Home Theater Forum and Systems banner

241 - 260 of 368 Posts

·
Registered
Joined
·
18 Posts
So for the moment I can trust the M30 and since I need to invest in more subwoofers to get meaningful response even down to 10Hz (now playing with two dual 15" sealed cabinets) a further calibration can wait a bit in favor of other needs such as a new SPL.

The calibration of the soundcard is another whole issue since I am using an HDMI out for the output. Yet, I have an Assus Essence STX in the HTPC that I could calibrate and use as the output device for measurements but then I will have to use my Onkyo 5507 crossovers to get sound on the sub instead of Jriver's active ones. While playing back I send the output signal to the amps through the balanced outputs but to calibrate I will have to use the RCA outputs for the loopback cable to the input device (Focusrite in this case). Oh well I will check all possible routes to see if there are discrepancies.

Regarding the criticality of absolute SPL precision, it sure is not important for FR but to calibrate the REW SPL for accuracy in max output measurements isn't it so?

However, how do I apply the correction curve to the Galaxy measurement of the signal REW outputs in order to calibrate its SPL? In other words I am trying to calibrate the REW SPL as close as possible to perfect. If of course it is possible.

Another option would be the purchase of a calibrator device and calibrate the ARTA SPL which also requires other data (I/O voltages, probably a scope and a calibrator device as well). Yet, even more expenses since also ARTA is not free if you want to save data.


Finally, does the calibrator device need calibration as well once in a while? Lol, what a calibration ordeal have I gotten into!
 

·
Premium Member
Joined
·
397 Posts
Hi all.



A) The Earthworks M30 came with a calibration file but only down to 770Hz. Contacting earthworks their engineer told me to get the calibration value form the 775Hz (not there) and copy it down to the desired frequency (10 or 5Hz) at appropriate intervals or leave it blank as the mic is accurate +-0.25d db down to 5Hz.

Is there a point to send the mic for an extra calibration to cross spectrum labs? Would extra information such as polar response be of value?
No. Those mics will be flat to well below 10 Hz. The Earthworks guys know what they're doing,

What a bout placement when measuring? (Vertical, facing)
For low frequency measurements, it won't matter, the Earthworks mics will be truly omnidirectional below 5 kHz.
 

·
Banned
Joined
·
4,838 Posts
If you are using HDMI output, ignore the comment about sound card calibration.

If you really need that kind of absolute accuracy in your SPL measurements, you need a reference SPL meter with guaranteed high absolute accuracy. For most in-home applications that extreme accuracy is simply not needed. If you are doing something commercial, then it might be called for. Of course, if it is important to YOU, then that is what matters. Once Room EQ Wizard is calibrated, working along with a properly calibrated mic, it can provide high accuracy regardless of level or frequency.

An accurate SPL meter, sent in for routine calibration, should satisfy your accuracy needs. When you start buying calibrators, then those have to be calibrated, too. If you are really concerned about calibration drift, I would buy a second SPL meter for crosschecking, then you always have a backup if one gets dropped or broken or is sent in for calibration. When selecting SPL meters, be sure to watch for guaranteed accuracy specs, including drift numbers for temperature.
 

·
Registered
Joined
·
18 Posts
First of all thank you for the clarifications.

The question then comes down to whether the Galaxy CM-140 Verified+ provides that accuracy needed for calibrating REW, serving as a reference point.

So is there a way to apply the correction curve of the Verified + Galaxy to the test tone output by REW during its SPL calibratrion.


It is cheaper for me based in Europe (probably) to buy a Galaxy from Cross Spectrum. Then after a period of time to buy a second one and then sell the first to someone not caring that much and keep on this cycle (replacing one of the two every six months).



P.S. The HDMI output will be replaced in the future with another digital interface (AES/EBU) on a Lynx AES to a Metric Halo ULN-8.
 

·
Premium Member
Joined
·
397 Posts
It is cheaper for me based in Europe (probably) to buy a Galaxy from Cross Spectrum. Then after a period of time to buy a second one and then sell the first to someone not caring that much and keep on this cycle (replacing one of the two every six months).
If you're concerned about SPL calibration, it's probably cheaper to just buy a Type 2/Class 2 calibrator for a couple of hundred Euros. As long as you don't drop it or otherwise abuse it, it likely won't drift more than a few tenths of a dB over a couple of years.
 

·
Banned
Joined
·
4,838 Posts
The question then comes down to whether the Galaxy CM-140 Verified+ provides that accuracy needed for calibrating REW, serving as a reference point.

So is there a way to apply the correction curve of the Verified + Galaxy to the test tone output by REW during its SPL calibratrion.
No, and it is not necessary, nor would it be very useful.

