• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Mic Calibration for Rode NT1A

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,422
Likes
4,029
Location
Pacific Northwest
I have a pair of Rode NT1A mics that I use for recording local acoustic music events. These mics have lower noise and distortion than the UMIK-1, but they have non-flat frequency response and they don't have individual calibration files. I created detailed calibration files for these mics, so I can now use them for room tuning. I have a pair and each measures slightly differently, so I created individual files for each. If you want a generic NT1A calibration curve just take the average.

NT1A Calibration Mic 1
NT1A Calibration Mic 2

How I created these: I measured my in-room system using Room EQ Wizard, at the same position & level with all 3 mics (using the UMIK-1's calibration file). I took 2 runs of each and averaged them, for consistency. I exported the response data from REW and differenced them in a spreadsheet, so if you apply the given deltas to the NT1A, their response exactly matches the UMIK-1. For example, in the files the value is about 9 dB at 60 Hz. This means the NT1A's response is +9 dB at 60 Hz, so you would subtract 9 dB to get "neutral".

That said, I'm not entirely sure of the results. Applying these corrections to the NT1A, their curves match exactly to the UMIK-1 in REW. That sanity-checks the calibration process. However, the NT1A response curves aren't even close to what Rode publishes for these mics, and it has a suspiciously large peak around 75 Hz, which makes me question the UMIK-1's accuracy or calibration in this range. Here's the resulting NT1A calibration, relative to the calibrated response of the UMIK-1.
1588441422951.png

Any tips to help resolve this suspicion?
 

NTK

Major Contributor
Forum Donor
Joined
Aug 11, 2019
Messages
2,656
Likes
5,819
Location
US East
The directional patterns of the mics are different. The Rode NT1-A are cardioid, and the UMIK-1 is omni. The UMIK-1 picks up sound that the Rode doesn't.
 
OP
MRC01

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,422
Likes
4,029
Location
Pacific Northwest
Ah you mean reflected sound (arriving at a different angle). So the more reflective the room, the more different the mics will measure.
At the anomaly frequency (around 75 Hz), the NT1A response was stronger than the UMIK1. So it's possible the UMIK1 was picking up a reflected frequency (arriving from a different angle) that was phase shifted, reducing 75 Hz.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,521
Likes
37,050
So your graph shows the two Rodes? How does this compare to the Umik 1? I'm assuming you measuring a speaker so is the 75 hz hump just from the speaker or are you saying these graphs show the difference vs the Umik 1 measure?
 
OP
MRC01

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,422
Likes
4,029
Location
Pacific Northwest
These graphs show the NT1A response relative to the UMIK1. Red & Blue are the 2 individual NT1A mics. Y=0 is the UMIK1 response. The graph shows mic differences, not room response.
 

NTK

Major Contributor
Forum Donor
Joined
Aug 11, 2019
Messages
2,656
Likes
5,819
Location
US East
Yes. The mics pick up the sound that came in from all different directions, and output the complex sum (i.e. including the phase differences) as an electrical signal. Since the directivities of the mics are different, you should expect the sums to be different. I think it is probably quite difficult to deconvolute the effects to make sense of the differences.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,521
Likes
37,050
Try this one I put together from published FR specs. It opens in to a .csv file which REW can use for cal files.

I've done this for several microphones. I find the response between two different mics of the same model closer than yours. And the results vs UMIK1 generally quite close, but not identical. Your results don't look right to me. Though the procedure you describe should have worked fine.

I would say you can take the response of the UMIk 1 and also pull up an un-calibrated result from your mic. In the window choose A/B or A divided by B and it will give you the difference between the two. That might let you make a better file using the difference between the two.
 

Attachments

  • Rode NT1a.zip
    821 bytes · Views: 949

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,521
Likes
37,050
Here is where I measured a speaker first with the Umik-1 and then with a cardioid Shure KSM32 using a file developed from published FR of the Shure mic. The up and down nature is the speaker response. While not spot on it is reasonably close. With cardioid's you typically see the response droop at the lower end starting around 200 hz like seen here. Red is the Umik-1 and green is the KSM32 Shure.
1588444292727.png
 
Last edited:

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,521
Likes
37,050
Here are my two KSM32's vs each other using the same cal file for each. One is an early version with a slightly different grill. The other was made almost a decade later. Again both measuring the same loudspeaker which is where the up and down variance is coming from.

1588444656030.png


Here are three Audio Technica AT4033's spanning several years. Same cal file for all. Not perfect, but pretty close.
1588444868729.png
 
OP
MRC01

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,422
Likes
4,029
Location
Pacific Northwest
Here's a second run, based on measurements taken on another day. Your curve shown in yellow. Not as extreme as the first, and closer to your curve.
1588449709695.png

But still that spike at 75 Hz doesn't make sense. It must be the UMIK-1 picking up an out-of-phase wave at 75 Hz that weakens the response, and it's a reflection coming from a different direction, so the NT1A isn't picking it up.

