EMM-1 Mic Vs Individually Calibrated Mics

[cawgijoe]

Got my Umik-1 from CSS.

Dumb question, do I keep the extension on the 90 degree file or change it to .txt?

I seem to recall that you just load it.

[ttocs]

I just loaded the .frd file and it worked.

[ttocs]

Jul 31, 2020 10:13:46 GMT -5 fbczar said:

Cross Spectrum Calibration File For UMIK-1 microphone.

I have seen a number of questions relative to the value of the UMIK-1 microphone and the calibration files by Cross Spectrum Labs. I thought the following letter from Herb Singleton of Cross Spectrum Labs would be helpful.

The discussion point I use about our calibration process is as follows:

Cross Spectrum Calibration File For UMIK-1 microphone.

“The primary differences our mics and the factory calibrations are that we calibrate over a wider range of frequencies (5 Hz to 25 kHz compared with the typical factory 10-20 Hz to 20 kHz calibration), and we provide two off-axis calibration data that is helpful for room calibration applications (the factory calibration is on-axis only which works best for applications where the mic is pointed at the source, such as loudspeaker measurements). We also calibrate the low frequency range and high frequency range in separate steps which we believe results in more accurate frequency response curves.

The downside of our calibrations is the process is more labor-intensive which is why our mics cost more.

That said, the factory calibration curves are certainly in the ballpark and are better than nothing.

There is also the question about whether the factory calibrations are done in batches or for individual mics - I'm not sure about the answer to that (I had thought they were done individually, but I'm not sure now), but our calibrations are performed individually for each microphone."

(Quoted from The Official Dirac Live Thread. AudioHTIT properly suggested it would be a good addition to this thread as well. I agree.)
This is an interesting and informative bit of info. Thanks!

I now own 3 mics which are Dirac/Emotiva compatible: EMM-1, an OEM UMIK-1, and a Cross-Spectrum Labs calibrated UMIK-1. I plan to do a comparo to demonstrate the differences between them in the near future. There are differences with the mics. When using Dirac and accepting the standard curve Dirac comes up with, the result when using the EMM-1 will be more low end bass output than either of the UMIK-1 mics. I'll find out more when I do an actual test of all three mics.

It's easy to see that all UMIK-1 mics from miniDSP have individual calibration files, just enter a couple serial numbers and download the files to compare. Here's a couple examples of consecutive serial numbers, one of which is for one of my mics. You will notice that the calibrations go down to 10Hz.


From looking at these calibration files it's apparent that there are differences from mic to mic, hence why individual calibrations are nice to have. Also, when comparing lots of calibration files the biggest differences I found were less than the tolerance of most speakers and probably all subwoofers. So based on this, I'd say that a group calibration is good enough because the variations are largely a "variable offset", which means that the variation is mostly just a sensitivity level adjustment that wanders up/down by maybe 1dB or so from mic to mic.

Questions:
Are these calibrations "accurate"?
Is the calibration method from one company any better than what another company might use?
What does "accurate" really mean?
Is the EMM-1 good enough?
What amount of tolerance is ok?
Should I really care?

Answers:
Probably good enough.
The companies think so. I don't, at least not for my usage.
Within a tolerance. See answer #1 and #5.
Absolutely.
Probably a dB or so, maybe two.
I'm not losing any sleep over this.

There are so many variables when doing in-room measurements that it really should be an exercise based in relativity, with relativity being one measurement result relative to another and tweaking based on the differences.

I'll throw this out there, I contend that when using calibration software the most important issues are phase and time alignment. Frequency response and levels are secondary.

[megash0n]

Jul 31, 2020 12:23:30 GMT -5 ttocs said:

Jul 31, 2020 10:13:46 GMT -5 fbczar said:

Cross Spectrum Calibration File For UMIK-1 microphone.

I have seen a number of questions relative to the value of the UMIK-1 microphone and the calibration files by Cross Spectrum Labs. I thought the following letter from Herb Singleton of Cross Spectrum Labs would be helpful.

The discussion point I use about our calibration process is as follows:

Cross Spectrum Calibration File For UMIK-1 microphone.

