Proper RTA measurement / actual results

OK peeps, as promised I had an extensive testing session with Mr. Floyd re: proper usage of the Audio COntrol 3055 AS PER IASCA SQ PROCEDURE, and I offer this as a device so that we may all learn from it and utilize this information in the future when tuning and tweaking our systems for competition. And although it was brought up in another post that the Audio COntrol user manual specifically states one way to do it, IASCA always has utilized the following "unique" method of doing basically the same thing. So, here we go>

Scenario: A/C 3055 with 26" mic stand and 136dB RTA mic placed in driver's seat of Jorge's Ram. Official IASCA pink noise track #28 set to play on repeat mode. RTA eq curve called up and ready to roll.

STEP 1- power up meter, set SPL range adjustment knob to 90dB, set increment knob to 3 dB steps, set sensitivity knob to ZERO.

STEP 2- Press and hold the SPL button so that HI and LO scroll, release the button when it is LO.

STEP 3- Adjust the system volume until the peak SPL level reaches 90.0 dB +/- 0.5

STEP 4- Press SPL button until "IASCA" is displayed and release, this puts the meter into the 30 point IASCA freq response analysis mode and will display 1/3 octave resolution of your curve on a window that is + and - 12dB referenced to center (so cumulatively a 24dB "range")

STEP 5- LEAVING the spl range selector at 90 dB, and setting the sampling rate to "MED" or medium, slowly adjust the input sensitivity up or down so that the peak (highest) frequency is exactly ON or as close to ON the CENTER REFERENCE LINE as possible. Depending on the car, this could be a small adjustment, or you may have to adjust this knob all the way to +10.

STEP 6- Once step 5 is completed, press FAST rate, count to 5, press MEDIUM rate, count to 5, press SLOW rate, count to 5, then press FREEZE. THis is the average frequency resolution of said vehicle referenced to the 90 DB level.

STEP 7- Press the scoring button on the back of the machine and this is your score out of a possible 30 points.

Thats the procedure, folks. Straight from the horse's mouth, and it always has been. Now, I have been to shows where they used only the MEDIUM rate setting and tried several attempts at FREEZING the best curve, I have also seen numerous attempts at doing the same on the fast and slow settings trying to get the competitor's best possible curve. NEVER have I seen it done where the dB range was then changed to 80 in a competition, and none of us can recall when this method was, if ever, used.

So, how'd we do with the sample testing??? Glad you asked, Johnson.

First, of course, we wanted to test the theory postulated by a couple of ya that frequency resolution of a given system changes with increases or decreases in overall system volume. I already knew the answer, but we had to get proof nonetheless.

BASELINE MEASUREMENT---89.6 DB using the above-mentioned method exactly as described.
Scores of 5 consecutive curves, each one measured after complete meter reset and re-calibrate as per above method.
Score 1- 29
Score 2- 29
Score 3- 30
Score 4- 29
Score 5- 29

- We checked the actual SPL measurement AFTER setting the sensitivity level as described above, just to see if the difference was as dramatic as some argued. IIRC, someone suggested the difference was as much as 10 dB, maybe even more! YIKES! well, I am happy to report that, when the meter is set correctly using the above technique, Unlce Audio is right again. I said it was a difference of +/- 3 dB, and the actual measurement differences we got during each of the above 5 runs were 2.6 dB to 3.0 dB of difference , and thats even with the sensitivity at damn near +10!

OK, so next, we set the SPL level to 100 dB, then followed the exact same steps as above, this time we had to decrease the sensitivity, of course.

Score 1- 29
Score 2- 30
Score 3- 29
Score 4- 29
Score 5- 29

Hmmm. Exact same findings.

OK, so next, I did 5 wide-open volume runs, peak SPL was 110.6 dB, and only varied .2 each time. And BOY HOWDY the speakers didnt like me very much lol!

Score 1- 29
Score 2- 29
Score 3- 29
Score 4- 30
Score 5- 28

Hmm, so this truck maxed out playing pink noise at 110 dB scores the EXACT same as it does at 90 dB. I think maybe the last run dropping to a 28 might have had something to do with battery voltage at this point...so we put on the power supply.

