Anonymous (not verified)
Anonymous's picture
radio shack's "New Digital" sound pressure meter - flawed or accurate?
ethanwiner
ethanwiner's picture
Offline
Last seen: Never ago
Joined: Sep 1 2005 - 2:26pm

> My memory is that the classic analog sound meter is flawed but the flaws are well known and built into everyone's solution <

All inexpensive SPL meters vary all over the place, from unit to unit and from one production run to the next. You can be sure whoever makes those for Radio Shack buys whatever microphone elements are cheapest at the time. I have an RS digital SPL meter I bought about a year ago, and it's surprisingly accurate, but only at low frequencies. Below about 800 Hz it tracks my expensive calibrated mike within 1 dB. Above 800 Hz it deviates by more than 12 dB. I assume this is typical for any inexpensive SPL meter, being acceptable at low frequencies and not so acceptable above the midrange.

--Ethan

papaned
papaned's picture
Offline
Last seen: 1 year 6 days ago
Joined: Sep 5 2005 - 11:21am

Possibly this may answer your question.
I have been using the Radio Shack SPL meter for several years to measure room and speaker response. However, I play the Rives Test CD2 which contains a full range of tracks [20 to 20k herz] which are corrected for the non-linearity of the meter.This may be the flaw you are referring to. I don't believe that the Stereophile Test CD's are so corrected.

BGLeduc
BGLeduc's picture
Offline
Last seen: 11 years 7 months ago
Joined: Sep 1 2005 - 9:46am

My understanding of the widely published correction factors was that they offset the "C" weighting that the meter applies, not to correct any errors that may also exist.

IIRC, this is what the Rives disc does; it's test tones are pre-compensated in level for the "C" weighting, so that the user can simply plot levels without further compensation.

Ethan, does that sound right to you?

Brian

ethanwiner
ethanwiner's picture
Offline
Last seen: Never ago
Joined: Sep 1 2005 - 2:26pm

Brian,

> does that sound right to you? <

I read somewhere once (uh oh) that somebody tested a large number of Radio Shack SPL meters and found they all varied by quite a lot. That sounds very believeable to me. This is not a precision calibrated instrument! It's a cheap device that probably costs all of $5 including manufacturing and shipping to the US. This is not to say it's not useful. I use one myself when I'm too lazy to schlep the extra electronics needed to use my "good" microphone. I assume all Radio Shack SPL meters are useful below about 1 KHz, and not so useful above that.

--Ethan

BGLeduc
BGLeduc's picture
Offline
Last seen: 11 years 7 months ago
Joined: Sep 1 2005 - 9:46am

In case anyone is interested, this link will bring you to a site that has a free prgram designed to work with a RS SPL meter and the Behringer Feedback Detroyer parametric EQ.

http://homepage.ntlworld.com/john.mulcahy/roomeq/index.html

The fellow that wrote the program was a designer with TAGMclaren. I can not comment on its overall accuracy, but I have used it to dial in my BFD, and the measurements were comparable to what I got with my Infinity RABOS kit.

Brian

SonnieP
SonnieP's picture
Offline
Last seen: Never ago
Joined: Dec 20 2005 - 9:24pm

I'd venture to say that the Radio Shack SPL Meter mic is NOT accurate. There are widely know correction values for use with the meter (especially for the bass region):

There have been some issues brought up in the various forums about the accuracy of the Radio Shack SPL Meter mic and whether or not the Correction Values are actually correct. There have been measured differences between a calibrated quality mic and the RS SPL Meter mic. There are also measured differences between the analog and digital versions of the RS Meter.

The differences between my RS mic and Behringer ECM8000 were quite dramatic. My ECM8k is out being calibrated and I should have it back within a few days. I will then remeasure against the RS mic, but I'm fairly certain that even with the corrections it is still going to be inaccurate and I will be recommending a different mic for measuring in the future.