“The primary differences our mics and the factory calibrations are that we calibrate over a wider range of frequencies (5 Hz to 25 kHz compared with the typical factory 10-20 Hz to 20 kHz calibration), and we provide two off-axis calibration data that is helpful for room calibration applications (the factory calibration is on-axis only which works best for applications where the mic is pointed at the source, such as loudspeaker measurements). We also calibrate the low frequency range and high frequency range in separate steps which we believe results in more accurate frequency response curves.

The downside of our calibrations is the process is more labor-intensive which is why our mics cost more.

That said, the factory calibration curves are certainly in the ballpark and are better than nothing.

There is also the question about whether the factory calibrations are done in batches or for individual mics - I'm not sure about the answer to that (I had thought they were done individually, but I'm not sure now), but our calibrations are performed individually for each microphone."

(Quoted from The Official Dirac Live Thread. AudioHTIT properly suggested it would be a good addition to this thread as well. I agree.)
This is an interesting and informative bit of info. Thanks!

I now own 3 mics which are Dirac/Emotiva compatible: EMM-1, an OEM UMIK-1, and a Cross-Spectrum Labs calibrated UMIK-1. I plan to do a comparo to demonstrate the differences between them in the near future. There are differences with the mics. When using Dirac and accepting the standard curve Dirac comes up with, the result when using the EMM-1 will be more low end bass output than either of the UMIK-1 mics. I'll find out more when I do an actual test of all three mics.

It's easy to see that all UMIK-1 mics from miniDSP have individual calibration files, just enter a couple serial numbers and download the files to compare. Here's a couple examples of consecutive serial numbers, one of which is for one of my mics. You will notice that the calibrations go down to 10Hz.
View Attachment

From looking at these calibration files it's apparent that there are differences from mic to mic, hence why individual calibrations are nice to have. Also, when comparing lots of calibration files the biggest differences I found were less than the tolerance of most speakers and probably all subwoofers. So based on this, I'd say that a group calibration is good enough because the variations are largely a "variable offset", which means that the variation is mostly just a sensitivity level adjustment that wanders up/down by maybe 1dB or so from mic to mic.

Questions:
Are these calibrations "accurate"?
Is the calibration method from one company any better than what another company might use?
What does "accurate" really mean?
Is the EMM-1 good enough?
What amount of tolerance is ok?
Should I really care?

Answers:
Probably good enough.
The companies think so. I don't, at least not for my usage.
Within a tolerance. See answer #1 and #5.
Absolutely.
Probably a dB or so, maybe two.
I'm not losing any sleep over this.

There are so many variables when doing in-room measurements that it really should be an exercise based in relativity, with relativity being one measurement result relative to another and tweaking based on the differences.

I'll throw this out there, I contend that when using calibration software the most important issues are phase and time alignment. Frequency response and levels are secondary.

the minidsp mics are group calibrated from my understanding. So, the serial number is valid, but the measurement may have been from a batch of 100 mics.

[ttocs]

Jul 31, 2020 12:28:20 GMT -5 megash0n said:

the minidsp mics are group calibrated from my understanding. So, the serial number is valid, but the measurement may have been from a batch of 100 mics.

Sorry, no. They are individually calibrated. Please see my attachment in the post of two consecutive serial numbers. You can do your own search as well and discover that no two calibration files match. For example, use the numbers 706-0100 and 706-0101 and see for yourself.

[megash0n]

Jul 31, 2020 12:37:50 GMT -5 ttocs said:

Jul 31, 2020 12:28:20 GMT -5 megash0n said:

the minidsp mics are group calibrated from my understanding. So, the serial number is valid, but the measurement may have been from a batch of 100 mics.

Sorry, no. They are individually calibrated. Please see my attachment in the post of two consecutive serial numbers. You can do your own search as well and discover that no two calibration files match. For example, use the numbers 706-0100 and 706-0101 and see for yourself.

ok. Everything I read in the past is they were group calibrated. If you wanted one individually calibrated, you had to get one from Spectra something another. Forget the name off the top of my head.
Edit- ok, I see they have updated their verbiage since I bought mine. They clearly state they are individually calibrated now instead of batch calibration. I apologize for being incorrect here. ☺

[ttocs]

Jul 31, 2020 13:54:48 GMT -5 megash0n said:

Jul 31, 2020 12:37:50 GMT -5 ttocs said:

Sorry, no. They are individually calibrated. Please see my attachment in the post of two consecutive serial numbers. You can do your own search as well and discover that no two calibration files match. For example, use the numbers 706-0100 and 706-0101 and see for yourself.

ok. Everything I read in the past is they were group calibrated. If you wanted one individually calibrated, you had to get one from Spectra something another. Forget the name off the top of my head.

