Strange RS SPL meter calibration results

[Ilkka]

A friend of mine wanted to have his newer type analog Radio Shack SPL meter calibrated, so he sent it to me. I noticed some strange behaviour while calibrating it against to my professionally calibrated microphone. The amount of correction needed wasn't anywhere near to what Sonnie's correction file suggest for this type of meter. I believe a few graphs explain it better. All the measurements were done with the REW (4.00), I merely used the TrueRTA to show them.

The first graph shows the difference between the correction files for my (i.e. my friend's) new analog meter and Sonnie's new analog meter. Quite a difference!



The second graph shows the difference between the correction files for my new analog meter, my old analog meter, and Sonnie's old analog meter. As you can see, they all match reasonably well (within ~3 dB). My new analog meter seems to be only a little bit better than the old ones.



So what does this all mean? Assuming both Sonnie and I took accurate measurements, some of the new analog meters are better then the others. The worse ones aren't much better than the old analog ones. This will definitely cause some strange measurement results when all the people having the new analog meter are using the same correction file. Their frequency response can look like having much more low frequency roll-off than it in reality has.

So it looks like some of the new analog meters are still having the insides of the old analog meter. How do we separate them? I took a picture of the QC stamp, which reads "10A05". Could people post their QC codes for all the RS meters they have?

Here's the list:

New analog (soft-look)
04A06 (tatkinson)
12A05 (Chrisbee, toecheese)
10A05 (Ilkka's friend)
06A05 (WillyD, Average Joe)
4A05 (norpus)
02A04 (JRace)

Old analog (boxy)
07A06 (Jeffeory)
11A05 (daxie)
12A03 (Ilkka)
10A02 (Peter De Smidt)
09A01 (snatcher)
9A4 (Chrisbee)

[brucek]

So it looks like some of the new analog meters are still having the insides of the old analog meter.

I can't see any other explaination. Granted, Sonnie only tested a few when the calibrations were done, but a random pick of two that reveal the same response ain't bad....

I seem to remember when Chrisbee said his new analog didn't give him anywhere near as good a response as he was measuring with the old meter. I wondered at the time why they weren't the same... This could be it.

I do have to agree with something you've commented on before, and that's if you want to test down to 10hz it is probably a better idea to get a good microphone like many of us have resorted to..

brucek

[Ilkka]

brucek wrote:

I can't see any other explaination. Granted, Sonnie only tested a few when the calibrations were done, but a random pick of two that reveal the same response ain't bad....

I believe RS made some changes (different mic capsule etc.) to their analog meters, but not necessarily at the same time they changed the outlook. It is more than likely that both Sonnie's meters were these "upgraded" ones, while my friend's isn't. Eventhough they all look the same.

I do have to agree with something you've commented on before, and that's if you want to test down to 10hz it is probably a better idea to get a good microphone like many of us have resorted to..

brucek

I agree 100%.

[Chrisbee]

Recently purchased soft-look analogue: QC 12A05 TEC

12 year old (?) boxy shaped analogue: QC 9A4 TEC

I shall do an REW sub response curve for each meter without loading any meter calibration files tomorrow.

[brucek]

I shall do an REW sub response curve for each meter without loading any meter calibration files tomorrow.

I would do it near field to remove as many spikes etc from the room that makes comparisons difficult.

I'd tape the two mics together.....

brucek

[Sonnie]

This could definitely throw a kink in things, especially for those purchasing newer analog meters.

I do have the $99 Galaxy CM-140 in hand, but haven't done any measurements with it. It may be that we simply need to advise people that they will need to either have someone check their RS meter for accuracy or purchase one of the CM-140's, if it turns out to be as accurate as we suspect it will be.

[Chrisbee]

Okay. Nearfield measurements at 4" from the dustcap of bottom driver of my IB.

Both meters carefully calibrated individually to 80dB prior to measurement.

Meter calibration file in REW cleared prior to each measurement to ensure no cal file loaded.

Here's the old analogue meter.



Here's the new analogue meter.



Like two peas in a pod!!!

[brucek]

Yep...... I guess you better use the cal file for the old meter.

Ilkka has made an interesting discovery?

What can we do about it?

brucek

[Chrisbee]

Those of us with two meters at least have a control (of sorts).

The chances of two meters being so alike is er.... very unlikely.

It certainly explains why my new meter behaves exactly like old meter when using the old RS analogue cal file.

But can we trust Sonnie to have got the old RS analogue cal file right?

How many were in his sample?

