Just off hand does anybody know how many db you increase in noise volume when you delete the cat on the evo x? With that whats the db increase between a hfc and test pipe?

 

No single number will be generally accurate. The increase in "loudness" will depend on many other factors, not the least of which is the cat-back. Also, what humans hear as "loudness" has a very particular frequency weighting; 10 more dB at one frequency will be much worse (or less bad) than 10 more dB at a different frequency.

In short, your question is too simple.

 

Too many variables. However when I had my VIII using a RadioShack Db meter in neutral at idle 90+ Db with full 3" turbo back O2 housing and cams. It was hello neighbors at 5:30am for pre-commute warmup.

 

I know metric is annoying, but it's dB, not db or Db. The d is for deci, as in tenths, and the B is for Bel, as in the log unit of sound pressure. Also, when using a sound meter, set it to A weighting and fast response, as that corresponds more closely to human hearing.

 

A weighting of course. Mostly we don't hear the low end of of the sound spectrum. However we hear the effect it has on the things around us, like the dishes rattling in the cabinets in my kitchen. Surely in the neighbors houses also at least a bit.

 

some vendors annotate dB levels for CBE w/ or w/out additional mods
you'd think that more vendors would supply this info instead of leaving it to the consumer to buy until they find what works.
either way....thank god for youtube

 

It would be nice, but there would need to be an established standard. Something like: WOT, third gear, 4000 rpm, measured inside cabin with all windows closed. But even that won't really provide what we want because the engine speed that produces the peak level of loudness depends on things like the placement of any mid-pipe resonators. One exhaust might be quieter than another at 4000 rpm, but deafening at a different rpm, simply because the resonator is a foot further back. So we'd end up needing the peak loudness from a complete gear-pull and I just don't see vendors going through the hassle.

 

As a test engineer I have to say that you're absolutely right. They'd have to ensure exact conditions are met in order to display dB ratings. I've only seen one vendor provide ratings but they did not mention specific RPM range. I'd guess it would read something like max hp or torque: 487hp @ 5300RPM but rather 92dB @ 3300 RPM/2Gear/WOT. Of course, as you noted certain conditions would have to be met. Say stock Evo X w/ Cobb CBE w/ Cobb HFC max dB ..... Luckily this would be a one time thing.

This could actually be a great sales point for companies by taking the guess work out of which exhaust is louder or quieter. I'm sure the California guys would love this. I hear the cops are a drag there. Personally I had a tough time with finding an exhaust b/c of the kids. Couldn't be too loud that the wife would complain but not completely neutered so I choose the Speciality X. I'm in love.

Any vendors want to experiment with this? I'm pretty sure you could find some stock Evo who'd love to do it.

 

I would suggest C-weighted and slow response to get an idea as to the actual change made by the exhaust mod. dBA is an EQ'd weighting emphasizing the mids and omitting the lows, generally speaking. As the bulk of the sound generated by the exhaust rests on the low end of the spectrum, an A-weighted number doesn't really exemplify the idea.

FWIW, the RS SPL meter is actually fairly accurate. Relative readings stand as such, but one does need to add a correction factor for a comparative numbers:
http://www.danmarx.org/audioinnovation/rsmeter.html

 

I guess it depends on what you want the readings for. If it's to see what it would be like to be in the car, then I'd stick with the settings that come closest to human hearing. If you want the actual sound pressure, then a flatter curve like C makes sense. Finally, if you're just trying to stay legal, then I'd find out what the sanctioning body or police use. FYI: the SCCA uses A-fast, 75', right angle to travel, for example.

 

Herein lies the rub. If we sample "normal" factory exhausts on, say, family cars we'll likely find the A-weighted reading to correlate fairly closely with one's perception of the noise. OTOH, if we consider high-performance exhausts that have a built in emphasis on the low-end (even some factory models) then dBA isn't as effective.

This being said, as you mentioned, much of this is speed-dependent. As speed goes up, all things staying equal, the lower the frequency the quicker it dissipates. How an exhaust with an elevated low end responds at speed, given a propensity to drone, is an educated guess. I tried finding a MLSSA plot to illustrate the idea, using a high-performance exhaust, but did not succeed.

In terms of approximating the human hearing, it sounds great on the surface. I mean, who wouldn't want a model of such a thing, right? As you know, the idea was based on the infamous Fletcher-Munson curve which was supposed to show the sensitivity of the human ear at a given frequency. Thus, the A-weighted EQ was born, OSHA standards materialized, and everyone and his grandmother started quoting a number in the low-100s as the threshold of pain and citing a jet takeoff at ~120dB (depending on the chart). But, does this really show how loud something may be? I'm sure that folks have had the experience of sitting on their respective front porches and feeling the ground tremble (along with a certain sense of personal discomfort) as a big truck drives by. This does not exist in the A-weighted world. As dBA eschews low frequencies altogether, such a thing would not show on a reading. Without having to write a tome, I think that folks can extrapolate the effect on a measurement of an exhaust. How how do we really measure "loud" in this application? Good question.


I think most use A-weighted.