Yeah, I was happily surprised that the factory calibration goes down to 10Hz. Going low is helpful. CSL goes down to 5Hz for those who can.

Factory goes to 20kHz, CSL goes to 25kHz. I don't really care about how high the cal goes for a couple reasons. First, if I can't hear it - it doesn't matter. Second, Dirac doesn't need to calibrate that high, and even if someone wants to drag the curve up in the upper freqs, the mic accuracy in that range becomes meaningless.

[bolle]

ttocs: I strongly disagree. Group calibrations are not good enough and might lead to quite audible deltas. Here is a picture of the measurements of 125 mics compared to a pro calibrated mic. If you would use just one middled calibration curve for these, it would lead to deltas in the range of a few dB for most.



An individual calibration doesn´t cost much compared to all of the other equipment most of us have and does make an impact. So for me, it´s a no brainer.

[ttocs]

miniDSP UMIK-1 mics are NOT group calibrated.

[ttocs]

Aug 1, 2020 5:31:04 GMT -5 bolle said:

ttocs : I strongly disagree. Group calibrations are not good enough and might lead to quite audible deltas. Here is a picture of the measurements of 125 mics compared to a pro calibrated mic. If you would use just one middled calibration curve for these, it would lead to deltas in the range of a few dB for most.

An individual calibration doesn´t cost much compared to all of the other equipment most of us have and does make an impact. So for me, it´s a no brainer.

Yep, a few dB. Just like a lot of speakers. I'm only suggesting that this is not an exact science when a mic - individually calibrated or not - is used in my great room which is not an anechoic chamber. 

For most people this shouldn't be something to fuss over. If the EMM-1 is as close in tolerance as the graph you provide, then that few dB is good enough.

But, if you're paying separately for a mic, then you're right, it's not good enough - except, what's the tolerance? miniDSP doesn't state a tolerance for their individually calibrated mics. CSL doesn't state a tolerance for their individually calibrated mics either. Do I care? No. I'm not building speakers, just using my three mics in my room with one set of speakers for a long time. If I don't like the tone, then I'll change it. Measure, load filter, listen, like or not like, view the graph and decide which way to go to change the way I expect and measure again.

Think about it. Those curves in that graph are what old school Bass and Treble tone controls used to alter. We are simply using all this stuff as really complicated tone controls, with the addition to the phase and timing.

[consul]

Hello,
anyone tried Beyerdynamic MM1 ?
www.beyerdynamic.de/mm-1.html

Based at the series number, the manufacturer provides an individual calibrating file..

Regards,
Alexander

[AudioHTIT]

Aug 3, 2020 8:38:19 GMT -5 consul said:

Hello,
anyone tried Beyerdynamic MM1 ?
www.beyerdynamic.de/mm-1.html

Based at the series number, the manufacturer provides an individual calibrating file..

Regards,
Alexander

Can’t find the detailed specs, but it looks like it’s phantom powered instead of USB, this would make it a little more difficult to interface with a laptop or computer. It might be possible to use the EMM-1’s digitizing cable with this, I’ll try to test that idea today with a generic condenser mic. They also don’t specify what type of calibration files are included (0 and 90?), and what the range and tolerance are. That said, Beyer makes excellent mics so this should be good, but at roughly double the price of a standard UMIK-1 and still substantially higher then a CSL UMIK, it better be pretty special. Too many unanswered questions for me to recommend.

[consul]

Aug 3, 2020 11:06:57 GMT -5 AudioHTIT said:

Can’t find the detailed specs, but it looks like it’s phantom powered instead of USB, this would make it a little more difficult to interface with a laptop or computer. It might be possible to use the EMM-1’s digitizing cable with this, I’ll try to test that idea today with a generic condenser mic. They also don’t specify what type of calibration files are included (0 and 90?), and what the range and tolerance are. That said, Beyer makes excellent mics so this should be good, but at roughly double the price of a standard UMIK-1 and still substantially higher then a CSL UMIK, it better be pretty special. Too many unanswered questions for me to recommend.