BFD GUIDE

ethanwiner
ethanwiner's picture
Offline
Last seen: Never ago
Joined: Sep 1 2005 - 2:26pm

Sonnie,

> There have been measured differences between a calibrated quality mic and the RS SPL Meter mic. There are also measured differences between the analog and digital versions of the RS Meter. <

In the test I did, mentioned above, with my RS meter and an expensive calibrated microphone, the two tracked each other very closely - within 1 dB - between 20 and 800 Hz. I have no idea if this is typical. My guess is that every batch of SPL meters Radio Shack buys from China probably has whatever electret element is cheapest on the market at the time. So the most likely answer is that they're all unknown because none of them are calibrated.

--Ethan

SonnieP
SonnieP's picture
Offline
Last seen: Never ago
Joined: Dec 20 2005 - 9:24pm

oops... guess I wasn't logged in on that last post.

Sonnie

ethanwiner
ethanwiner's picture
Offline
Last seen: Never ago
Joined: Sep 1 2005 - 2:26pm

Sonny,

> I know Ilkka over at AVS came up with some pretty dramatic differences when he measured against his pro mic. <

I don't recall that post so I can't comment except to say that all "pro" microphones are not equal. There's also a lot that can go wrong when measuring room response, so (without impugning anyone's knowledge) it may be he didn't measure properly. For example, if the microphones are even an inch or two apart, that alone could account for a large difference in the measured response.

> below 20hz it was off by as much as 7db <

I didn't measure below 20 Hz. But who cares what happens that low anyway? If a movie explosion is off by a coupla dB below 20 Hz, no big deal.

--Ethan

Lamont Sanford
Lamont Sanford's picture
Offline
Last seen: 1 year 1 month ago
Joined: Mar 31 2006 - 8:32pm

My poor man method. This affordable meter has its weaknesses that can be overcome with ballpark results using the meter's default settings.

I prefer to average the measurements. Like taking the average for 20, 31.5, 40 Hz for the 31.5 reading to be plotted on a graph. You'll find that the average compared to actual (with the ever popular adjustment charts) will remain true to the actual measurement until you get into the higher ranges that will be all over the place with the actual measurements. Averaging mellows out the meter's weakness at high ranges. The logic being that the average is more realistic than the actual measurement at higher ranges. It's a cheap way out of getting your hands on some pricey measuring instruments only to end up with approx. the same graph. Also, at the most lowest ranges you are not going to get anything above 50dB so you need to put the mic up against the the face of the loudspeaker cabinet, take the reading and subtract 6dB (in addition to adding the published adjustments) for each metric meter of distance you are actually testing from. I prefer the 1 metric meter distance for the RS meter.

Better yet, if you can register the meter at 80dB with a pink track, switch the meter to the 70dB or if need be, 80dB range and get through an entire log sweep track (20-20k) without the meter going below or over the set range than you can avoid all the graphing and consider yourself fortunate as far as acceptable frequency response is concerned. No different than suspecting a bad alternator but you have no equipment so you disconnect the battery while the motor is running to see if it dies or not. There is more than one way to skin a cat. It just depends on your tools and how delicate you want to do it. After all, it is only good for frequency response to begin with. This isn't rocket science. It's low grade statistical sampling (sort of). Close estimations.

Also, if you do the whole graph thing and then do the whole log sweep thing described above and get the same results than you know you are close enough for your ladies.

imispgh
imispgh's picture
Offline
Last seen: 9 years 5 months ago
Joined: May 23 2006 - 10:37am

Isn't that the useful range anyway? At higher frequencies don't comb filtering affects "confuse" the device more?

imispgh
imispgh's picture
Offline
Last seen: 9 years 5 months ago
Joined: May 23 2006 - 10:37am

An alternative is to get an RTA program for the PC and use that (paying attention to the microphone used). I use RAL and it works great. (I actually use the mic that came with my PC. It seems to work really well. When I compare it to the RS the RS seems to drop off too much below 50hz. Maybe it's output feed is not as linear as the meter?)

ethanwiner
ethanwiner's picture
Offline
Last seen: Never ago
Joined: Sep 1 2005 - 2:26pm

> Isn't that the useful range anyway? At higher frequencies don't comb filtering affects "confuse" the device more? <

Yes and Yes. I use the ETF software which can do 1/3 octave etc, and also does very high resolution.

