Measurement Mic Shootout (EMM-6, WM-61A, RS 33-2055, Audyssey) - Page 2

Looking at the graph below the blue is the digital RS SPL meter set to C-Weight with no calibration file loaded into REW and the C-Weighted SPL meter checkbox unchecked in REW. The green is with the cal file from HTS loaded (the one notnyt mentioned above) and C-Weighted SPL meter checkbox checked in REW (this only has an effect below the range of the .cal file though).

I believe when notnyt mentions C-weighted he means SPL meter on C-Weight and C-weighted SPL Meter checkbox checked in REW. I didn't check that configuration but it can be derived from what I did measure (it would be about 7dB low at 10hz and about 3dB low at 20hz).

With my RS (digital) meter set to C weight (not REW) my meter was close but still a little high (most are low) for the below 20hz material. Per everything that I have read you were supposed to use C weighting and the correction table. Doing that gave me ridiculously high below 20hz readings.

Per REW at Theater Shack:
"An SPL meter with a line level analogue output. The Radio Shack meter is perfectly adequate for low frequency room acoustics work, either the analogue or digital display version. The Galaxy CM-140 meter has better tracking of the C-weight curve and better behaviour above subwoofer frequencies than the RS meter, but is more expensive. Calibration files for the various models of the RS meter and the CM-140 can be found in the Downloads area of the subwoofer equalization and calibration forum at www.hometheatershack.com/forums/"

"Your SPL meter's range should be set to the value normally used for speaker level calibration and must not be altered while using REW. If you are using the Radio Shack meter, select the 80dB range if you calibrate your system at 75dB (this is the standard level recommended by DolbyTM).

Set your meter to C weighting and "slow".

I just ran a few sweeps with my ACOPacific mic/pre/PS vs the Audessey mic (no calibration file) and my digital RS meter that's 10 years old:

ACOPacific vs Audessey:



ACOPacific vs RS Digital Meter with C-Weight CHECKED:



Bosso

It's a 33-2055. The Manual is copyright 2000.

The receiver is MultEQ.

The interface is down -1dB at 3 Hz. No Calibration.

Bosso

Actually, I have the opposite reaction. If the Audessey mic is any more accurate than this, why are we buying measurement microphones?



Bosso

Yes. Opposite. I find it interesting that the other graphs don't show the Aud mic rolling off at all.

I think my result is more in keeping with the mics spec:

It appears, when compared to the ACO mic, that my Aud mic is flat to around 22 Hz, then rolls off below that, which makes much more sense than seeing virtually no roll off.

As a matter of fact, that's the only reason I did the comparison.

Bosso

I think nontyt is saying that you agree with him, not that your reaction is the opposite.

Gotcha. Sorry for the word twist confusion.

Fact is that if someone hadn't pointed out what should be the obvious, one would be left with the impression that an Aud mic is flat to single digits with no calibration file.

The question remains, what is going on that causes that result? Since I've purposely built a measurement system that requires no calibration files, I'm not at all familiar with those systems or the calibration file process and use.

Bosso

Those results are kind of interesting. The 18 to 28Hz frequency range that has that deep dip on your RS SPL meter sure looks abnormal to me. How much smoothing did you use in your curves? What were your measurement conditions?

Why not try a close mic of a sealed subwoofer with both microphones? That should be more accurate for the low end.

Funny thing is I can make 10 different measurements with the same mic and the FR curves all look different. In the free field a difference of few inches one way or the other (up, down, left, right, forward, back) makes the FR chart reading all look different.

I used the RS Digital SPL meter to close mic measure my NHT speaker's 10" sealed subwoofer driver. The straight SPL C scale factors were used for compensation. While I am not sure of the absolute accuracy of the RS SPL meter, the sealed speakers roll off on my chart does not show any abnormal dips. Then again, I replaced the original microphone element in my meter a few years ago.

The close mic measurements are probably the most reliable for the low frequency evaluation. That should eliminate the room effects as much as is practical. Kind of easy to place each mic in a near identical position for a close mic low frequency test.

Yes, there's definitely something wrong with the RS meter sweep. Probably a near-dead battery. I plan to try again.

But, what you seem to be saying is that you could have saved the $$ and effort and just used the RS meter with a C-Weight cal file, which I ain't buyin'.

Bosso

I'm having a hard time reconciling that as well...

The way I understand what Bosso is getting at is, if the RS meters measure just as good as other more expense mic setups, why would anyone purchase the more expense ones since the RS meters are just as good?

Example, Ricci is doing all of the hard work measuring the subs as he has been, if the RS is just a good, why would he not use the RS meter instead of the setup he did use?

And just to make sure, I am in no way saying anything bad towards Ricci (or what - how he did those tests since it is past my experience level), just using that a comparison between the setup he used for measurements and why he chose not to use an RS meter.

If I read Bosso's posts incorrectly than I do apologize.

James

wow, great info. thumbs up.

this will be very usefull when i finish my sub builds.

notny, a few questions.

1) which audyssey mic did you use (i assume there are quite a few around). i have one that come with the Integra 9.9

2) did you use the "mic in" input on your motherboard sound? or did you use an external sound card? the sticky over at the HTS suggest using the "line in" and not "mic in" input?

thanks

NYT,

I have no disagreements with that part. I have seen many a post in different forums where results are questioned since it was done using a RS meter.

Once I can get REW setup correctly I will try to compare my Omni, RS and Audyssey mics. My Audyssey mic is different than the one you showed a picture of but came with my Onkyo 885.

James

thanks, and you did mentioned that there was not mic cal on the audyssey?

which sound card did you use?

and you mentioned you used HDMI for audio out?

Ahhh, you're killin' me, man.

I'm not questioning that you get a similar FR with various electrets through your measurement signal chain/calibration files. That much is obvious.

I'm saying flat out that none of those systems approaches the accuracy of the ACOPacific/Edirol UA1000 system below 20 Hz.

The Aud mic roll off should be evidence enough. Do you really think that if the Aud mic is flat to 7 Hz, they would spec it as being down -3dB at 20 Hz? I used to have the Aud mic FR graphs, but they're gone and I see that Audyssey has removed them from their site, but that's irrelevant. There's no chance in this universe that your rig and the Aud cap are flat to single digits. None.

AV Talk's Slartibartfast uses the LinearX M53 with a calibration file (which includes cap/preamp). Here's the LinearX graph of its uncorrected FR, with the grayed out area representing the variation spectrum from one mic to the next.

I scaled and laid over the ACO 7012 mic uncorrected FR. Each ACO mic comes with a FR graph to 10 Hz, made at the time of manufacture. It will always be ruler flat below 10k Hz. The top end is where the ripple is, down -2dB at 40k Hz.



Of course, his measurements are accurate to 15 Hz, the bottom of his measurements, to within 1dB because his mic/preamp are professionally calibrated and a file created and burned to disc that he loads into his software, but the point here is that the ACO rig does not require a calibration file for subwoofer measurements.

Here are the published specs for the ACO mic, pre, PSU and Edirol UA1000:



I bet a months pay that the Aud mic will not be +/- 1dB with the ACO rig.

Bosso

So either the EMM-6 is not calibrated that well or the Audyssey mics can differ.