpm at which these effect occurs.

Then you also have some sounds coming from all the other parts of the engine. You don't really hear that much combustion sounds for petrol engine (in regard to diesel I mean) but you may hear the camshaft also.

In the end, engine sound is a big mix of fluid dynamics and mechanical vibrations.
Hmm right. Well this only gives me more questions. I want to get the essence right, at least, and then see what I'm missing..

I'd also be very happy if anyone knows someone who might be able to help. I have to start somewhere.
Well since you are in to sound engineering I'm gonna ask you a question. Do you know any free program that can analyze the frequency of a sound file and show it to me in a graphic form?

I sometimes produce quite a lot of sounds for the sims I'm playing and to match the rpm of the files I have to typically use my ear in a similar way that artists use to tune their guitars. When I have a sound clip of a 4 cyl engine going at an unknown rpm, I compare it to an another clip of a 4 cyl engine with known rpm. IE. when I play a clip of 4cyl in steady 8000 rpm and at the same time also play the another clip, I will gradually match play speed of the unknown clip. When the sound stops "oscillating" and they match and if I end up with, let's say a 2x play speed for the unknown clip I know that it's a steady 4000 rpm clip.

But sometimes even clips that sound steady, may have little oscillation on the rpm which is hard to tune out just by ear. If I could see a graphic of the clip showing the frequency(s) second by second, I could use this information to tune the rpm of the clip to a genuinely steady state.

A while ago I tried to find a program that could actually tell me the engine rpm from a sound clip, but I realized that seeing only the frequency would work good also.

Quote from Darkone55 :Hmm right. Well this only gives me more questions. I want to get the essence right, at least, and then see what I'm missing..

I'd also be very happy if anyone knows someone who might be able to help. I have to start somewhere.

Basically what he means is that your math is basically right, it's just isn't the combustion that makes the sound. The intake valves and exhaust valves open at the same frequency than the combustion does too. You basically have two instruments, one being intake valves and the other exhaust valves, playing at the same time at the same frequency.

You also have to know that some engines do have irregular firing order and others don't. IE. the American V8 get's it's thunderous sound from irregular firing "rhythm".
Exhaust pipe lengths also play a part, the main reason why boxer-4 cyl sounds different from straight-4 is that they typically have different length primary pipes on the headers for the different cylinder banks. IE if the exhaust exits at the left side of the car, the left exhaust pipes are shorter length than right, so the exhaust pulse created by the exhaust valves opening/closing will reach the end of the exhaust pipe in a shorter time than the right bank will.

It's true that also mechanical sounds are a part of the engine sound (valve train etc.) but the most important is the intake and exhaust notes. When it comes to engines that make enjoyable sounds (ie. racing engine at full tilt) you can barely hear the mechanical noises from under the pulses that are the "note". Especially if you are far away from the sound source.
Quote from BigPeBe :Well since you are in to sound engineering I'm gonna ask you a question. Do you know any free program that can analyze the frequency of a sound file and show it to me in a graphic form?

I sometimes produce quite a lot of sounds for the sims I'm playing and to match the rpm of the files I have to typically use my ear in a similar way that artists use to tune their guitars. When I have a sound clip of a 4 cyl engine going at an unknown rpm, I compare it to an another clip of a 4 cyl engine with known rpm. IE. when I play a clip of 4cyl in steady 8000 rpm and at the same time also play the another clip, I will gradually match play speed of the unknown clip. When the sound stops "oscillating" and they match and if I end up with, let's say a 2x play speed for the unknown clip I know that it's a steady 4000 rpm clip.

But sometimes even clips that sound steady, may have little oscillation on the rpm which is hard to tune out just by ear. If I could see a graphic of the clip showing the frequency(s) second by second, I could use this information to tune the rpm of the clip to a genuinely steady state.

A while ago I tried to find a program that could actually tell me the engine rpm from a sound clip, but I realized that seeing only the frequency would work good also.