The price in Germany:
- miniDSP UMIK-1 USB: EUR 110-120
- B-Stock MM1: EUR 135-140

[AudioHTIT]

Aug 3, 2020 11:21:02 GMT -5 consul said:

Aug 3, 2020 11:06:57 GMT -5 AudioHTIT said:

Can’t find the detailed specs, but it looks like it’s phantom powered instead of USB, this would make it a little more difficult to interface with a laptop or computer. It might be possible to use the EMM-1’s digitizing cable with this, I’ll try to test that idea today with a generic condenser mic. They also don’t specify what type of calibration files are included (0 and 90?), and what the range and tolerance are. That said, Beyer makes excellent mics so this should be good, but at roughly double the price of a standard UMIK-1 and still substantially higher then a CSL UMIK, it better be pretty special. Too many unanswered questions for me to recommend.

The price in Germany:
- miniDSP UMIK-1 USB: EUR 110-120
- B-Stock MM1: EUR 135-140

Certainly buying in country makes sense. The other questions remain, and I can now (unscientifically) confirm that a generic phantom powered condenser mic does not work with the EMM-1 cable. So if the Beyer is phantom powered, the biggest issue would be how to get its input into the laptop. I would try something like the Focusrite 2i2 (as I own one), but that roughly doubles to cost. If you have another method, the Beyer may be a good choice.

[ttocs]

I still have a lot to learn about REW and how to measure subwoofers. In furthering my career of beating my head against rocks, I came across something today that is still a work in progress but shows a direct comparison of the EMM-1 and UMIK-1 microphones.

I've been testing maximum output of my subs using one, two, and three sub combos and discovering what affects what, and how (as the Three Stooges used to say). I was questioning why it is that there's a 94dB brick wall when using the UMIK-1. At first, I suspected my subs just can't push any more air, but after testing this week, that just didn't seem correct. I find all sorts of videos showing output in large rooms easily surpassing 100dB. So with 3 subs, why can't I experience something similar albeit at a reduced SPL due to the 5400 cubic feet of air in my space?

Well, it has everything to do with REW, and the UMIK mic with it's calibration file.

When I use the UMIK from miniDSP, REW starts throwing clipping errors at around 93dB. When I use the UMIK from Cross-Spectrum Labs the same thing happens. When I use the EMM-1, which REW will allow me to calibrate using the calibration tool I have, I get up to 100dB.

So then I discovered that there's a process that can be used to get the UMIK and REW to be able to measure above 94dB. BTW, why is the magic number 94dB? Because that's one of the calibration points that mic calibration tools use, the other is 114dB. I'll check more into this process for getting the UMIK to measure higher, but for now I'll just use the EMM-1. When using the SPL meter function in REW it will tell the max SPL the calibration is good for. Like I said, that max for the UMIK is 94dB, but for the EMM-1 it's 105 without doing anything extra other than loading the calibration file and running the calibration in SPL Meter in REW.

Here's a comparison of one of the UMIK's and the EMM. Note two things. First, the frequency response is very similar and based solely on this would not make me want to spend the extra money on the UMIK. Second, the EMM allows the measurement to be 7dB louder without clipping. These measurements were done at a lower level so I could directly compare them and not have to deal with possible clipping errors distorting the results. Also, I've been moving my subs around testing different things, so all I did was shove the stack of subs (not the Sack-O-Suds in the movie My Cousin Vinny) against the wall and test.

[ttocs]

I found an official answer regarding the SPL limit on the UMIK mic. According to miniDSP the UMIK-1 is setup at the factory for 18dB gain for "mostly" lower SPL measurements. There's a dip switch inside that is user adjustable for other settings.

From miniDSP: "Latest versions are indeed with 18dB gain as per the general request to increase the sensitivity for Dirac software and other measurement software. It will work even better at lower levels. You can always change the gain if you want to listen @ higher dB levels by opening the ring and changing the internal DIP switch."

So I'm going to make the adjustment on the miniDSP UMIK mic and leave the CSL mic the way it is. FWIW all three mics are so close on measurements that it's a toss up for me. I could be blindfolded and select one of the mics out of a hat and be happy.

