Getting started with Room Equalising

[Ron Wagstaff]

I've just purchased the AudioTool app and an IMM-6 microphone to see if I can improve my home listening experience. 

I'm working with the Peak Store readings.  I've generated a Log Sweep from 10-20000 Hz with Freq Scale Limits set to 20-20000 Hz.   When I look at the 1/6 Octave values in the stored file, why is the first value (20hz) always negative?  

[Brian Greenwood]

I can't speak to the specifics on the app and how it's representing 20hz as negative.  But I would encourage you not to be too concerned about it.

Most home Subwoofers can't produce an audible 20hz.  Typically their roll off is around 30hz if not a little higher.  The internal circuitry often filters lower frequencies in order to protect the speaker driver from the damage that could be caused by trying to produce ultra low frequencies.

Honestly you don't want 20hz playing in your living room or media room.  In fact you don't want it in your house.  In my experience (15 yrs as a Pro AV Installer and Systems Designer) frequencies below 30hz are only useful when you need to carry sub bass frequencies a long distance.  I'm talking about filling 1,000+ seat auditoriums and outdoors concerts with strong bass.  Small and even mid sized churches don't typically need anything much below 40hz.  The instruments themselves aren't playing that low.  The harmonics from a bass guitar can produce below 30hz, but not the direct tones.  Bass players correct me if I'm wrong here but I believe the bottom note on a bass guitar is tuned around 40 - 45 hz. 

Please don't misunderstand me, I'm not suggesting you change the setup of your log sweep.  You want to start it below what you expect it to play, and end above as well.  So 20 - 20K is just fine. Just understand that you should expect little to no sound at the bottom frequencies.  You should see a dramatic change once you hit the bottom end of what the Sub Can play.

Hope that helps,

Brian

On Sun, Jun 28, 2020, 9:18 AM Ron Wagstaff <ronwag...@gmail.com> wrote:

I've just purchased the AudioTool app and an IMM-6 microphone to see if I can improve my home listening experience. 

I'm working with the Peak Store readings.  I've generated a Log Sweep from 10-20000 Hz with Freq Scale Limits set to 20-20000 Hz.   When I look at the 1/6 Octave values in the stored file, why is the first value (20hz) always negative?  

--
You received this message because you are subscribed to the Google Groups "AudioTool Discussion Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to audiotool-discussio...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/audiotool-discussion-group/e250792a-e801-49b5-b1ee-a777a5680b09n%40googlegroups.com.

[Ron Wagstaff]

Thank you for your reply, Brian.  As you point out this isn't relevant to my project but I'm a mathematician and I automatically hone onto numerical abnormalities.

[Julian Bunn]

Thanks, Brian for this excellent information!

Ron: it may be that there is zero signal in that band, and with the global offset subtracted, the app shows the resulting  negative number. 

Julian

To view this discussion on the web visit https://groups.google.com/d/msgid/audiotool-discussion-group/aa07627c-6b73-434e-988b-054f7df2d5aen%40googlegroups.com.

[Jean-Baptiste ROMANET]

dB scaled levels can be negative if the level is below the reference. If the pressure calculated at 20 Hz is below thz reference 20 micropascals, the level in dB is negative.
High end sound level meters with high sensitivity microphones have a self noise below -10 dB in each 1/3 Oct band (so below -13dB in each 1/6 Oct band).

[bmampl...@gmail.com]

Ron, When I got Audiotool the first thing I did was plot the frequency response of the phone I was using.  I found it dropped off rapidly below 100Hz, so much so that it was impossible to compensate for. So, unless you have a device with a known and compensated for frequency response, your results at low frequencies are likely very unreliable.

Nick

[Ron Wagstaff]

Thanks for your answers guys, but as usual, they raise more questions than answers.  Now this may just highlight my lack of understanding in this whole area (for instance I had to look up what micropascals where).  

Firstly, Julian I haven't set a Calibration Offset (it is still 0.0) if that is what you are referring too.  

I get the same negative value (-35.8) irrespective of the sound source, whether its my hi-fi speakers or headphones. I also loaded the software onto my old TAB-3 tablet and got the same value.

I don't understand the relation between most of the data and the final 1/6 Octave figures that I'm looking at, but there are no other negative numbers appearing anywhere in the saved data.  I've attempted to attach a swp file but I can't post with it attached.

I guess my main concern was that I had spotted a flaw in the system that wasn't just an issue for me.  At the end of the day, I'm happy if you're happy.    