This leads to a perceptual question: is our hearing directional like a cardiod? Or is it more omnidirectional? I would guess omni, since that's the kind of mic used to measure room response.
 

Attachments

  • 1588449582746.png
    1588449582746.png
    15.5 KB · Views: 274
OP
MRC01

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,422
Likes
4,029
Location
Pacific Northwest
Here's how these 3 mics compare, with UMIK1 corrected, and the NT1As uncorrected. In-room, listener position, no EQ.
umik1-nt1a-comp-raw.png

Here's how they compare after I create and apply the relative correction curves for the NT1As:
umik1-nt1a-comp-corr.png
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,521
Likes
37,050
Here's a second run, based on measurements taken on another day. Your curve shown in yellow. Not as extreme as the first, and closer to your curve.
View attachment 61508
But still that spike at 75 Hz doesn't make sense. It must be the UMIK-1 picking up an out-of-phase wave at 75 Hz that weakens the response, and it's a reflection coming from a different direction, so the NT1A isn't picking it up.

This leads to a perceptual question: is our hearing directional like a cardiod? Or is it more omnidirectional? I would guess omni, since that's the kind of mic used to measure room response.

It could be noise. Sometimes hard to ferret out in the below 100 hz frequencies. I've seen such things AC on vs AC off because some long duct was resonating when the AC is on and feeding it into a room.

Here is a thought. Try measuring with the Rode facing the speaker, then turn it around and measure again. If you are careful placing the diaphragm at exactly the same place when you turn it around 180 degrees, the two should combine for an omni response.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,521
Likes
37,050
Here's how these 3 mics compare, with UMIK1 corrected, and the NT1As uncorrected. In-room, listener position, no EQ.
View attachment 61514
Here's how they compare after I create and apply the relative correction curves for the NT1As:
View attachment 61515
That looks good, and the Umik is the one with the strange result having that dip at 75 hz. Does 75 hz fit as a room cancellation for any of your room's long dimensions. It would be in the ballpark for 8ft ceiling height.
 
OP
MRC01

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,422
Likes
4,029
Location
Pacific Northwest
I eliminated noise when taking the measurements, but flipping the mics is a good test.

I believe the 74 Hz phenomena is LBIR from the back wall. The distance to back wall is 44 inches, which is 1/4 wavelength at 74 Hz, which makes total distance traveled 1/2 wavelength, which makes a null. I placed a few Scopus T70s behind the listener position and it weakened the null by 6 dB.

So, it looks like the NT1A isn't picking up that null because it doesn't respond to the back wave reflecting forward.

PS: the above curves were measured with the room treatments removed.
 

Steezyryder

New Member
Joined
May 9, 2021
Messages
3
Likes
0
I’m not sure a directional microphone like the Rode is good for HT speaker calibration. If you were to use it, then probably pointing the front of the mic (with the dot) upwards would be the way to do it.
that said, since I do not have a mini dsp, nor do I want to buy one for $100, I plan to use my rode nt1-a with my audio interface and laptop to perform a Dirac Live room calibration with my NAD T787 AVR with upgraded atmos sound card for a 7.2.4 home theater.
Thanks for posting you calibration files. Which one do you think will work best for my situation? I though Rode might have one posted but I can’t find it in their site anywhere... I read something about an anniversary post with a cal file but all I see is a graph.
 
OP
MRC01

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,422
Likes
4,029
Location
Pacific Northwest
I’m not sure a directional microphone like the Rode is good for HT speaker calibration. If you were to use it, then probably pointing the front of the mic (with the dot) upwards would be the way to do it. ...
From what I read, human hearing is not omnidirectional, but more cardiod shaped. Of course it is more cardiod at high frequencies, approaching omni at low frequencies. Which is not unlike the response of cardiod mics like the NT1A. Also, the NT1A has lower distortion than the UMIK1 and the difference shows in measurements, if your speakers have low enough distortion.

Example of omni vs. cardiod: post 14 above. My room had a dip at 74 Hz which was caused by a reflection from the wall behind the listener. The UMIK1 (being omni) showed it as a steep null, but playing tones through that range, perceptually the dip was there but more slight, wasn't nearly as deep as the UMIK1 showed it to be (human hearing isn't fully omnidirectional, the reflected wave, coming from behind, is perceived at a lower level than the incoming front wave). The NT1A showed less of a dip and matched closer to my perception. This makes me wonder why we use omnidirectional mics for room tuning when our hearing isn't omnidirectional. If the mic's response pattern doesn't match our hearing, it will in some cases show a response that doesn't "reflect" :) what we hear.