Now, to give the "changing response curve" theory the benefit of the doubt, we tried a LOWER volume than the normal 90 dB. We set the SPL range to 80 dB, set the overall SPL to 80.1 dB, then followed the proper steps as above>

Score 1- 28
Score 2- 29
Score 3- 30
Score 4- 30
Score 5- 29

OK, so the scores are STILL the same, and therefore the theory that frequncy response changes with system volume is bunk! I said it before, I'll state it again- if a system's response changes with overall volume, then that system is either 1) not designed properly and has a bad linearity issue 2) HAs signal clipping somewhere in the chain 3) is experiencing driver failure or high distortion levels 4) is improperly damped 5) some other wacky phenomenon that would for all intents and purposes preclude it from being World Champ material in the first place.

NEXT ITEM UP FOR BIDS, BOB...
we decided to test MY theory that the event noise floor, NOT JUDGES ERROR , was the actual reason for the low scores at NOPI.

I jumped in and boosted JUST THE SUBS by 6 dB ( I didnt want to go any higher because any more could be dismissed as too much noise floor, but any less would be deemed "not enough"...the fellow chaps agreed.)

We then followed the EXACT same steps mentioned above referenced to a 90.1 dB spl level, and here are the results>

Score 1- 18
Score 2- 18
Score 3- 20
Score 4- 18
Score 5- 20

AMAZING! The same system that immediately scored in the 29 range just moments before, now had TOO MUCH BASS and could not muster better than a 20. THere were between 3 and 6 bands not even present on the screen at any given time. And this is solely because, when we do the "adjust the sensitivity" step, the boosted bass hits the CENTER "zero-reference" line too soon and prevents the rest of the curve
from being visualized. All three of us agreed that if a vehicle has too much bass, and obviously some of them did, then the scores will suffer in an otherwise good car. So, this means that a major reason that some of the cars did poorly the first go round was less likely a judging error and MORE likely a matter of boosted noise floor and the fact that the competitors did NOT account for it! I dont know how I can be any more clear.

TO further prove the bass mal-adjustment was the sole cause of the poor scores, Terry told me to got in an cut the subs down 3 dB less that the original 29-curve setting. I did, and we got the following results>

Score 1- 28
Score 2- 28
Score 3- 26
Score 4- 25
Score 5- 26

Interesting, eh? Clearly the scores are better when we cut the subs than having them boosted too much, and looking at the curves the overall curve was slightly louder than the 29 curve, obviously due to the peak frequency shift into the midbass region during 90dB level setting, and the main source of the points deductions was the sharp roll-off below 60 Hz now.

So whether or not this proves anything is up to you guys. What it says to me is, while of course we need to standardize this method of RTA testing across the board for everyone, there is more to the lesson for all of us, Especially the noobs that tuned for RTA the night before, when it was quiet>
some factors to consider are
1) of course, ambient noise floor
2) Vehicle deadening/ dampening
3) Temperature
4) Barometric pressure
5) Humidity
6) adjusting vehilce gain structure for peak full-range spl to at least hit 105 dB (see linearity section)

The above reasons are why RTA should be double-checked right before entering the testing lanes so that all of these factors are accounted for. Trust me, these things change the response of a car dramatically. The cars that have been competing a long time have got it nailed for the most part, and you guys will notice vast improvements in your scores if you pay attention to the above factors. They can even drastically change your SPL scores too.

Thanks for listening, and if any of you have any questions, please feel free to email me and I will do my best to answer or put you in touch with Terry---he loves to teach

Peace!

Ahh, the plot thikcens...Great point! Now I can discuss further what I just mentioned to Andy, but address specifically the very FACTORS of frequency response I originally posted about..AWESOME!

OK, Leonard's car, for example, is a car that I would consider a "seasoned" car, being that he is not a rookie and has been competiing for what, 3 or 4 years, right?
OK, now that we have established this, lets look at one of the factors I mentioned...Vehicle damping and resonance control.