Now Nick has got me worried with his reply.  Does this mean that this setup (IMM-6 with AudioTool running on a Samsung S9) will be no use to me setting up my subwoofer?  

How exactly do I plot the frequency response of my phone?

[Julian Bunn]

Hi Ron,

What do you see in the 20Hz band in the octave RTA modes (not the sweep mode)? I don't discount that there may be a bug in the sweep mode, but I've had no reports of anything similar. If the phone or tablet has zero response at 20Hz that would not be unusual.

There is a built in global offset of 20dB which as a user you don't see - the user supplied global offset is applied in addition to the built in value.

Julian

--
You received this message because you are subscribed to the Google Groups "AudioTool Discussion Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to audiotool-discussio...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/audiotool-discussion-group/02a2f396-76c7-4388-b289-04fb56dcaa3en%40googlegroups.com.

[Ron Wagstaff]

Hi Julian

Here are the first few lines from a 1/6 Octave Run:

AudioTool Spectrum File Wed Jul 01 10:52:40 GMT+01:00 2020 Rate:44100 Length:4096 dB Offset:0.0 Columns are: Hz, Volts, dB, peakVolts, valleyVolts

0.0 0.0 0.0 0.0 0.0

10.7666015625 0.10703126712701448 4.790592431555364 0.34507722106789357 0.0

21.533203125 0.5328435017110413 18.732008770437492 1.171985427239618 0.0

32.2998046875 3.1989332622079725 34.300104018471174 9.949938687528121 0.0

43.06640625 2.107183852501624 30.67404957011604 35.92375635000283 0.0

And then the 1/6 Octave data:

1/6 Octave Data: Columns are Frequency(Hz), dB, dB(valley), dB(peak)

20.0 -35.8 84.19999565705736 -35.8

22.4 18.732008770437492 84.19999565705736 25.57844739362418

25.0 18.732008770437492 84.19999565705736 25.57844739362418

28.0 18.732008770437492 84.19999565705736 25.57844739362418

31.5 34.300104018471174 84.19999565705736 44.15640813566599

35.5 34.300104018471174 84.19999565705736 44.15640813566599

40.0 34.300104018471174 84.19999565705736 44.15640813566599

45.0 30.67404957011604 84.19999565705736 55.30763484655795

The 1/12 Octave data is even more illuminating:

AudioTool Spectrum File Wed Jul 01 10:56:55 GMT+01:00 2020 Rate:44100 Length:4096 dB Offset:0.0 Columns are: Hz, Volts, dB, peakVolts, valleyVolts

0.0 0.0 0.0 0.0 0.0

10.7666015625 12.384320473761381 46.05744366606403 32.124454471601766 0.0

21.533203125 85.39488988735798 62.828637657906484 188.73877504092513 0.0

32.2998046875 220.5644588605765 71.07071065498617 952.9294090753184 0.0

43.06640625 200.9064582673426 70.2598779530704 2907.3589320512706 0.0

1/12 Octave Data: Columns are Frequency(Hz), dB, dB(valley), dB(peak)