Does this mean I can now have my bass back which Sonnie stole by giving us the new rs analogue file?

[Ilkka]

Chrisbee wrote:

Okay. Nearfield measurements at 4" from the dustcap of bottom driver of my IB.

Both meters carefully calibrated individually to 80dB prior to measurement.

Meter calibration file in REW cleared prior to each measurement to ensure no cal file loaded.

Like two peas in a pod!!!

Thank you! Just what I thought. You should definitely use Sonnie's old analog correction file for both of them.

And might I say, that's a fine looking NF response. When you put the correction file in, you'll have loads of low frequency gain! How does it look at the LP?

[Chrisbee]

Ilkka wrote:

Thank you! Just what I thought. You should definitely use Sonnie's old analog correction file for both of them.

Agreed.

And might I say, that's a fine looking NF response. When you put the correction file in, you'll have loads of low frequency gain! How does it look at the LP?

This was my best curve after a 16dB 120/60BW boost at 20Hz using the BFD.

[Ilkka]

Chrisbee wrote:

This was my best curve after a 16dB 120/60BW boost at 20Hz using the BFD.

That looks strange. Could you take a NF measurement (4" from the dustcap) WITH the right correction file loaded? Do not use any BFD boost etc.

[brucek]

How many were in his sample?

The purpose of the exercise to create the new calibration files came from the widely held claim that the new RS meters were different and more accurate than the old RS meter and so Sonnie agreed to test a few against his calibrated ECM8000. Then we could come up with closer cal files than the low resolution one that had been circulated for years.

Sonnie got two new digital meters, two new analog meters and his own old analog to compare against the ECM8000.

A new file was created for the old RS meter against his single old RS meter.
A new file was created for the new analog meter against two new analog meters.
A new file was created for the new digital meter against two new digital meters.

The decision was made that if two meters compared the same, then that was as large a sample as we would use. The two new analogs tracked well and the two new digitals tracked well. Arguable a small sample, but since Sonnie paid for these himself, I think it was good enough.

Once completed, Sonnie did a retest to compare the final results against his ECM. It was close enough and we posted the results for anyone who dared to use them.....

Here is a jpg of a final test of the ECM8000 against the two new digital and two new analog with their posted cal files loaded. This pic didn't have the old analog in it, but it was just as accurate when compared with its cal file at the time.

It would indeed appear though, that they may have switched the mic elements in the new ones at some time after they started producing them? What can you do.



final_four_meters_with_new _cals_loaded.jpg

brucek

[Chrisbee]

Ilkka wrote:

That looks strange. Could you take a NF measurement (4" from the dustcap) WITH the right correction file loaded? Do not use any BFD boost etc.

Nearfield without boost:



Thanks for the explanation brucek.

[Sonnie]

Also keep in mind that Ilkka created correction values for an old analog meter with a completely different mic and we were within 1.5db of each other, except at 10Hz where I still say he had a blurp...

[Ilkka]

Thank you. But that doesn't make sence...? How can THIS...



Turn into THIS when you apply the correction file? The deep bass should go UP, not down.

[brucek]

Turn into THIS when you apply the correction file?

Scaling.....

With BFD boost nearfield old analogue file new meter.

Your scaling makes it hard to interpret.

Can you repost it with vertical scale 45dB - 105dB and horizontal 10Hz - 200Hz.....

brucek

[Chrisbee]

brucek wrote:

Scaling.....

Your scaling makes it hard to interpret.

Can you repost it with vertical scale 45dB - 105dB and horizontal 10Hz - 200Hz.....

brucek

Both plots are in exactly that format.

I should have scaled to 800 not the 880 I have been using.

I have also lifted the target curve to 80dB which is my calibration level in REW.

[Chrisbee]

The first plot is nearfield with the BFD boost but no cal file.

The second plot is no BFD but with cal file.

I think.

[Ilkka]

Chrisbee wrote:

The first plot is nearfield with the BFD boost but no cal file.

The second plot is no BFD but with cal file.

I think.

Arrgh...

[Chrisbee]

We can't all be experts like you, Ilkka.

I now have a whole mass of completely meaningless graphs because you questioned the old V new RS meter.

I now have whole threads which are completely meaningless because they are based on the wrong meter cal files.

But I bet my bass is better than yours even if I can't manage a simple graph.

[Ilkka]

Chrisbee wrote:

We can't all be experts like you, Ilkka.

I was just kidding. Although it does help if you tell ALL the settings etc. used for each graph.