• 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!

Time vs. Frequency Domain

OP
MRC01

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,422
Likes
4,029
Location
Pacific Northwest
Audacity 2.1.2, low pass filter (effect, plugins 1-15), with corner frequencies from 10k to 20k at 1/6 octave intervals, each one 6 dB / octave.
BTW I am adding the above recording to the list I use when reviewing gear. It should be ideal for detecting any audible differences in the top octave. I should have done this years ago.
 
Last edited:

andreasmaaan

Master Contributor
Forum Donor
Joined
Jun 19, 2018
Messages
6,652
Likes
9,399
Audacity 2.1.2, low pass filter (effect, plugins 1-15), with corner frequencies from 10k to 20k at 1/6 octave intervals, each one 6 dB / octave.
BTW I am adding the above recording to the list I use when reviewing gear. It should be ideal for detecting any audible differences in the top octave. I should have done this years ago.

Thanks :)

So it's this effect:

1550020564748.png

With these settings (obviously with various frequencies from 10KHz up in 1/6 steps):

1550020626909.png


?
 

andreasmaaan

Master Contributor
Forum Donor
Joined
Jun 19, 2018
Messages
6,652
Likes
9,399
If that's the case, your issue is that the filter affects the amplitude of frequencies well below the corner frequency.

On the left is the spectrum of white noise with no filter applied, while on the right is the spectrum of the same white noise with the filter applied that I took screenshots of in my previous post (i.e. 6dB/octave @ 20KHz, which is the one I believe you're applying):

1550020994581.png


Sorry the scales don't quite line up (Audacity does that automatically and it can't seem to be changed), but as you can see, the filter begins rolling off at around 4KHz and is (for example) nearly 2dB down at 10KHz. This should definitely be audible.

To do this test correctly, you need to apply a very steep linear phase filter that doesn't even begin to roll off until above the highest frequency you can hear.

To my knowledge this isn't possible in Audacity, although I'm not 100% across all its functions. Certainly doesn't seem to be from what I can tell. What you want to do might require something a bit more specialised I think.

(As a side note, that kind of minimum-phase filter will also be affecting the phase of the signal well below the corner frequency, but I doubt that's specifically the cause of the difference you're hearing.)
 
OP
MRC01

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,422
Likes
4,029
Location
Pacific Northwest
That sounds about right. IIRC, a typical low pass filter is something like 3 dB down at the corner frequency. You're right, a steeper filter would be better. Then it starts looking like the classic "sharp" sinc(t) Whittaker-Shannon reconstruction filter. If you do something that steep in minimum phase, it can create audible passband phase distortion.

And we've only scratched the surface here. It's a glimpse of the effort required, and respect earned, by folks who get these tests right to the extent that the results are reliable and pass peer review.
 

andreasmaaan

Master Contributor
Forum Donor
Joined
Jun 19, 2018
Messages
6,652
Likes
9,399
The closest thing I think you could manage with Audacity would be a 48dB/octave (I presume Butterworth) filter. It gets you closer in the frequency domain (just a narrow band of attenuation below the corner frequency), but will of course still not get you there in the time domain:

1550024161344.png
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,521
Likes
37,050
The closest thing I think you could manage with Audacity would be a 48dB/octave (I presume Butterworth) filter. It gets you closer in the frequency domain (just a narrow band of attenuation below the corner frequency), but will of course still not get you there in the time domain:

View attachment 21825
You can open the equalization effect and make your own curves.

You also can do this with Sox or Sox plugin for foobar.

You also can perform the filter more than once.
 

andreasmaaan

Master Contributor
Forum Donor
Joined
Jun 19, 2018
Messages
6,652
Likes
9,399
You can open the equalization effect and make your own curves.

You also can do this with Sox or Sox plugin for foobar.

You also can perform the filter more than once.

This is great to know, thanks :)
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,198
Likes
16,981
Location
Riverview FL

Sergei

Senior Member
Forum Donor
Joined
Nov 20, 2018
Messages
361
Likes
272
Location
Palo Alto, CA, USA
Me too. Best explanation I can think of is that that 25dB of audio band IMD the super tweeter created poked through the masking threshold at a couple of points in playback of the 2nd stimulus.

Anyone got any other theories?

Yes. Note: the article mentions that the effect was only observed in longer-duration tests. This could be related to physiological responses described here: Therapeutic Ultrasound in Physical Therapy.

Heating and cavitation in the tissues of the hearing system, caused by application of intense enough ultrasound for long enough time, could lead to either direct bodily sensations, or to physiological changes affecting perception of sound in the frequency range of direct hearing.

Read more about how ultrasound can be felt, rather than heard: Intense focused ultrasound can reliably induce sensations in human test subjects in a manner correlated with the density of their mechanoreceptors.

That could further explain why some of the subjects detected the difference, while other didn't. Despite their hearing frequency ranges being similar, the density of their mechanoreceptors in the areas of the body exposed to ultrasound could be significantly different.
 
OP
MRC01

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,422
Likes
4,029
Location
Pacific Northwest
I've noticed that as sounds approach the upper limit of my hearing, they stop sounding like a clean tone and my perception changes to a sense of active "deadness". In this threshold frequency range I can detect the presence or absence of such sounds but they don't sound like normal "sounds". Then above that range they fade out entirely. Also if I play higher frequencies than I can hear, at moderate to loud levels, even though I can't hear them, after a while I feel symptoms similar to a headache. Seem as if the impulse pressure affects the ear even if it's not actually "hearing" it.
 
Top Bottom