--Ethan

cyclebrain
cyclebrain's picture
Offline
Last seen: 2 months 1 week ago
Joined: Jun 16 2006 - 11:40pm

Sorry, I couldn't help it.
The new digital SPL meter doesn't have the same smooth results as the analog meter.
The digital meter works best with CD's and server based audio.
The analog meter works best with vinyl and FM radio.

Come on people, they are both analog sensors. The field that they both measure is analog. Air pressure. This analog unit is then converted to a digital output.

But I am really interested in how they calibrate a microphone.
What is the reference?

ethanwiner
ethanwiner's picture
Offline
Last seen: Never ago
Joined: Sep 1 2005 - 2:26pm


Quote:
What is the reference?


One way uses a mechanical "clapper" that creates a sound at a known SPL level. A friend of mine has an expensive microphone that came with a calibrator. The calibrator slides over the microphone so the "clapper" is very close to the microphone element. You push a button and it makes a click sound.

I'm not an expert on this and there may be other methods.

--Ethan

Elk
Elk's picture
Offline
Last seen: 3 years 7 months ago
Joined: Dec 26 2006 - 6:32am


Quote:

The digital meter works best with CD's and server based audio.
The analog meter works best with vinyl and FM radio.


SAS Audio
SAS Audio's picture
Offline
Last seen: 4 months 4 weeks ago
Joined: Jun 6 2007 - 6:56am

Unfortunately, there are some huge discrepancies between Ethan's lastest mics graph and his earlier graphs of the Earthworks and RS meters. His earlier graphs show the mics response is quite close to the Earthworks.

First the two earlier graphs from Ethan. Notice the Earthworks mic approx 250hz to 20khz.

Now compare to the Radio Shack spl meter, 250hz to 20khz.

Notice the similarity and how close they measure. Now check out his recent latest graph.

Notice the huge 20-22db discrepancy at 20khz, 13db at 12khz in Ethan's latest graph. Not even close in the first two graphs he presented earlier.

The average overall response from mid frequencies to high frequencies is minimally affected by narrow band comb filtering peaks and nulls and room nodes do not affect the overall average.

Here are some accurate tables/ easy to graph correction figures.

http://www.digital-recordings.com/audiocd/radio.html

http://www.gearslutz.com/board/post-prod...tion-table.html

http://www.audioasylum.com/audio/general/messages/49147.html

Dear IM,

No, comb filtering minimally affects the average response at higher frequencies because the sharp peaks and nulls are such narrow bandwidths.

Hope this helps.

Jim Tavegia
Jim Tavegia's picture
Offline
Last seen: Never ago
Joined: Sep 1 2005 - 4:27pm

Behringer makes a measurement mic that I believe is fairly accurate, the model 8000?, that sells for around $49. Even with a cheap mic pre like the M-Audio Audio Buddy, could be used to make some in room comparisons. It is a cheap omni, but good enough to try. The noise level is a little high, but a pair of them can make some pretty decent 2 channel recordings, especially for your kids or a school concert.

An inexpensive downloadable audio recording program from NCH Swift Sound as a built in graphic analyzer for reviewing recorded music which you could use. It is called Wave Pad. With a good sound card you can even record at 24/96.

SAS Audio
SAS Audio's picture
Offline
Last seen: 4 months 4 weeks ago
Joined: Jun 6 2007 - 6:56am


Quote:
Behringer makes a measurement mic that I believe is fairly accurate, the model 8000?, that sells for around $49. Even with a cheap mic pre like the M-Audio Audio Buddy, could be used to make some in room comparisons. It is a cheap omni, but good enough to try. The noise level is a little high, but a pair of them can make some pretty decent 2 channel recordings, especially for your kids or a school concert.

An inexpensive downloadable audio recording program from NCH Swift Sound as a built in graphic analyzer for reviewing recorded music which you could use. It is called Wave Pad. With a good sound card you can even record at 24/96.

Thanks Jim! Much appreciated. However, the RS meter is not that far off. The 3 correction tables will provide close approximation. The RS meter is actually pretty consistent and flat, except some 7db low at 20hz, 4db at 30hz and a couple of db low at 40hz.

