Home Theater Forum and Systems banner

1 - 10 of 10 Posts

·
Registered
Joined
·
16 Posts
Discussion Starter #1
When I saw this bug fix below I decided to buy a UMIK-1 from Cross Spectrums Labs and upgrade REW to beta release 23.

• Bug fix: SPL readings from USB mics were 6 dB low (verified on UMIK-1, suspected to affect Omnimic and UMM-6 also).

REW recognized the UMIK-1 and I used the zero degree cal file from Cross Spectrums Labs with the following line added to the top of the file that I got from the MiniDSP site.

"Sens Factor =0.4938dB, SERNO: 7005156"

I also have a Radio Shack 33-2050 SPL meter. When I use the REW pink noise generator and compare the UMIK-1 via the REW SPL Meter to the Radio Shack SPL meter, REW reads 4 db higher than the SPL meter (75 db with the RS meter and 79 db with REW). The Radio Shack meter is set to “C” weighting and meter response set to slow; REW is set the same way. I have also tried the REW “Z” weighting but it makes no difference. I am measuring just my left and right stereo speakers.

What accounts for this difference and more importantly, what is my true reference level? Is it 75 db as the RS SPL meter reads or the 79 db REW reports?
 

·
Premium Member
Joined
·
1,562 Posts
What accounts for this difference and more importantly, what is my true reference level? Is it 75 db as the RS SPL meter reads or the 79 db REW reports?
I would suggest you use the cal for the UMIK-1 if you are typical user of REW for home audio setup.

Additional Comments just FYI:
4dB is more difference than would be expected/common, but it is a very possible result considering the variability in calibration of these inexpensive units. If the RS meter was not individually calibrated it may be more suspect than the mic. You could also just split the difference if you like.

Some of the measured difference may be due to using 2 speakers for the level measurements. Unless the 2 mic/meter capsules were very close to each other comb filtering may be involved. You may want to try again using only one speaker and holding the RS meter close to the mic.

If you are using REW for general setup of an audio system as many of us are then either of the 2 values can be used with no significant issues. We are mostly concerned with variation across the frequency range and not the absolute SPL level. The SPL level is a less significant factor in sound quality and most here adjust level to preference in actual usage anyway.

If you have special requirements for highly accurate level calibration, that is another subject.
 

·
Registered
Joined
·
16 Posts
Discussion Starter #3
Jtalden,

Thank you for your reply. You are correct I am just using REW for home audio setup so a highly accurate level calibration is not absolutely necessary. However with beta release 23, I had hoped these level differences had been resolved. See the link below to a post regarding the exact same issue where the poster is using a Galaxy CM-140 SPL meter and a UMIK-1. In his case the SPL meter was reading 5 dB higher than REW, in my case the opposite occurs where REW is reading 4 db higher. The date of the post is 5/12/14 and the beta 23 release was first available on 10/19/14.

http://www.hometheatershack.com/forums/rew-forum/78170-spl-levels-cm-140-vs-umik-1-rew.html

For my tests, the RS SPL meter was mounted 6”behind and 3”above the UMIK-1 on the same mic stand. Following your suggestion I removed the RS SPL meter from the stand and held it beside the UMIK-1, then started the pink noise generator (low cut 200 Hz - high cut 2000 Hz) and still get the same 4 db difference. Next I tried using just one speaker and again got the same 4 dB difference. My RS SPL meter has never been professionally calibrated. In my case perhaps using a REW release before Beta 23 would give closer results, but Beta 23 was suppose to fix the issue not make it worse.
 

·
Premium Member
Joined
·
1,562 Posts
Yes, 4 dB is greater than I would expect as well. I don't know of any easy way to resolve the issue though.

If you are questioning if the offset used by REW SPL meter for the UMIK-1 is really correct, only JohnM can confirm that. There were early problem with both REW offset and the UMIK-1 factory calibration process I believe. My impression was like yours that these have been sorted out.
 

·
Registered
Joined
·
16 Posts
Discussion Starter #5
John,

Both jtalden and I would be very interested in your thoughts on this problem. Please reply when you can and thank you in advance for your response.
 

·
Premium Member
Joined
·
6,711 Posts
The REW change was made after checking 4 mics with a 94 dB calibrator using the MiniDSP sensitivity figures, with these results:

700-0139: 93.8 dB
700-0355: 94.3 dB
700-3718: 93.1 dB
700-4653: 93.5 dB

The calibrator claimed to be accurate to 0.5 dB, but that is just a manufacturer's claim so it could be out.
 

·
Premium Member
Joined
·
1,562 Posts
John,

Both jtalden and I would be very interested in your thoughts on this problem. Please reply when you can and thank you in advance for your response.
Really? It's a legitimate question/concern, but it is yours, not mine.

I made clear that my understanding was that the problems have been resolved. You hinted that you questioned this and I pointed you to JohnM as the only resource to answer your question.

I'm probably overly sensitive to this, but don't appreciate your referencing me in this. I put up with this technique too often in my professional life. It may be corporate favorite technique, but don't need to see it here. It may have been a simple misunderstanding, but I am not sure how. If it is an honest mistake, fine. If it was an attempt to add credibility/importance to your concern then it is both uncalled for and unnecessary. JohnM has been very responsive to concerns addressed to him from all HTS members.

Please don't speak for others unless you are sure they are onboard.

Thanks
 

·
Registered
Joined
·
16 Posts
Discussion Starter #8
Jtalden I am very sorry if including you in my question to John offended you. That certainly was not my intent, please accept my sincere apology.

Going forward I will either get my Radio Shack SPL meter calibration checked or purchase a Galaxy CM-140 to compare against the Radio Shack meter. This may take me a week or so, but I will report back my findings.
 

·
Registered
Joined
·
16 Posts
Discussion Starter #9
I purchased a Galaxy CM-140 to compare against my Radio Shack 33-2050 SPL meter. Using the REW Signal Generator (pink noise - 500 to 2000 Hz) at 75 dB I found the Radio Shack meter reads 2.5 dB lower than the CM-140. I consider the CM-140 to be one of the best type 2 SPL meters available and will use it as my reference. I removed the "Sens Factor =0.4938dB, SERNO: 7005156" statement from the Cross Spectrum Labs 0 degree cal file so I could calibrate the REW SPL meter, which I set the same as the CM-140.

What do you experts think, is my method correct?
 

·
Banned
Joined
·
4,838 Posts
If I follow the thread correctly, it sounds like you have two devices that have been professionally calibrated and agree with 1.5 dB (class 2 SPL devices have a +/- 2 dB tolerance, I believe), and one which has not been professionally calibrated and is farther off and can be ignored. While it would seem "nice" to easily get greater absolute SPL accuracy, it is really of little consequence, as has been stated, other than having coincident measurements agree as a matter of convenience. Your choice is a little arbitrary but not unreasonable. You might as well change the sensitivity value in the mic calibration file so it automatically matches your SPL meter and save the trouble of calibrating REW at all. That is what I would do.

But your suggested route will work just as well. The 1.5 dB difference will never bother you in any practical way.
 
1 - 10 of 10 Posts
Top