Is it possible to measure Infrasonics with a UMIK-I?

[Plompudu]

I've added a SVS SB-2000 to my setup (2x Swissonic A306 + SB-1000) and I'm wondering how to properly measure the Infrasonic response of my system.
The calibration file of my UMIK-I goes down to "only" 10Hz and the measured response via a RTA average looks like this before any EQ is applied:

After EQ (Left Channel/ Right Channel / Both Channels / Both RTA):

(Issue at 70Hz is not yet fixable since I can't EQ and time/phase align the Subs separately yet)

The Roll-off of the Calibration file is 8dB/oct.

Is another solution needed or is a UMIK-I enough?
Thanks in advance!

 

[Multicore]

My UMIK-1 calibration file goes down to

4.33 -3.34 0

So, yes, it is possible, if 4.33 Hz is low enough for you.

What, btw, do you want to shake the air at 4 Hz for?

 

[Oddball]

Do you have any measurements at higher SPL? For some music and definitively HT, low end SPL of 100+dB (for peaks) seems to be desirable.

 

[Plompudu]

My UMIK-1 calibration file goes down to

4.33 -3.34 0

So, yes, it is possible, if 4.33 Hz is low enough for you.

What, btw, do you want to shake the air at 4 Hz for?

Ah so I just got one that isn't calibrated for infrasonics.

Is it fair to assume that the roll-off continues with the 8dB/oct like in my case?
Can you send your calibration for comparison?

 

[Multicore]

Is it fair to assume that the roll-off continues with the 8dB/oct like in my case?

Who knows?

What harm could be done if you did so assume?

Can you send your calibration for comparison?

See DM.

 

[Oddball]

Ah so I just got one that isn't calibrated for infrasonics.

Is it fair to assume that the roll-off continues with the 8dB/oct like in my case?
Can you send your calibration for comparison?

I dont think think you will really suffer from potential UMIK-1 discrepancies. But then again 80dB is not really much even if in 80hz range and much less so in infra range? Sub SPL is 115db for HT.

I would like to see anyone that would really tell if you have 8hz amplitude properly aligned. As to the phase, this will likely cause all kind of issues. Can’t even imagine as did not they.

 

[Plompudu]

Do you have any measurements at higher SPL? For some music and definitively HT, low end SPL of 100+dB (for peaks) seems to be desirable.

I just did with my mic at roughly the the MLP (roghly ~15cm off), so that's why the FR looks worse.

But i had to stop at the 100dB sweep cause it hit too high of a distortion at 18-30Hz + slight compression and moved so much air that it knocked the mic of the seat haha. (so more than I personally need for continuous loudness)

The SB-1000 is likely the limiting factor, but it seems like it's not worth testing for my usage.

Compression Sweeps:

Brown looks weird in the higher frequencies, cause it dropped the mic.

Distortion for the 85dB Sweep:

Distortion for the 90dB Sweep:

Distortion for the 95dB Sweep:

Distortion for the 100dB Sweep:

The EQ that I use:

If I EQ the 18-30Hz range down I'll likely be able to hit noticeably higher levels, but having a flat FR is more worth it imo.

But with >40Hz being below 2% THD even at 100dB I'm totally fine for pretty much all music I throw at it during casual listening, i would guess?

Would you recommend creating a separate EQ for loud levels that doesn't try to flatten the 18-30Hz Range?
Would you go for a Low Shelf below 35Hz or how would you approach it?

I also attached the compression test + FR of "L+R After Calibration" + RTA.

 

[Oddball]

Seems like you know well what you are doing. Peaks beyond 100dB are natural for both HT and music if listening at near reference levels and would think that would be one of the points of design. I would prioritise response in the >20hz area as that seems to matter most.