Ethan's earlier graphs confirms that the RS meter is close to the Earthworks mic. If the RS meter were that far off, that information would have been posted by many other individuals on many forums.

Thanks again Jim.
Steve

ethanwiner
ethanwiner's picture
Offline
Last seen: Never ago
Joined: Sep 1 2005 - 2:26pm


Quote:
Behringer makes a measurement mic that I believe is fairly accurate, the model 8000?, that sells for around $49.


Yes, and if cost is a factor, for $10 less you can get the Nady which is basically the same thing. See my Comparison of Ten Measuring Microphones article.

--Ethan

ethanwiner
ethanwiner's picture
Offline
Last seen: Never ago
Joined: Sep 1 2005 - 2:26pm


Quote:
Ethan's earlier graphs confirms that the RS meter is close to the Earthworks mic.


Man, you don't give up do you?

Neither of those graphs you posted were measured with an Earthworks microphone. Steve, if you want to learn about room measurements, you have to read my articles more carefully.

In any case, the tests I did in my article Comparison of Ten Measuring Microphones are accurate, and trump all the various "correction" curves made by amateurs you'll find around the 'net.

--Ethan

ethanwiner
ethanwiner's picture
Offline
Last seen: Never ago
Joined: Sep 1 2005 - 2:26pm

Heh, I can't let this one pass:


Quote:
Here are some accurate tables/ easy to graph correction figures.
http://www.digital-recordings.com/audiocd/radio.html
http://www.gearslutz.com/board/post-prod...tion-table.html
http://www.audioasylum.com/audio/general/messages/49147.html


Steve, none of those tables agree with each other, so how can they all be accurate? The first link says the RS meter is off 1 dB at 20 kHz, and the third link says it's off by 11.2 dB. Yet according to you they're both correct?

--Ethan

Elk
Elk's picture
Offline
Last seen: 3 years 7 months ago
Joined: Dec 26 2006 - 6:32am

Interesting article, Ethan.

You are one of the few that mentions how to point an omni mic when measuring room response. Even the manufacturers and others, such as Velodyne that include mics, don't mention this. Kudos.

SAS Audio
SAS Audio's picture
Offline
Last seen: 4 months 4 weeks ago
Joined: Jun 6 2007 - 6:56am


Quote:
Heh, I can't let this one pass:


Quote:
Here are some accurate tables/ easy to graph correction figures.
http://www.digital-recordings.com/audiocd/radio.html
http://www.gearslutz.com/board/post-prod...tion-table.html
http://www.audioasylum.com/audio/general/messages/49147.html


Steve, none of those tables agree with each other, so how can they all be accurate? The first link says the RS meter is off 1 dB at 20 kHz, and the third link says it's off by 11.2 dB. Yet according to you they're both correct?

--Ethan

I am just trying to figure out how you came up with such erroneous measurements concerning the RS spl meters (in the 10 mic comparison).

You see Ethan, the link/article with the 11db correction at 20khz was for the analog RS spl meter. The other links describe correction for the RS digital spl meter.

Something else you may not have thought of, which will help you see that there is a problem with your latest RS meter measurements (10 mic comparison) is the comment http://forum.stereophile.com/forum/showf...part=5&vc=1 (#43786)
when you stated you could not remember if you had used the super linear AKG mic or the RS spl meter when making other room measurements.


Quote:
Mackie HR824 but I don't recall if I used my Radio Shack SPL meter or my AKG C451 with a calibrated CK22 capsule.

The mics were basically indistinguishable. If the RS meters were 24db down at 20khz as shown on your graph, one would have easily seen the huge high frequency rolloff compared to the AKG, and been able to identify which mic/meter one was using (RS or AKG with calibrated CK22 capsule).

Another way of looking at it is if the RS meter read 24db down at 20khz, 22db lower than the AKG or Earthworks mic, the RS mic would only read/measure only approx 6% of the signal compared to those mics. I hope this helps you to see the problem with your latest RS measurement in the 10 mic comparison.

It is also obvious from your earlier graphs I posted (post #50367) that something is obviously wrong with your latest RS meter 10 mic graph/measurements.