2.5 -35.8 84.19999565705736 -35.8

2.6 -35.8 84.19999565705736 -35.8

2.8 -35.8 84.19999565705736 -35.8

3.0 -35.8 84.19999565705736 -35.8

3.1 -35.8 84.19999565705736 -35.8

3.4 -35.8 84.19999565705736 -35.8

3.5 -35.8 84.19999565705736 -35.8

3.8 -35.8 84.19999565705736 -35.8

4.0 -35.8 84.19999565705736 -35.8

4.3 -35.8 84.19999565705736 -35.8

4.5 -35.8 84.19999565705736 -35.8

4.8 -35.8 84.19999565705736 -35.8

5.0 -35.8 84.19999565705736 -35.8

5.3 -35.8 84.19999565705736 -35.8

5.6 -35.8 84.19999565705736 -35.8

6.0 -35.8 84.19999565705736 -35.8

6.3 -35.8 84.19999565705736 -35.8

6.7 -35.8 84.19999565705736 -35.8

7.1 -35.8 84.19999565705736 -35.8

7.5 -35.8 84.19999565705736 -35.8

8.0 -35.8 84.19999565705736 -35.8

8.5 -35.8 84.19999565705736 -35.8

9.0 -35.8 84.19999565705736 -35.8

9.5 -35.8 84.19999565705736 -35.8

10.0 -35.8 84.19999565705736 -35.8

10.6 46.05744366606403 84.19999565705736 54.33671523092707

11.2 46.05744366606403 84.19999565705736 54.33671523092707

11.8 46.05744366606403 84.19999565705736 54.33671523092707

12.5 46.05744366606403 84.19999565705736 54.33671523092707

13.2 46.05744366606403 84.19999565705736 54.33671523092707

14.0 46.05744366606403 84.19999565705736 54.33671523092707

15.0 46.05744366606403 84.19999565705736 54.33671523092707

16.0 46.05744366606403 84.19999565705736 54.33671523092707

17.0 46.05744366606403 84.19999565705736 54.33671523092707

18.0 46.05744366606403 84.19999565705736 54.33671523092707

19.0 46.05744366606403 84.19999565705736 54.33671523092707

20.0 46.05744366606403 84.19999565705736 54.33671523092707

21.2 62.828637657906484 84.19999565705736 69.71722264108706

22.4 62.828637657906484 84.19999565705736 69.71722264108706

23.6 62.828637657906484 84.19999565705736 69.71722264108706

25.0 62.828637657906484 84.19999565705736 69.71722264108706

26.5 62.828637657906484 84.19999565705736 69.71722264108706

28.0 62.828637657906484 84.19999565705736 69.71722264108706

30.0 62.828637657906484 84.19999565705736 69.71722264108706

31.5 71.07071065498617 84.19999565705736 83.78121460491099

33.5 71.07071065498617 84.19999565705736 83.78121460491099

35.5 71.07071065498617 84.19999565705736 83.78121460491099

37.5 71.07071065498617 84.19999565705736 83.78121460491099

40.0 71.07071065498617 84.19999565705736 83.78121460491099

42.5 70.2598779530704 84.19999565705736 93.46997303022518

45.0 70.2598779530704 84.19999565705736 93.46997303022518

The points of note I see are:

The negative numbers are always -35.8

The DB(valley) are always 84.199.. (that's through the whole frequency range)

DB(peak) is often lower than DB(valley)

One thing I should point out is that I'm running version 8.3.3.  I don't seem to have the option to update to v8.4.  Is that because I'm in the UK?

[Julian Bunn]

Hi Ron,

For Version 8.4, "8.3.3" is the version number you see if you look at "Version" in the Menu - an oversight when I submitted it. So you have the latest version.

The dB "valley" numbers will not be meaningful unless you have turned on the Valley setting in AudioTool. After doing so, all valley levels are initially set to some high value, and as soon as new data arrive, the valley levels are adjusted to be the minimum of the incoming data and what is already there.

For the low frequency dB levels, the difficulty is that the FFT measurements are quantised to frequency intervals of SampleRate/SampleLength, which in your case is about 10Hz. This means that the first dB level for which there is a measurement is at 10Hz, the next at 20Hz etc.. Any frequency lower than that we don't have a measurement for, and so AudioTool will use some artificial, low value. And, for frequencies higher than that, AudioTool tries to spread the dB level it has a measurement for evenly into surrounding bins. So, for 1/12 octave you will see many of the low frequency 1/12 octave bins share the same values. This is the best it can do.

The takeaway is that the very low frequency measurements (like below around 20Hz) are not reliable.

I hope this helps.

Julian

To view this discussion on the web visit https://groups.google.com/d/msgid/audiotool-discussion-group/8e4636b5-b567-4628-9bc9-2a2525ec4994n%40googlegroups.com.

[Ron Wagstaff]

Hi Julian 

Thank you for persisting with me; I know I can be a bit anal-retentive.  

My last say on this matter is, may be the artificial low and high values aren't extreme enough and perhaps you should have chosen obvious numbers like -9999 and 9999. 

I shall now get on with what I purchased the software for.

Regards

Ron

[bmampl...@gmail.com]

Hi Ron,

I determined the frequency response by feeding the mic input with a sine wave signal generator and setting to the centre frequency of each band in turn. Looking at my  notes, it was a while back, the drop at 80Hz and below was

-23dB @ 31Hz

-23dB @ 40Hz

-11dB @ 50Hz

-4dB @ 63Hz

0db @ 80Hz

Nick

[Neo Bass]

The low B on a five string bass (guitar)  has a fundamental of 31Hz and the actual wavelength is I think 36 feet. The lowest note of a Contrabassoon is something like 29Hz and reminds me of someone flapping their hand by my ear.