Same engine, completely different sound

[gijim]

I was kind of shocked to hear these two recordings of an LS1, one in a C5 vette, and the other in an LS1 F-body.

The F-body has the typical "muscle car" sound:

http://www.corsaperf.com/cfflash.html

But the C5 has an almost exotic (Ferrari-esque) sound:

http://www.corsaperf.com/flashc5.htm

Amazing, no?

My question is this... Is it one thing, or a combination of things which cause the difference?

Facts:
The F-body has a Y-pipe which joins both banks to a single muffler, back to 2 rear exits. Exhaust manifolds are typical stamped and welded steel "clamshell" type.

The C5 has true dual exhausts, with an H(? possibly x? I couldn't confirm) pipe somewhere upstream of the mufflers. Exhaust manifolds are similar to "shorty headers", almost a 4-2-1 design:

http://i13.ebayimg.com/04/i/07/7d/51/e7_1.JPG

So is it a combination of the above? One more than another? Will true duals on an F-body with an H-pipe give the same sound? I'm curious as to your thoughts....

 

[Marquis]

gijim, sorry,I almost forgot: a three pass silencer is very similar to the one you posted, but that is a two pass and a three pass has three runs of pipes inside it (not great for back presure!)

And if subjectively you're satisfied that your engine/vehicle set up is similar to your target 16 valve car, then job done and well done- It's good enough to your subjective ear! However if you compared order plots and recorded each engine in a semi anachoic chamber I think the NVH waterfall or order plots may show a different story. In the vehicle other factors come into it which effect the Noise transfer function into the cabin....

 

[GregLocock]

I was peripherally involved in an active noise cancellation project that added noise back into the cabin to simulate teh sound of a V8.

The lead engineer on this was a very talented NVH engineer, and also a bit of a musician.

He struggled for many weeks trying to match the waterfall plot of the synthesised sound to a V8.

It did not work, subjectively. Finally he ended up tuning it by ear and got a great result.

I'd add, I doubt that any code can really predict enough of the engine sound to give the right answer first time in hardware, certainly 8 years ago a welding torch and a large supply of parts was the quickest way to a good exhaust tune.

Incidentally Goldwave will produce waterfall plots.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[UdjatEye]

Last weekend I programmed a software sampler to simulate any engine configuration based on the firing order and exhaust manifold pipe lengths. The details are boring, but basically for a V8 sound I have 8 single cylinder pulses playing at the same time, but phased 1/8 apart, and each with a resonant filter tuned to a frequency corresponding to the path length from the cylinder to the outlet of the exhaust manifold. It's not as accurate as using a true time delay plus manifold resonance, but the sound ends up similar and it's simple.

I can recreate the sound of the C5 corvette and the Camaro with a nearly identical exhaust system consisting of dual exhaust with a crossover pipe. If my model is right, the C5 has an extremely symmetrical exhaust system with outlets very close together. A symmetric system with outlets far apart, or an asymmetric system with outlets close together will still sound like a lumpy muscle car. The make the symmetric/close outlet system sound like a muscle car in my model, all that has to be done is move the location of one side of the crossover pipe 6" forward or back. The throbbing muscle car sound is due to a path length difference between the sound from the left and right cylinder banks whether that difference is caused by pipe length difference or geometry from tailpipes to listener.

This is a link to a sound file of the same symmetrical dual exhaust with crossover pipe and close outlet, but one side of the crossover pipe is shifted rearward differing amounts compared to the other side. In the first segment the shift is 2", in the second it is 8", and in the third it is shifted 14".

http://media.putfile.com/exhaust-sound-V8

-Eric

 

[GregLocock]

Interesting. There's a bit around 39 seconds that seems to have some V8 character, but most of it sounds too even. Also way too much high frequency, sounds like a recording under the hood!

Are you saying that if you are equidistant from each outlet you won't hear the V8 throb? I beg to differ.



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[gijim]

Eric, that's awesome!!

Remember though, the LS1 does not have a 180 degree crank. There are 2 cylinders firing at the same time. 8-7, 2-6, 5-4, 3-1. I'm told the "doubling up" on even/odd sides makes the v8 idle. I'd be interested if you can somehow simulate that...

 

[UdjatEye]

My simulation is of a cross plane V8, and I haven't done a flat crank V8 yet. In a cross plane V8 no two cylinders fire at the same time, the power strokes are all spaced 90 degrees of crank rotation apart. However, each bank has an uneven firing order so that it isnt r-l-r-l-r-l-r-l but r-l-l-r-l-r-r-l. The uneven pattern of firing in each bank gives the V8 its rumble. However, I discovered in my model that the rumble mostly goes away if the exhaust system is symmetric. The rumble is caused by one cylinder bank dominating the sound of the other either by being louder or getting to the tailpipe first.

A certain amount of nonlinear distortion arises in the sound traveling down exhaust pipes and contributes to the sound of engines. Basically the high pressure part of the wave is hotter and has a higher sound speed, so it catches up with the low pressure part of the wave and distorts the overall sound. The two sequential pulses on a single bank of a V8 can overlap enough to cause more distortion and a popping sound exiting the tailpipe, making single exhaust and small diameter exhaust V8s sound even more lumpy. An interesting special case is the peculiar sound of a diesel V8 that I'm sure you've heard. I can simulate that by using guitar distortion and cranking the low frequencies up. The distortion from a low quality speaker makes these synthesized engine sounds more realistic.

Heres a comparison of the synthesized single exhaust, dual exhaust and side pipe V8 sounds. In the side pipe version you can clearly hear the sound of the left bank of a V8 engine.

http://media.putfile.com/V8comparo

-Eric

 

[gijim]

Now when you say symmetric, do you mean equal lengths, or equal lengths and muffler in the same place, or all that and same number and location of bends?

I would imagine the muffler location and the length of the pipe is critical, but what about the number of bends and location?

My application is a transverse V8, which automatically makes an equal length system harder but not impossible.

 

[UdjatEye]

When I say symmetric, I mean the same on both sides in form and function. However, what is important according to my model is the total time it takes for the sound pulse to get from exhaust port to tailpipe. I experimented with moving a simulated muffler forward and back on one side of an otherwise symmetric system, and all it did was change the resonance frequencies of the tailpipe on one side without adding or subtracting from the V8 rumble. Actually the sound was smoother with the muffler located in a different spot. With regard to bends, I dont know. From working with brass instruments I can say that bends really dont matter as long as the acoustic length of the tube is right, and you can find that by slapping the end and listening to the resonance frequency.

-Eric

 

[gijim]

I'd be interested in hearing a symmetric exhaust with asymmetric muffler location.

 

[SpeedyG]

About 2 years ago i wrote a realtime 0D/1D engine sound simulator. I never finished it but some sounds are quite
interesting.
It runs fine with up to 12 cylinders on an P4 win 3.2GHz.
There are hundreds of parameters, i played some examples with my USB throttle pedal

Example 1:
The very beginning. Switching through engine types. Without room simulation.
Example 1

Example 2:
Supercharged Streetracer (little bit room simulation)
Example 2

Example 3:
Large V8 (perhaps Truck or Diesel Railway Engine, heavy room simulation)
Example 3

greets,
Speedy

 

[kenre]

Nice SpeedyG.

First one sounds like it has bad ignition breakdown about 1/2 way through the file.

Second very V8 like at high revs, down low sound like a 4 stroke single motorbike.

Last file could be a truck engine, except when it hit those high revs

Ken