[LCSeminole]

Sept 26, 2020 18:48:46 GMT -5 LCSeminole said:

Sept 26, 2020 17:59:23 GMT -5 ttocs said:

I found an official answer regarding the SPL limit on the UMIK mic. According to miniDSP the UMIK-1 is setup at the factory for 18dB gain for "mostly" lower SPL measurements. There's a dip switch inside that is user adjustable for other settings.

From miniDSP: "Latest versions are indeed with 18dB gain as per the general request to increase the sensitivity for Dirac software and other measurement software. It will work even better at lower levels. You can always change the gain if you want to listen @ higher dB levels by opening the ring and changing the internal DIP switch."

So I'm going to make the adjustment on the miniDSP UMIK mic and leave the CSL mic the way it is. FWIW all three mics are so close on measurements that it's a toss up for me. I could be blindfolded and select one of the mics out of a hat and be happy.

So have you done this yet?.....if so I too am curious as to what the DIP switches should be on for the 18dB gain. I've opened my UMIK-1 and see 3 DIP switches, switch 1 is "ON" or in the up position, switch 2 & 3 are "down" or I assume the "OFF" position. What combination gives the 18dB gain?

Did a little digging myself and have figured out by the table on the back of the PCB board in the UMIK-1 that to attain an 18db gain the "three" DIP switches are either set to "0" or "1". 0 being in the down position and 1 being in the up position. My UMIK-1 came preset to 12db, so I too will be giving 18db a try. I also just read in several posts(by John Mulcahy the REW software writer)on the AV Nirvana, HomeTheaterShack and miniDSP forums, that DIRAC suggests the 18db setting.

First # in the table below is "DIP SWITCH #1", Second # in the table below is "DIP SWITCH" #2, Third # in the table below is "DIP SWITCH #3". So for example, my UMIK-1 came from CSL with the 12db setting, (ie: 1 0 0)

To gain access to the PCB board, there are two small pin-holes on the bottom end of the UMIK-1. I simply used a push pin and put in one of the holes and rotated counter-clockwise, and the "Ring" that ttocs described above will come out. Now simply and gently pull on the inside edges where the "Ring" was and pull out the PCB board(remembering that there are a few wires attached to the other end of the PCB board) far enough to expose the DIP switches on one side and you'll see the table that I copied below on the other side of the PCB board. I used the push pin, that I unscrewed the Ring with, to change the DIP switch settings. Now simply do everything in reverse to put it back together. Now my UMIK-1 is set to (0 1 1), ie: 18db. I bought my UMIK-1 from CSL right before I received my XMC-1(6 or 7 years now, so this explains why my UMIK-1 was set to 12db, and why miniDSP is now setting them by default from around 2017-2018.

1 1 1 = 0db
1 1 0 = 0db
1 0 1 = 6db
1 0 0 = 12db
0 1 1 = 18db
0 1 0 = 24db
0 0 1 = 30db
0 0 0 = 36db


..............just for conversation sakes, I had my eyes dilated today at the ophthalmologist so I could get prescription reading glasses. Meaning if I can do this gain setting change with temporary blurry vision, then just about anyone can do it.

[megash0n]

Sept 26, 2020 19:07:30 GMT -5 LCSeminole said:

Sept 26, 2020 18:48:46 GMT -5 LCSeminole said:

So have you done this yet?.....if so I too am curious as to what the DIP switches should be on for the 18dB gain. I've opened my UMIK-1 and see 3 DIP switches, switch 1 is "ON" or in the up position, switch 2 & 3 are "down" or I assume the "OFF" position. What combination gives the 18dB gain?

Did a little digging myself and have figured out by the table on the back of the PCB board in the UMIK-1 that to attain an 18db gain the DIP switches are either set to "0" or "1". 0 being in the down position and 1 being in the up position. My UMIK-1 came preset to 12db, so I too will be giving 18db a try. I also just read in several posts(by John Mulcahy the REW software writer)on the AV Nirvana, HomeTheaterShack and miniDSP, that DIRAC suggests the 18db setting.