And no one else has ever posted that the RS spl meters are that far off, 24db down at 20khz, as your latest RS meter measurements (10 mic comparison) show. Comments on the RS meter's failure regarding this problem would be all over the forums.

Can anyone imagine an owner trying to raise the tweeter 24db; anyone for an ultrasonic ear cleaning and splitting headache?

Obviously your latest RS meter measurements (10 mic comparison) are grossly inaccurate and needs to be corrected without delay.

ethanwiner
ethanwiner's picture
Offline
Last seen: Never ago
Joined: Sep 1 2005 - 2:26pm

Steve, all I can suggest is you learn a little about room acoustics and room measuring, and learn a little manners, then maybe we can have an intelligent discussion.

To anyone who cares, it may not be clear from this thread, but Steve apparently has some sort of vendetta. He has been all over the web claiming my microphone tests are incompetent. One thread here was locked after he couldn't refrain from insults. Another over at Audio Central was deleted outright, and when he pulled this at the AVS forum the regulars told Steve he doesn't know what he's talking about. Which I believe is quite clear in this thread.

Also, when Steve started making these accusatory posts I called him on the phone and offered to discuss it so we'd both understand the issues. He hung up on me. So clearly Steve has some sort of "issue" with me. Which is a pity because he could learn much from me.

BTW Steve, say Hi to Joel for me.

--Ethan

SAS Audio
SAS Audio's picture
Offline
Last seen: 4 months 4 weeks ago
Joined: Jun 6 2007 - 6:56am


Quote:
Steve, all I can suggest is you learn a little about room acoustics and room measuring, and learn a little manners, then maybe we can have an intelligent discussion.

First, continued success, but please correct the error in your latest RS meter measurements (10 mic comparison).

I simply presented the evidence in my last post, including your own evidence Ethan (the graphs which is still on your website), just the evidence. But I see no response to the evidence I presented. The fact is, the evidence is true, including your own earlier evidence I presented.

I simply reponded to your posts that you placed at all the forums I frequent concerning the latest RS meter graph (10 mic comparison). So why are you posting all over? I mean you thought the RS meter used to be pretty good yourself as your earlier measurements show.
Now your view of the RS meter has substantially changed.

Earlier you measured and found the RS meter very close to both the AKG and Earthworks mic at high frequencies. So the RS meter is quite accurate at high frequencies as you and others have demonstrated.

Now, you claim the RS meter is incapable of measuring highs, being 24db down at 20khz, 13db down at 12khz. How can you undo the accuracy you and others have already shown to be true?? Obviously, something has drastically changed on your part.

Here is the link to the AVS string Ethan refers too in his post.
http://www.avsforum.com/avs-vb/showthread.php?t=1063837&page=2

First, notice there are some who measured the electronics and supported Ethan. However, after I posted Ethan's own evidence, his support leaves later on page 2. (I got these graphs from Ethan's website, he has two websites.)

Also notice in this string, which occurred several weeks ago, Ethan does not complain about the Earthworks mic being used in one of the graphs I posted (although one can still edit one's post in that string).
But now/here, he objects, stating the Earthworks mic was not used in either graph.


Quote:
Neither of those graphs you posted were measured with an Earthworks microphone

The point is that the RS meter response measured quite close to both the Earthworks and AKG meter/mic in earlier measurements (high freqencies accurate) that Ethan performed. And now Ethan claims the RS meter does not, and is grossly inaccurate. How do you undo the accuracy that has already been demonstrated multiple times?

If some think the RS meter seems to be ok at high frequencies in their homes, that is because it is accurate. You are correct.


Quote:
Also, when Steve started making these accusatory posts I called him on the phone and offered to discuss it so we'd both understand the issues. He hung up on me. So clearly Steve has some sort of "issue" with me. Which is a pity because he could learn much from me.

Not quite the truth Ethan. First it was another string you posted. Second, he asked if I would give him the federal investigator's name who was watching him. I stated I should not have even mentioned that as the Feds were upset with me. Pretty sick conversation and so I hung up on him. After the call, Ethan posted and taunted me to name the investigator. since he knew I could not respond.

