Accuracy of measurements...

**Exocer** · **09-06-06, 02:23 PM**

I own the newer RS analog SPL meter and am getting excellent measurements with my Sealed Avalanche 15 sub. I wonder if they're too good to be true? Im dying to get a calibrated mic asap though... Here are some measurements.

Nearfield.

Listening position

What are some ways I can test for accuracy? The spl meter seems to have found the Tuning frequency of my Natalie P's to be 28hz (in-room) instead of my goal to tune them to 32hz...pretty close. My Velodyne DPS 10 looked like this before EQ

and after

Seems like the thing is working...but I cant believe the Avalanche is that flat in-room with no sort of EQ behind it.

**brucek** · **09-06-06, 03:52 PM**

Quote:

What are some ways I can test for accuracy?

Well, when Sonnie and I were creating the calibration files for the three RS meters, measured against the ECM8000 that Sonnie had professionally calibrated, we overlayed our final responses using the cal files we had developed.

We felt that if we measured the same signal with the three different RS meters (using our newly generated cal files) against the ECM8000 and the responses were fairly close, then we'd be OK.

Here is a set of response graphs of the three types of RS meters using the cal files posted here at the Shack against the ECM 8000. We actually had two digitals at the time.

Here's the 10Hz to 150Hz graph....

final_four_meters_with_new _cals_loaded.jpg

Here we look a little closer from 10hz to 60Hz to check accuracy.

final_four_meters_with_new _cals_loaded_resolute.jpg

Here we separate the different graphs to get a closer look.... You see the ECM8000, an old analog, a new analog and two new digitsl using the new cal files......

final_four_meters_with_new _cals_loaded_trace_seperate.jpg

They're close enough for me. Is that close enough for you?

brucek

**Sonnie** · **09-06-06, 06:14 PM**

Robert... that is remarkable response on your Avalanche 15.... awesome indeed with no eq at all.

I wouldn't think the meter would be more than a db or two different, but then again you could have gotten one with a bad capsule inside... only way to know for absolutely sure is to measure it against another SPL meter or a calibrated mic.

**Exocer** · **09-06-06, 06:57 PM**

Bruce, they're definitely close enough for me. I wonder about my particular SPL meter, not your calibration files...You guys have done a remarkable job

Sonnie, I guess that'd be the only way to know for sure huh?
I guess i'll be confident with the measurements i've got so far. At least they're not totally unbelievable like 120db at 10hz from a single sealed 15" sub.

**Sonnie** · **09-07-06, 01:54 AM**

You seem to do a lot of speaker projects that need measurement... I'd recommend the Behringer ECM8000, especially for full range. Go ahead and get it calibrated. Leonard has just sent his off to another fellow to compare calibration with our recommended guy... so you might wanna wait til he gets his back and lets see how close mine is first. If his is drastically different, I may end up getting my re-calibrated.

Here we go again???

Maybe not!

**Exocer** · **09-07-06, 06:29 PM**

I'm hoping the measurements are close enough for satisfaction. Is there a thread about this? I am very interested in keeping track of this experiment

**Ilkka** · **09-25-06, 04:59 PM**

Without some heavy EQ/LLT, the near field response you show can't be accurate. It should start sloping much higher with around 12 dB/oct. slope. I recommend getting a calibrated mic (or at least non-calibrated ECM8000) and measure again.

**kesa32** · **10-09-06, 07:14 AM**

Hi all, l'm wondering if would be able to steer me in the right direction here.......l have a newer analog rs meter and am wondering what allowances l will have to make with it just using it manaully c weighted for my sub.......any information would be appreciated........cheers ken

**brucek** · **10-09-06, 08:09 AM**

Quote:

l have a newer analog rs meter and am wondering what allowances l will have to make with it just using it manaully c weighted for my sub

See here and download the Excel workbook for your specific meter and the compensation values are built into the file.