Basically what he means is that your math is basically right, it's just isn't the combustion that makes the sound. The intake valves and exhaust valves open at the same frequency than the combustion does too. You basically have two instruments, one being intake valves and the other exhaust valves, playing at the same time at the same frequency.

You also have to know that some engines do have irregular firing order and others don't. IE. the American V8 get's it's thunderous sound from irregular firing "rhythm".
Exhaust pipe lengths also play a part, the main reason why boxer-4 cyl sounds different from straight-4 is that they typically have different length primary pipes on the headers for the different cylinder banks. IE if the exhaust exits at the left side of the car, the left exhaust pipes are shorter length than right, so the exhaust pulse created by the exhaust valves opening/closing will reach the end of the exhaust pipe in a shorter time than the right bank will.

It's true that also mechanical sounds are a part of the engine sound (valve train etc.) but the most important is the intake and exhaust notes. When it comes to engines that make enjoyable sounds (ie. racing engine at full tilt) you can barely hear the mechanical noises from under the pulses that are the "note". Especially if you are far away from the sound source.

Ah, thanks!

Well, you can try SPEAR: http://www.klingbeil.com/spear/
It analyzes audio and then synthesizes that. It does so pretty well. Here's an example: http://i.imgur.com/oCEHcnZ.jpg

It's from this video: https://www.youtube.com/watch?v=RqaJKTRs-Kg So it's the external sound. So yea, doppler effect plays a major role, but that's not important for now. The reverb from the buildings and trees changes the sound too. But onboard video's have such a crappy audio quality, that's very hard to analyze that. Anyway, in this instance it's driving away, so the pitch is lower than the actual revs. 200 Hz = 12000 RPM. Makes sense. I filtered out all 'partials' below -20dB, so only the loudest sine-waves are kept. So, the lowest is just over 200 Hz (G3/G#3), the next is just over 500 Hz (B4/C5). That's indeed an octave and a third interval. The next note is +/- 720 Hz. An F#5? Huh? Then C6/C#6, then over 1000 Hz (makes sense), then another G# and a C and then +/- 2600 Hz, about a D#7 or E7. This could be a harmonic overtone of the G3/G#3 note?

So the hardest thing is to get a very clean engine sound. If you listen to the video I mentioned, you can hear quite heavy shifting, backfiring and traction control very prominently. The quality of the recordings is very bad too. I don't think I have ever heard a recording of an F1 car that sounded even close to it in real life. I've even tried analyzing a V10 in LFS (3L V10, made with LFSTweaker), but even then.. The sound is so complicated, with so many partials, it's hard to see which ones really matter. And even then you get only an impression. It's like talking to someone through the phone, you can still recognize their voice, but it's not the full voice.

About the source of the sound then. You mention the intake valves and exhaust valves. What are they and what do they do exactly to produce a sound? And what sound(s)?

I have indeed noticed that the more regular it fires, the 'tighter' the sound is. The tighter the sound is, the more it is associated with fast cars, I think. About the exhaust. I listened to this video: https://www.youtube.com/watch?v=aP0wzXlWkPA. The character of the sound seems to be about the same, except the filtering, the eq is different. In other words, they are the same partials, but with different volumes. That's what I'm hearing.
Quote from Darkone55 :Well, you can try SPEAR: http://www.klingbeil.com/spear/

Thanks for the tip, I'll see at some point if it can help me with my sound samples.

Quote from Darkone55 :About the source of the sound then. You mention the intake valves and exhaust valves. What are they and what do they do exactly to produce a sound? And what sound(s)?

Well I think it's time for you read about the basic principle of the 4-stroke engine. Quick google found a wiki page http://en.wikipedia.org/wiki/Four-stroke_engine.