Thanks for posting you calibration files. Which one do you think will work best for my situation? I though Rode might have one posted but I can’t find it in their site anywhere... I read something about an anniversary post with a cal file but all I see is a graph. ...
Anyway, the calibration files I made by comparing with the UMIK1 seem "overfitted". So I made this one, which is simplified and smoothed. It's based on a mix of measuring against a known calibrated mic, reference to the Rode frequency response curve, and simplifying based on subjective perception. So it could be incorrect, use with a grain of salt.
NT1A mic calibration
 

Steezyryder

New Member
Joined
May 9, 2021
Messages
3
Likes
0
Thanks for the reply. I agree that our hearing is not omnidirectional (mainly due to the shape of our ears), so what you’re saying makes sense. I thought maybe the EQ manufacturers took that into account due to measurements at various seating arrangements? But perhaps that isn’t possible for the mic to discern, not knowing which direction it came from, but merely how loud it is and how long it took to get there... I thought that the various positions would help it recognize direction though... Which brings me back to the loudness part. How did you orientate the microphone? The high frequency response on an nt1-a is very much in the front and center. I would think there is a dramatic calibration difference. Since the calibration software states that when calibrating stereo only, one should point the microphone toward the speakers, and upwards when calibrating for home theater; I think the nt1-a would be best with the dot facing up. Perhaps you would not lose that signal at 7500 Hz? Or perhaps that would negate the correlation to our directional hearing?

All this lead me to order a Dayton emm-6 right after my first post for my upcoming Dirac Live 7.2.4 atmos room calibration. Looks like I’ll be doing 3 calibrations to see what I like better. Emm-6, nt1-a forward, ndt1-a upwards....

My AVR has very good signal to noise ratio of 93 dB with total harmonic distortion of <0.08 dB. That said, most professional testing rated my NAD T787 on the bright side of the high spectrum. I feel like the nt1-a calibration will miss more of the highs resulting in an even brighter sound. We shall see. I’ll update when I get it back from the module upgrades at the dealer. Fortunately, the auddesey calibration is still stored in the AVR (although it’s not able to run auddesey calibration after the upgrade) so I will really be able to do a good A/B comparison between these calibrations.

Thanks again for you response and cal file.
 

Steezyryder

New Member
Joined
May 9, 2021
Messages
3
Likes
0
From what I read, human hearing is not omnidirectional, but more cardiod shaped. Of course it is more cardiod at high frequencies, approaching omni at low frequencies. Which is not unlike the response of cardiod mics like the NT1A. Also, the NT1A has lower distortion than the UMIK1 and the difference shows in measurements, if your speakers have low enough distortion.

Example of omni vs. cardiod: post 14 above. My room had a dip at 74 Hz which was caused by a reflection from the wall behind the listener. The UMIK1 (being omni) showed it as a steep null, but playing tones through that range, perceptually the dip was there but more slight, wasn't nearly as deep as the UMIK1 showed it to be (human hearing isn't fully omnidirectional, the reflected wave, coming from behind, is perceived at a lower level than the incoming front wave). The NT1A showed less of a dip and matched closer to my perception. This makes me wonder why we use omnidirectional mics for room tuning when our hearing isn't omnidirectional. If the mic's response pattern doesn't match our hearing, it will in some cases show a response that doesn't "reflect" :) what we hear.


Anyway, the calibration files I made by comparing with the UMIK1 seem "overfitted". So I made this one, which is simplified and smoothed. It's based on a mix of measuring against a known calibrated mic, reference to the Rode frequency response curve, and simplifying based on subjective perception. So it could be incorrect, use with a grain of salt.
NT1A mic calibration
So I finally got my avr hooked up and started with the omnidirectional Dayton emm-6. Well that took about 3 hours. Honestly I wasn’t immediately pleased with the fully flat calibration across all frequencies. So I adjusted the curtain to only correct up to 400 Hz on my floor level speakers and left the atmos ceiling speakers flat. After listening. Both of them are growing on me actually. It was way too much work to do again with a different microphone. I was looking for some amazing change, but it was already calibrated with audyssey before and that was a huge improvement over uncalibrated. Anyway. Now I’m looking for new speakers lol. These old pioneers are crap
 
OP
MRC01

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,422
Likes
4,029
Location
Pacific Northwest
... I wasn’t immediately pleased with the fully flat calibration across all frequencies. ...
Me neither. My current room sounds best with a bass tilt, a gradual drop of about 12 dB over several octaves to the treble. That's a bit more than the typical 1 dB / octave, but every room is different.
 
Top Bottom