Or, you can download the XXXX.cal file from that page (which is just a text file) and use the values of meter compensation.

brucek

**kesa32** · **10-09-06, 08:31 AM**

Quote:

brucek wrote:

See here and download the Excel workbook for your specific meter and the compensation values are built into the file.

Or, you can download the XXXX.cal file from that page (which is just a text file) and use the values of meter compensation.

brucek

Thaks for that brucek
this means that my ib is down around 25db @10hz ( nearfield )...that sound about right??...cheers ken

**brucek** · **10-09-06, 10:21 AM**

Quote:

this means that my ib is down around 25db @10hz ( nearfield )...that sound about right??...

Well, if you use the newrsanalog.cal file, you would add +7.29db @10hz to the reading you get with the meter to compensate for the inaccuracy at that frequency.

brucek

**kesa32** · **10-09-06, 05:40 PM**

Quote:

brucek wrote:

Well, if you use the newrsanalog.cal file, you would add +7.29db @10hz to the reading you get with the meter to compensate for the inaccuracy at that frequency.

brucek

thanks bruce........yeah that is corrected level......my drivers were brand new when l did the measure a few weeks ago so maybe they're broken in more now and have changed slightly...either that or there is some attenuation going on in my avr's lfe.....cheers ken

**wenwenguo** · **06-18-08, 12:49 PM**

i AGREE WITH YOU

**Ricci** · **07-03-08, 11:53 AM**

Bruce,
I've got a question about the CM140 cal file. How was this file determined? I'm interested in the correction below 20hz and most importantly the 10hz and lower correction values. I recently measured my output at 5hz with one of my subs with slow C weighting and 80-130db range, 5hz sine wave. The meter showed 105.1db. Now if I add the correction value of 24db to this my true SPL would be 129.1db at 5hz? That's why I'm wondering about the correction values down low.

**brucek** · **07-03-08, 12:45 PM**

Quote:

I've got a question about the CM140 cal file. How was this file determined?

It was created by comparison measurements taken near field, using Sonnies subwoofer and his calibrated ECM8000 (which is calibrated down to 5Hz). Quite a few Galaxy meters were tested and they were actually quite consistent. When the cal file was complete we loaded it and did a comparison against the ECM (with its own cal file loaded) and the two results tracked very closely.

But it matters not, because we found that the galaxy tracked quite close to a real C-Weight curve and so even if you cleared your cal file and simply checked the C-Weight box instead, it would still be fairly accurate. Our file just trims it up a bit. A real C-Weight curve compensate about 25dB at 5Hz (so you can see the cal file is close).

Either way, I'm not a fan of measuring anything below 10Hz...... why bother.

brucek

**Ricci** · **07-03-08, 04:34 PM**

Thanks.

Why? Because there is content in movies down that low, and I'd like to try to reproduce it if possible. Also because why not?

**NEO Dan** · **Yesterday, 07:50 PM**

Bruce,
Like Ricci I am very concerned about one thing in particular:

The LF consistency of the ECM8000 and CM-140 has been called into question.

Ilkka has basically declared that unless the Mic is calibrated it is not to be trusted for LF measurement.

Has anyone found this to be the case?

**lcaillo** · **Today, 05:21 AM**

This is the case with any measurement tool, particularly at the extremes of the operating range.

**Ilkka** · **Today, 05:22 AM**

Quote:

NEO Dan wrote:

Bruce,
Like Ricci I am very concerned about one thing in particular:

The LF consistency of the ECM8000 and CM-140 has been called into question.

Ilkka has basically declared that unless the Mic is calibrated it is not to be trusted for LF measurement.

Has anyone found this to be the case?

As harsh it may sound, I wouldn't trust in any microphone that hasn't been individually calibrated, especially when it comes to the extreme ends of the operating range. Down to 20 Hz is usually okay, but measuring 5 Hz or similar with an uncalibrated microphone (and the rest of the measuring loop) is like taking a stab in the dark.

Advertisement