• 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!

Where from does the clicks come in this "unhearable" file?

OP
tomtoo

tomtoo

Major Contributor
Joined
Nov 20, 2019
Messages
3,722
Likes
4,821
Location
Germany
I would say what you hear are the audio frequencies of your speaker trying to follow the signal, and not being able to, to put it simply your speaker does the best it can to respond to the electrical impulse, and it makes a pigs breakfast out of it.

That was my first thought too. But thats not what realy happens. If you record the signal you can see that the speaker follows the signal.
 

JustAnandaDourEyedDude

Addicted to Fun and Learning
Joined
Apr 29, 2020
Messages
518
Likes
820
Location
USA
Basically, you get a frequency domain spectrum that has harmonics at multiples of the repetition frequency. So if the repetition rate is low, the harmonics will be pretty rich in the audible range. Limiting case: a unit impulse of a width that approaches zero time has a frequency spectrum that includes all frequencies.

This tech note explains tone bursts pretty well. Skip though the baffling math and concentrate on the conclusions.
https://ietlabs.com/pdf/GR_Appnote/IN-105 Frequency Spectrum of a Tone Burst.pdf

Pretty cool link. Never heard of a tone burst before, but that tech note explains it very well. They spared us the derivations, which are probably hairier (particularly for the amplitude) than the neat expressions shown on the right hand sides.
 

bennybbbx

Addicted to Fun and Learning
Joined
May 27, 2020
Messages
691
Likes
124
Location
germany
Maybe the sonogram of the file can help you :

View attachment 68482

What we hear are the vertical blue bars.
and what db level are the blue lines ?. I think in such spectogram algorithm is something wrong. I think the time length of the clicks are much to much in blue lines. to do such FFT analyze there need windowing do. https://download.ni.com/evaluation/pxi/Understanding FFTs and Windowing.pdf maybe the window method produce this low level long blue lines as a side effect. the click have more level as the 20 khz sine . and blue is less level.I only hear the better the speaker or headphone the less low frequency hear from the click.

Maybe have a good headphone or speaker that let not hear the click, then virtual surround work better. so if somebody here have a high end headphone that is in frequency response list at more than 30 khz please let know if you hear the click in the example with music. the best headphones are the stax, but beyerdynamic DT770 reach 35 khz. adam speakers reach 50 khz
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,250
Likes
17,200
Location
Riverview FL
Here's the original file, and a copy of the first non-zero sample.

It is clearly audible as a click,


1591949620246.png


It occurs to me that that first sample is 124dB above 24bit digital silence...

1591950670804.png


Ok, only 76db above 16bit silence, but still...
 
Last edited:

bennybbbx

Addicted to Fun and Learning
Joined
May 27, 2020
Messages
691
Likes
124
Location
germany
yes its -20 db and click in middle is -14 db. it seem it is not possible to produce transient perfect with kali and headphones i have. I use start. the trust gxt322c headset look as best. 1. picture is sony is ZX 310 2. picture trust 3. picture kali lp6
sony start.jpg

trust start.jpg
kali start.jpg
trust start.jpg
 

Attachments

  • kali start.jpg
    kali start.jpg
    160.4 KB · Views: 73
  • trust start.jpg
    trust start.jpg
    141.5 KB · Views: 95
Top Bottom