The SPL meter is used to calibrate Room EQ Wizard with pink noise running and the calibration mic as input. Then the calibration mic's correction curve provides measurement accuracy over the entire frequency range. There is no way to put the SPL meter correction curve into REW. You could probably manually fake it into the soundcard calibration file slot, but then you would have to remember to only activated when you were doing a cross check to verify your SPL measurement accuracy at different frequencies. That is a lot of work for very little benefit. Trust REW to do that work for you. It will.

The SPL meter is calibrated to be accurate at 1 kHz reference frequency. SPL meters use RMS sensing measurement algorithms, so an 85 dB SPL sine wave and an 85 dB SPL band of pink noise centered at 1 kHz will give the same result. Calibrate using REW using the speaker cal pink noise setting, and you will have an extremely accurate calibration. Use pink noise when calibrating, not a sine wave, because with a sine wave, even a small separation between the SPL meter and the calibration mic positions can result in error. Use the speaker cal setting because it gives you a band of pink noise between 500 Hz and 2 kHz, where the SPL meter is most accurate. From that point forward, your measurements will be most accurate using REW and your earthworks mic, because the earthworks mic is far more accurate at the low frequencies, the SPL meter will have a large correction factor there. That is the way I would do it.

It is cheaper for me based in Europe (probably) to buy a Galaxy from Cross Spectrum. Then after a period of time to buy a second one and then sell the first to someone not caring that much and keep on this cycle (replacing one of the two every six months).
That's a lot of SPL meters. It is your business, but I am having just a teeny tiny problem understanding why that level of absolute SPL accuracy is necessary, all due respect. Enough said. It is your business, and you may have a very good reason for it.

If you're concerned about SPL calibration, it's probably cheaper to just buy a Type 2/Class 2 calibrator for a couple of hundred Euros. As long as you don't drop it or otherwise abuse it, it likely won't drift more than a few tenths of a dB over a couple of years.
I agree, rather than buying several SPL meters per year, this is probably a better long-term choice.

I truly wish you the very best. Perhaps I went into unnecessary detail, I was just trying to anticipate further questions and get them answered in one shot. Hopefully our responses have been helpful to you, that is our intention.:sn:
 

·
Registered
Joined
·
18 Posts
Thanks a lot for the wise advice. It certainly makes things clear now so I will do as you suggest.

I will buy a calibrator since I know now it does not drift as much as I thought over time and it will be stored under ideal conditions in a dry and "dark" spot with stable temperature.

With that calibrator I might be able to recalibrate my existing SPL meter since it has a calibration pot on it but the company which made it never replied to mails regarding its calibration procedure and I had no idea back then that devices such as a this existed.

With regards to accuracy, I do not expect 100% absolute values since that is truly hard. However, I want to stay as close as possible when measuring max SPL values of output especially in the very low frequencies where it is hard to achieve meaningful output and a 3db difference can be missing the target or not. Not the end of the world but it could mean an extra subwoofer before you hit reference SPL or not considering diminishing returns as well.

Yet, I will have to deal with fine tuning where placement options or specific EQ filters (e.g. LT in sealed cabinets) combined produce increased output down very low but a decrease past a higher point or vice versa. Having the best possible reference point (i.e. +/- 1.5db) when measuring makes things easier while exploring the effects of various factors given specific limitations and/or narrow margins some times and might be a hit or miss case of a detail that a wide measurement variation could otherwise hide.


You are certainly helping a lot here though! Thanks again.



P.S. Any specific calibrator you could suggest and which type? (94 - 114db?)
 

·
Premium Member
Joined
·
397 Posts
P.S. Any specific calibrator you could suggest and which type? (94 - 114db?)
At the Type 2/Class 2 price points, there's not going to be a lot of difference from unit to unit. In terms of 94/114, if you have to choose, I typically recommend choosing a level closest to the level you usually measure at. So if you make measurements at 80 dB, get a 94 dB calibrator (or use the 94 dB setting on calibrators with switchable levels), if you measure at 120 dB, get the 114 dB calibrator. All of my calibrators (2 GenRad 1986's, 4 Larson-Davis CAL200's) are switchable, but I use 94 dB pretty much exclusively.
 

·
Registered
Joined
·
18 Posts
Well after some research the choice for an acoustic calibrator comes down to these 3 models:

This one at 199 Euros (94-11-db)

http://www.content.ibf-acoustic.com/catalog/product_info.php?cPath=31_36&products_id=42

This one at 330 Euros +40 shipping (94db only)

http://www.cirrusresearch.co.uk/acoustic_calibrators/


And this one at 288 Euros (94-114 db)

http://www.thomann.de/gb/galaxy_audio_cm_c200.htm


The Galaxy has some issues with the Earthworks M30 with the adapter that does not fir tight so a custom might be needed.

The Cirus is the most expensive but it is a reliable company and does not have a 114db setting.

The iSEM one is the cheaper and they can send it with a custom adapter (23 Euros) for a perfect fit of the Earthworks M30/.