The rest of Ethan's post is not worth mentioning.

In the final analysis, the evidence is most important. In previous experiments, Ethan demonstrated the RS mic was accurate to both the AKG and Earthworks mics. Now, the RS digital meter is measured as being grossly off by 24db at 20khz and 13db off at 12khz.

Continued success, but please correct the gross error in your latest RS meter measurements (10 mic comparison).

ethanwiner
ethanwiner's picture
Offline
Last seen: Never ago
Joined: Sep 1 2005 - 2:26pm

I rest my case.

bifcake
bifcake's picture
Offline
Last seen: Never ago
Joined: Nov 27 2005 - 2:27am


Quote:

...he asked if I would give him the federal investigator's name who was watching him. I stated I should not have even mentioned that as the Feds were upset with me. Pretty sick conversation and so I hung up on him. After the call, Ethan posted and taunted me to name the investigator. since he knew I could not respond.

Bahahhahahahaha I can just see it now: Ethan being investigated for dealing acoustic treatments!!

Ethan standing on a street corner: Wanna bass trap? I got bass traps... want some traps? I got good Jamaican traps here...


Quote:

Continued success, but please correct the gross error in your latest RS meter measurements (10 mic comparison).

Concerning the RS measurements, I remember reading that the RS meters weren't all that accurate across the entire frequency range. I don't have the time right now to search for the article (I'm at work), but I think Ethan's measurements are pretty accurate.

pbarach
pbarach's picture
Offline
Last seen: 2 weeks 6 days ago
Joined: Sep 8 2005 - 3:10am

Newbie question: How do you position the mic on the Radio Shack meter at the listening position? Pointed up at the ceiling, or directly forward? And does anyone know what mic position was used to come up with the "corrections" included on the Rives Test CD2 or posted here and there?

Elk
Elk's picture
Offline
Last seen: 3 years 7 months ago
Joined: Dec 26 2006 - 6:32am

As it has an omni-directional mic, it should point up when measuring room acoustics.

The best method is to use a tri-pod, both because it is repeatable and because the presence of your body can influence the readings, especially at higher frequencies.

My assumption is that Ethan's measurements, and corrections, were created in this fashion but I do not know for certain.

SAS Audio
SAS Audio's picture
Offline
Last seen: 4 months 4 weeks ago
Joined: Jun 6 2007 - 6:56am


Quote:
As it has an omni-directional mic, it should point up when measuring room acoustics.

The best method is to use a tri-pod, both because it is repeatable and because the presence of your body can influence the readings, especially at higher frequencies.

My assumption is that Ethan's measurements, and corrections, were created in this fashion but I do not know for certain.

Hi Elk,

I believe the RS meter is very directional. If the meter is pointed off axis from the speakers, or visa versa, severe high frequency rolloff will occur.

Hope this helps.

Elk
Elk's picture
Offline
Last seen: 3 years 7 months ago
Joined: Dec 26 2006 - 6:32am


Quote:
I believe the RS meter is very directional. If the meter is pointed off axis from the speakers, or visa versa, severe high frequency rolloff will occur.


This may be true. I don't know. Have you played with one to see?

ethanwiner
ethanwiner's picture
Offline
Last seen: Never ago
Joined: Sep 1 2005 - 2:26pm


Quote:
How do you position the mic on the Radio Shack meter at the listening position? Pointed up at the ceiling, or directly forward?


Most omni microphones will be closer to flat when pointed at the source. But when balancing speakers in a surround system you need to point it at the ceiling. The quote below is from my article Comparison of Ten Measuring Microphones. The last graph in that article compares a Radio Shack SPL meter pointed both ways.

--Ethan


Quote:
Although omnidirectional microphones supposedly receive sound equally from all directions, when measuring rooms and loudspeakers the convention is to "aim" the microphone upward. No omni microphone has exactly the same frequency response from all directions, though microphones with tiny diaphragms and slim bodies are often more uniform than larger models. So when balancing loudspeaker volume levels on a surround system, pointing the microphone toward the ceiling favors all of the loudspeakers equally.

Log in or register to post comments
-->
  • X