Basically when the intake valve opens the piston draws a lot of air to the engine, so the speed of the air in the intake port quickly increases. When the intake valve closes the speed of the air suddenly drops. This creates an air pulse and as we know, sound is basically just pulsating air, so it makes a sound. The same happens with the exhaust valves, but the air just exits the engine this time to the exhaust piping. Exhaust gases of course are a lot higher in temperature, so it creates a big pressure and when it's released to the exhaust the sound is a lot louder than the intake sound (well at least before it gets muffled by the silencers in the exhaust).
Quote from BigPeBe :Thanks for the tip, I'll see at some point if it can help me with my sound samples.

Well I think it's time for you read about the basic principle of the 4-stroke engine. Quick google found a wiki page http://en.wikipedia.org/wiki/Four-stroke_engine.

Basically when the intake valve opens the piston draws a lot of air to the engine, so the speed of the air in the intake port quickly increases. When the intake valve closes the speed of the air suddenly drops. This creates an air pulse and as we know, sound is basically just pulsating air, so it makes a sound. The same happens with the exhaust valves, but the air just exits the engine this time to the exhaust piping. Exhaust gases of course are a lot higher in temperature, so it creates a big pressure and when it's released to the exhaust the sound is a lot louder than the intake sound (well at least before it gets muffled by the silencers in the exhaust).

Cool. Thanks for the link, I've been readling a lot of stuff about engines since yesterday. Guess I'll have to learn about them quite a bit before I can ask more direct questions.
Your sound experiment is way over my head...but I will say that trying to understand the sound from a single cylinder will be easier to interpret....
...then making sense of the resonance at inlet/exhaust should become more apparent...two stroke can be easier or more complicated...depending on your train of thought- one crank rotation or two?
#9 - col
Quote from BigPeBe : ...This creates an air pulse and as we know, sound is basically just pulsating air, so it makes a sound. The same happens with the exhaust valves, but the air just exits the engine this time to the exhaust piping. Exhaust gases of course are a lot higher in temperature, so it creates a big pressure and when it's released to the exhaust the sound is a lot louder than the intake sound (well at least before it gets muffled by the silencers in the exhaust).

Also significant would be that when pressure waves in the exhaust reach transitions in diameter of the tube it causes a change in pressure, this results in some of the wave will be reflected back up the pipe. This also happens at the end of the pipe.
Have a look at Lego engines on YouTube. Despite only running on compressed air they surprisingly like the equivalent gasoline engines, just by using channeling air through two valves per cylinder.
Quote from Anthoop :Your sound experiment is way over my head...but I will say that trying to understand the sound from a single cylinder will be easier to interpret....
...then making sense of the resonance at inlet/exhaust should become more apparent...two stroke can be easier or more complicated...depending on your train of thought- one crank rotation or two?

You're right, I should start with a bit simpler stuff.

Quote from ACCAkut :Have a look at Lego engines on YouTube. Despite only running on compressed air they surprisingly like the equivalent gasoline engines, just by using channeling air through two valves per cylinder.

Thanks, I will!
Quote from ACCAkut :Have a look at Lego engines on YouTube. Despite only running on compressed air they surprisingly like the equivalent gasoline engines, just by using channeling air through two valves per cylinder.

Likewise there isn't really air moving through this system at all, as it's not air tight and some of the passages are fake, but it sounds quite like the real thing despite being really only mechanical sound (and plastic at that)

https://www.youtube.com/watch?v=6JPplxM0hGk
#13 - 5tag
Don't know if you're still into this, but regarding intake sounds have a look at these:


What's really nice is that you can see the throttle opening and closing and the sound in relation to that.

Now what these sound like from onboard?
Stripped race car:


"Street cars":



What they sound like from outside:


note that all of these are 4 cylinder engines.
Intake noises of race engines give them a really agressive sound under load that just puts a giant smile on my face. It's one of the things I miss most about LFS' sound engine.
Absolutely, thanks!

What really makes an engine sound
(14 posts, started )
FGED GREDG RDFGDR GSFDG