First # in the table below is "DIP SWITCH #1", Second # in the table below is "DIP SWITCH" #2, Third # in the table below is "DIP SWITCH #3". So for example, my UMIK-1 came from CSL with the 12db setting, (ie: 1 0 0)

To gain access to the PCB board, there are two small pin-holes on the bottom end of the UMIK-1. I simply used a push pin and put in one of the holes and rotated counter-clockwise, and the "Ring" that ttocs described above will come out. Now simply and gently pull on the inside edges where the "Ring" was and pull out the PCB board(remembering that there are a few wires attached to the other end of the PCB board) far enough to expose the DIP switches on one side and you'll see the table that I copied below on the other side of the PCB board. I used the push pin, that I unscrewed the Ring with, to change the DIP switch settings. Now simply do everything in reverse to put it back together. Now my UMIK-1 is set to (0 1 1), ie: 18db. I bought my UMIK-1 from CSL right before I received my XMC-1(6 or 7 years now, so this explains why my UMIK-1 was set to 12db, and why miniDSP is now setting them by default from around 2017-2018.

1 1 1 = 0db
1 1 0 = 0db
1 0 1 = 6db
1 0 0 = 12db
0 1 1 = 18db
0 1 0 = 24db
0 0 1 = 30db
0 0 0 = 36db

interesting. When using the calibration file... It says 18dB gain. Is this just in the file, or does it understand how these dip switches are set?
Sorry if this was mentioned already and I skimmed over it.

[LCSeminole]

Sept 26, 2020 19:29:33 GMT -5 megash0n said:

Sept 26, 2020 19:07:30 GMT -5 LCSeminole said:

Did a little digging myself and have figured out by the table on the back of the PCB board in the UMIK-1 that to attain an 18db gain the DIP switches are either set to "0" or "1". 0 being in the down position and 1 being in the up position. My UMIK-1 came preset to 12db, so I too will be giving 18db a try. I also just read in several posts(by John Mulcahy the REW software writer)on the AV Nirvana, HomeTheaterShack and miniDSP, that DIRAC suggests the 18db setting.

First # in the table below is "DIP SWITCH #1", Second # in the table below is "DIP SWITCH" #2, Third # in the table below is "DIP SWITCH #3". So for example, my UMIK-1 came from CSL with the 12db setting, (ie: 1 0 0)

To gain access to the PCB board, there are two small pin-holes on the bottom end of the UMIK-1. I simply used a push pin and put in one of the holes and rotated counter-clockwise, and the "Ring" that ttocs described above will come out. Now simply and gently pull on the inside edges where the "Ring" was and pull out the PCB board(remembering that there are a few wires attached to the other end of the PCB board) far enough to expose the DIP switches on one side and you'll see the table that I copied below on the other side of the PCB board. I used the push pin, that I unscrewed the Ring with, to change the DIP switch settings. Now simply do everything in reverse to put it back together. Now my UMIK-1 is set to (0 1 1), ie: 18db. I bought my UMIK-1 from CSL right before I received my XMC-1(6 or 7 years now, so this explains why my UMIK-1 was set to 12db, and why miniDSP is now setting them by default from around 2017-2018.

1 1 1 = 0db
1 1 0 = 0db
1 0 1 = 6db
1 0 0 = 12db
0 1 1 = 18db
0 1 0 = 24db
0 0 1 = 30db
0 0 0 = 36db

interesting. When using the calibration file... It says 18dB gain. Is this just in the file, or does it understand how these dip switches are set?
Sorry if this was mentioned already and I skimmed over it.

Not sure if this is in the file or not(I'm guessing probably not), since the actual PCB board has the three DIP switches to change the gain settings, thus I would think it would just be going by how the switches are set.

[megash0n]

Sept 26, 2020 19:34:03 GMT -5 LCSeminole said:

Sept 26, 2020 19:29:33 GMT -5 megash0n said:

interesting. When using the calibration file... It says 18dB gain. Is this just in the file, or does it understand how these dip switches are set?
Sorry if this was mentioned already and I skimmed over it.

Not sure if this is in the file or not(I'm guessing probably not), since the actual PCB board has the three DIP switches to change the gain settings, thus I would think it would just be going by how the switches are set.

thank you sir. I'm curious to see where these tests go. May or may not be of any benefit, but it's interesting nonetheless.
ttocs, are you saying the mic won't read above 94dB, or that it clips? I'm pretty sure I've measured ab