Exhaust Noise Measurements

[mikester]

The following summarizes the results of some noise measurements and sound recordings conducted on Saturday, December 4, 2004. Thanks to Randy, Lillie (Modern Wedgie) and her Dad Gil (Ancient Wedgie), Patrick (PatrickO) and Shinoo, we were able to measure exhaust noise from five different systems: the Standard exhaust; Stage II Quicksilver; Quicksilver Track; and Arqray Twin Tip with Sport Cat. I also want to thank Robert Puertas and Justin (Vantage) for their help.

The measurements were conducted using a Larson Davis 824 Sound Level Meter/Spectrum Analyzer (SLM), which is a Type I (“Precision” grade) instrument. The SLM was recently laboratory calibrated. The calibration of the SLM was verified before and after each measurement session. Measurements were conducted using the “slow” meter setting; one-third octave band frequency data was collected unweighted, and then post-processed using Excel to A-weighting, so the data is here shown for both what the exhausts actually produce (the unweighted graphs) and what the typical human ear actually hears (the “A-weighted” graphs)

It is important to note that all data reported here was collected at a distance of 20” from the exhaust tips, at a 45 degree offset from the centerline of the exhaust tip opening. Several RPM levels were measured, including approximately 75% of maximum RPM. This methodology is consistent with international automotive exhaust testing procedures (e.g., "SAE J1169 Measurement of Light Vehicle Exhaust Sound Level under Stationary Conditions", “Blue Book norm for Saloon and Sports Car Races (E12.17.8 Section A)”. In all cases the SLM was attached to a mini-tripod, to ensure that the microphone-exhaust distance and orientation was uniform. Measurements were conducted at Idle (approximately 1,000 RPM), 4000 RPM, and 6,200 RPM.

Sound recordings were conducted in tandem with the noise measurements, by means of a handheld Sony Digital Audio Tape (DAT) recorder. An audio cable was connected from the line-out of the SLM to the line-in of the DAT, without gain of any kind. The calibration levels fed into the SLM were also recorded onto the DAT, for purposes of Quality Assurance and for level playback settings. For those of you who may have a sound meter (“Survey” grade meters can be purchased, for instance, at Radio Shack for not a lot of money, IIRC less than $100, and are also useful in setting up your home sound system!), the reference level would be 94 dBA very close to the speaker (say approximately ½” to 1”). The sound recordings were converted to WAV files on a PC using the Windows Sound Recorder. The links to these sound files (stored in the Gallery section of Elisetalk) are attached to this thread.

Some background for you on noise and acoustics is attached, here:

[mikester]

We should not consider these measurements conclusive – remember that each set of data represents only one series of measurements from each setup, from one Elise and one exhaust, on one particular day. So we should look at this as a “snapshot” only, for our own interest. Some of the cars and exhausts were “run in” (the Standard, the Stage II, and – [I believe] the Quicksilver) and some were not (the Arqray and the Quicksilver Track). Also, after doing these measurements, I think there’s still lots more that could be done, including systematic interior and exterior measurements at speed, which will give us a more “real-world” idea of what the different systems sound like. And, of course it’d be great to be able to measure additional systems, such as the Stage I, a KMS, or others as they become available.

Here you go:

[mikester]

First Up: a comparison of the spectral noise measurements (in 1/3rd Octaves) on an unweighted basis, at idle:

[mikester]

At 4,000:

[mikester]

6,200:

[mikester]

Now the same data but adjusted for A-weighting, which closely approximates the frequency response of the human ear (what the typical human hears).

At idle:

[mikester]

4,000 RPM:

[mikester]

6,200 RPM:

[mikester]

The next set of tables presents the overall A-weighted sound pressure levels (noise levels), again at 20", two ways: one is by exhaust type, and one is by RPM level.

[mikester]

by exhaust type:

[mikester]

Well that's it for now, the sound files are on my home computer so hopefully tonight I'll upload those to the gallery.

[Patricko]

Wow, that is a lot of data

What did we learn?

Is there a big difference between them?

[mikester]

Originally posted by Patricko
Wow, that is a lot of data

What did we learn?

Is there a big difference between them?

In a word, no !

Differences not as great as I was expecting, TBH.

Personally, I'm kind of liking the shape of the Arqray Exhaust spectrum, because it seems to have a bit more of the energy down in the lower frequencies and it's smoother down there, but I would like to hear what it's like in the cabin and outside at speed.

When we were working the course, I really liked the way the Quicksilver track exhaust sounded, on Randy's car. But, yours sounded really good to me as well !

In terms of overall A-weighted levels, your (Stage II) is 5 dBA louder than a standard (which is a noticeable difference) and 1 dB than the Arqray (which is not a noticeable difference, typically). But at 4,000 and 6,200, the Quicksilver is slightly higher.

[shay2nak]

wow! thanks for all the work! Seems like the stage 2 & QS are the loudest. But I hope we can get data on the Arqray single tip and even either one of the Arqrays with the straight pipe.

[LarryB]

mikester- thanks for all the data, nice to see the frequency spectrum. One thing I was wondering about, the Quicksliver vs track, A-Weighted is only 1db lower at 6200 rpm. Off hand thought the numbers that shinoo was given it would have been lower. Given that Randy was flagged for being 1-2db over (with standard QS) it doesn’t seem to allow for much more margin.

[BrianK]

Did shinoo's car (with the Arqray) have sport cats? If so, I'd imagine that would increase the db over the stock cat a touch, yes?

How do you pronounce Arqray anyway? ark' ray'?

[mikester]

It was an Arqray, twin tip, with a sports cat, yes. And yes IIRC that pronunciation seems right.

[shinoo]

Originally posted by BrianK
Did shinoo's car (with the Arqray) have sport cats? If so, I'd imagine that would increase the db over the stock cat a touch, yes?

How do you pronounce Arqray anyway? ark' ray'?

Yep, it had the Sport Cat. You got the pronounciation correct.

Mike,

Thanks for the great work and data.

[shinoo]

Originally posted by LarryB
Given that Randy was flagged for being 1-2db over (with standard QS) it doesn’t seem to allow for much more margin.

Actually the Track version is supposed to have a much larger reduction at 7k rpm than the std version. Mike also took measurements from the distance SCCA uses and we found that none of the systems were louder than the SCCA requirement. We need to look into why SCCA readings were so much higher.

[LarryB]

Doubt the real world track measurements are done to the level Mike as gone through. Then again in the real world as you'd mention you've got the full rev range(under load), intake, tire noise, reflections... makes it hard to know what would be "safe".