I am willing to bet that Leonard, as well as the other veteran competitors, have gone to some lengths, even great lengths, to properly dampen their vehicles to acheive great sound performance characteristics. At least we can all pretty much agree on this, right? Speaking for myself, the Big Meat is literally covered with Dynamat over the ENTIRE cabin, including inner and outer door skins, and the floors have been treated with Cascade Audio VB3 lead-neoprene sheets. This makes my cabin VERY VERY quiet, one reason for the great sound in there.
Using the Civic as an example, I have gone to even greater lengths to deaden the environment, everything from dynamat to spray-on Stingeer Roadkill to modelling clay to acoustic foam to CONCRETE! (Thats right folks, there is concrete in the rear 1/4 panels and some on the floor to add mass and dampen vibration). Thats an example of my Pro Ultimate car's dampening. I am willing to bet that Leonard, like others who have dedicated more than a single year to competition, have gone to similar lengths.

So, what does this mean?

THis means that a VETERAN CAR is LESS susceptible to ambient noise floor issues than perhaps as much as all ROOKIE (first-time) cars because of the additional damping that occurs with perpetual tuning and system improvements.
How many rookies can afford to address or even think of addressing, in their first competition year, proper damping and resonance control? Typically, if a guy just dropped $5 to $10K or more on a competition system, would he have also added additional fundage to cover materials and labor to dampen and isolate noises, etc? Sure it COULD happen with one or two of them, but lets be realistic and note that it is generally an afterthought, as many vibrations and acoustic treatment problems arent foud until AFTER the system is done and all the panels and whatnot "settle".

With me so far?

Using myself as an example, again. Big Meat has a 3 band parametric EQ. My system relies more on proper damping, proper gain structure, proper crossover selection and slope, and proper arrival times to arrive at its current state of great sound. I have never been able to get better than a 26 on RTA with this set-up because I simply do not have enough EQ flexibility to adjust the 4 anomalies that are present. ..I did not touch my setting for RTA testing, did not even recheck it the day of the show! And yet, my score was 25! Hmm, noise floor didnt affect my score much, maybe I'm wrong.

BUT WAIT! Let's use Jorge's truck as the next example and compare! Jorge's truck has MINIMAL sound deadening in it compared to mine. Yes he has the requisite dynamat Xtreme over the whole cab, but not the floors. Yes he has dempening in his speaker enclosures, but not the dash or underdash like I do. Also, the rear of his cab is not reinforced like mine is due to my large cut-through...allowing his cabin to "resonate" when the back wall flexes ever so slightly.

Jorge DOES have adequate EQ to get a great frequency response score (as evidenced by the original test results), but when I double-checked HIS response with the meter before going into the lanes, it was outta whack! What was causing it??? Well, chaps, it's very very simple---the noise floor. Why was it affecting Jorge? Quite simply, his cabin was resonating with the ambient frequency.
Sooooo I re-tuned as best I could to accomodate that resonance and scoring in the neighborhood of 28-29, we ran it through the lanes, and we scored a 24. Why? Well, at 930-1000AM when I retuned the system, the noise floor was X-loudness, but by the time we waited in the lanes to get tested, another hour had passed, and now the noise floor was Y-loudness, obviously a higher value.

So, lets come full circle here and compare the two systems...Jorge, in his second season competing, does NOT have the amount of acoustic damping and resonation treatments that I do in my Pro-class truck. Therefore, his vehilce is more susceptible to ambient resonation that mine. THis caused his otherwise great scoring vehicle that routinely (see original post) gets a 29 0r 30 to suddenly garner a 24, EVEN AFTER ADDITIONAL ADJUSTMENT to try to correct for the noise floor. THis is further proven that I scored ON PAR with my system, which I NEGLECTED TO ADJUST BEFORE TESTING, becuase my vehilce is FAR LESS SUSCEPTIBLE to these conditions.

Would it be safe to assume that almost all rookie cars there could probably use some additional dampening and resonance control to perform like the PRO and Expert class cars did? YOU BET YOUR TUSHY!