Not a big fan of infrasonic though and roll off the subs with pretty much a brick wall after 15hz

 

[Roland68]

I've added a SVS SB-2000 to my setup (2x Swissonic A306 + SB-1000) and I'm wondering how to properly measure the Infrasonic response of my system.
The calibration file of my UMIK-I goes down to "only" 10Hz and the measured response via a RTA average looks like this before any EQ is applied:

View attachment 438202

After EQ (Left Channel/ Right Channel / Both Channels / Both RTA):
View attachment 438201
(Issue at 70Hz is not yet fixable since I can't EQ and time/phase align the Subs separately yet)

The Roll-off of the Calibration file is 8dB/oct.

Is another solution needed or is a UMIK-I enough?
Thanks in advance!

A quick note on infrasound measurements. The problem is the ratio of frequency to amplitude; at 1 Hz, the wavelength is 340 m.
From 5 Hz, you should be able to measure easily with your microphone if the sound level is greater than 20 dB.

For lower sound levels, or below 5 Hz, you need a measuring device with microbarometers.

 

[Plompudu]

My UMIK-1 calibration file goes down to

4.33 -3.34 0

So, yes, it is possible, if 4.33 Hz is low enough for you.

What, btw, do you want to shake the air at 4 Hz for?

Ohh i totally forgot to answer the second part - Main Reason was honestly just a the beginning of show that i first heard with headphones and i was already like "damn that's some deep Bass I wonder how it sounds like with a good Infrasonic Subwoofer". All the other advantages of adding the Sub also made it worthwhile

Kaiju-no.8 EP1 beginning SPL log at -13dB:

RTA peak:

The Full infrasonic response is taken advantage of in this show and the mix is absurd, but quite enjoyable, haha

 

[Multicore]

Ohh i totally forgot to answer the second part - Main Reason was honestly just a the beginning of show that i first heard with headphones and i was already like "damn that's some deep Bass I wonder how it sounds like with a good Infrasonic Subwoofer". All the other advantages of adding the Sub also made it worthwhile

Kaiju-no.8 EP1 beginning SPL log at -13dB:
View attachment 438460

RTA peak:
View attachment 438461

The Full infrasonic response is taken advantage of in this show and the mix is absurd, but quite enjoyable, haha

What do the various lines on the first graph represent?

 

[NTK]

What do the various lines on the first graph represent?

It is from the REW SPL Logger screen. Below are the plot legends if using the default trace colors.

 

[Plompudu]

What do the various lines on the first graph represent?

I accidentally didn't screenshot the labels, my bad

Here is it with the Labels below:

 

[GXAlan]

Is another solution needed or is a UMIK-I enough?

Did you adjust the mic preamp? That is often the limiting factor for the UMIK-1 and you will want to bring that down to test for high SPL. As you get infrasonic, it is also more susceptible to noise, which affects the accuracy of those infrasonic measurements

 

[lucian]

so most important question: WHICH show?

 

[Salt]

Some sophisticated developers implement a 20 or 30 Hz highpass in active driven implementations.

 

[Plompudu]

so most important question: WHICH show?

You mean the name of the show?
It's called Kaiju No. 8
Basically a Anime about Japan if "Godzilla" was happening every other week and the MC being

Spoiler: Spoiler EP1

turned into a Monster

, while also wanting to join the team that fights them.

Pretty simple show, but pretty impressive Audio/Bass imo.

 

[Plompudu]

Did you adjust the mic preamp? That is often the limiting factor for the UMIK-1 and you will want to bring that down to test for high SPL. As you get infrasonic, it is also more susceptible to noise, which affects the accuracy of those infrasonic measurements

Thanks for the input!
No I use the stock config + calibration. How do I change it?

 

[Heinrich]

I accidentally didn't screenshot the labels, my bad

Here is it with the Labels below:
View attachment 438554

It would support the scientific spirit to mark the scales compehensible. Instead of time there should be frequency.

 

[Plompudu]

It would support the scientific spirit to mark the scales compehensible. Instead of time there should be frequency.

It's not a FR or RTA like the ones I shown above.

It's a SPL Log to show the dynamic range of the content / what my system reproduces over the 3 minutes of the beginning of the Show I've talked about, since the RTA doesn't show the peak values/dynamics properly.

Or did you mean something else by your comment, that I'm missing?