So it seems I should go for the iSEM one.

What about their reliability and accuracy? Does anyone know about those three and how accurate they are?
 

·
Premium Member
Joined
·
397 Posts
What about their reliability and accuracy? Does anyone know about those three and how accurate they are?
With respect to reputation I would probably rank them 1) Cirrus; 2) Galaxy, and 3) iSEM, but I suspect any of the three will be adequate.
 

·
Registered
Joined
·
18 Posts
I thought so. The iSEM is officially being carried by on the online store of the SMAART developers making me think they have at least tested it as well as their microphones that they also sell.

The thing with the Galaxy is reports from other users of the Earthworks M30 not fitting 100% (loose) but that was a post back from 2008. My M30 came with a 1/2" adapter so maybe it is perfectly fine.

In the end I will flip coins I think! Lol. Nonetheless, for what I want it as long as the iSEM is not totally out of specs and instead of +-0.5db variance it has +-3db then it seems to be the best deal out of the three price wise.
 

·
REW Author
Joined
·
6,711 Posts
No, the factory Dayton file doesn't give a sensitivity value
I think it is in the first line of the Dayton cal files, e.g.

"Sens Factor =-23.555dB, SERNO: 1280303"

and I don't trust measuring it simply because I don't know what settings constitute zero gain.
For these USB mics it is typically the rms value of the data from the mic in dB FS when driven by a 94 dB SPL calibrator with the Windows volume control set to 1.0, but there can be some ambiguity as to whether the rms level of a full scale sine wave is treated as -3 dB FS or 0 dB FS, and the effect of a 1.0 volume setting can be different between Windows XP (which I think the factory uses) and Windows 8. One option could be to simply copy the Dayton line across as supplied so that the files retain the same info.
 

·
Premium Member
Joined
·
397 Posts
I think it is in the first line of the Dayton cal files, e.g.

"Sens Factor =-23.555dB, SERNO: 1280303"
You are correct, I forgot that I deleted that line from the test files I downloaded because it screwed up my import script.

For these USB mics it is typically the rms value of the data from the mic in dB FS when driven by a 94 dB SPL calibrator with the Windows volume control set to 1.0, but there can be some ambiguity as to whether the rms level of a full scale sine wave is treated as -3 dB FS or 0 dB FS, and the effect of a 1.0 volume setting can be different between Windows XP (which I think the factory uses) and Windows 8.
That ambiguity is exactly what prevents me from offering sensitivity measurements for USB mics. There are ways to confirm it, say by sticking the USB mic in the same field as one of my reference mics + meters and seeing if plugging in the factory (or other) sensitivity numbers into ARTA/etc give the same SPL results, but that's enough of a pain to not be worth the effort for me.

One option could be to simply copy the Dayton line across as supplied so that the files retain the same info.
Yeah, but the implication would be that I endorse those factory sensitivity number, which I don't. Not to say they're wrong (all indications are that they're at least pretty close), but if my name is on it, I want to be able to measure it directly.
 

·
Registered
Joined
·
1,298 Posts
For Cross - Spectrum Calibrated Dayton EMM-6 and Behringer ECM8000 microphones with:

0 degrees calibration files
45 degrees calibration files
90 degrees calibration files

Can you describe the senarios (when they should be used) for each of these calibration files?

Which mic orientation should/ideal be used for each senario/file?

Thanks in advance
 

·
Premium Member
Joined
·
397 Posts
For Cross - Spectrum Calibrated Dayton EMM-6 and Behringer ECM8000 microphones with:

0 degrees calibration files
Loudspeaker measurements, microphone pointed directly at speaker.

90 degrees calibration files
Room acoustics/calibration measurements, microphone pointed straight up at ceiling or tilted slightly toward front speakers.

45 degrees calibration files
For experimentation purposes.
 

·
Banned
Joined
·
4,838 Posts
For Cross - Spectrum Calibrated Dayton EMM-6 and Behringer ECM8000 microphones with:

0 degrees calibration files
45 degrees calibration files
90 degrees calibration files

Can you describe the senarios (when they should be used) for each of these calibration files?

Which mic orientation should/ideal be used for each senario/file?

Thanks in advance
My suggestion: For room measurements, pick one and use it for all measurements. Otherwise it becomes difficult to compare measurements taken at different times, under different conditions. For close speaker measurements, always use 0 degrees.
 

·
Premium Member
Joined
·
397 Posts
Effectively immediately, we're no longer selling calibrated ECM8000 mics. We will continue to offer Dayton EMM-6/UMM-6 mics and Galaxy CM-140 sound meters (and very soon, UMIK-1 mics).

To make a long story short: ECM8000 quality has declined to a point where they are utter and I'm sick of messing with them.
 
241 - 260 of 368 Posts
Top