• 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!

Audibility thresholds of amp and DAC measurements

luft262

Senior Member
Joined
Mar 25, 2021
Messages
465
Likes
236
Location
Phoenix
Where could I find a brief and simple guide that explains each of the tests and what they mean? Thanks.
 

wwenze

Major Contributor
Joined
May 22, 2018
Messages
1,284
Likes
1,827
I should try recording the output from a China TDA1543 DAC that measures around 0.3% IIRC. Any nomination for the music?
 

192kbps

Addicted to Fun and Learning
Joined
Apr 3, 2019
Messages
534
Likes
614
In 'Sound Reproduction: The Acoustics and Psychoacoustics of Loudspeakers and Rooms' Floyd Toole says the following: "The simplest deviation from flat is probably a spectral tilt. There is some evidence that we can detect slopes of about 0.1 dB/octave, which translates into a 1 dB tilt from 20 Hz to 20 kHz — not much."

"If the FR rises by 0.05 dB in one area and drops by 0.05 dB in another, the total discrepancy will be 0.1 dB."

Frequency response: ±0.05dB
Channel balance: 0.1dB

Am i right?
 

Robbo99999

Master Contributor
Forum Donor
Joined
Jan 23, 2020
Messages
6,874
Likes
6,672
Location
UK
In 'Sound Reproduction: The Acoustics and Psychoacoustics of Loudspeakers and Rooms' Floyd Toole says the following: "The simplest deviation from flat is probably a spectral tilt. There is some evidence that we can detect slopes of about 0.1 dB/octave, which translates into a 1 dB tilt from 20 Hz to 20 kHz — not much."

"If the FR rises by 0.05 dB in one area and drops by 0.05 dB in another, the total discrepancy will be 0.1 dB."

Frequency response: ±0.05dB
Channel balance: 0.1dB

Am i right?
(Is that anything to do with this thread? Not read back through the pages)

But he's talking about a 1dB tilt all the way from 20Hz to 20kHz being audible, not a 0.1dB tilt across that same range. And you're only going to hear those differences if it's a consistent tilt, not just some random +/-1dB variation across the whole range.
 

BYRTT

Addicted to Fun and Learning
Forum Donor
Joined
Nov 2, 2018
Messages
956
Likes
2,452
Location
Denmark (Jutland)
(Is that anything to do with this thread? Not read back through the pages)

But he's talking about a 1dB tilt all the way from 20Hz to 20kHz being audible, not a 0.1dB tilt across that same range. And you're only going to hear those differences if it's a consistent tilt, not just some random +/-1dB variation across the whole range.

He said "0,1dB/octave" and that's not same as 0,1dB tilt across that same range :) 0,1dB/octave arrive at a 1dB difference @20Hz/20kHz..

Robbo99999.png
 
Last edited:

Robbo99999

Master Contributor
Forum Donor
Joined
Jan 23, 2020
Messages
6,874
Likes
6,672
Location
UK
He said "0,1dB/octave" and that's not same as 0,1dB tilt across that same range :) 0,1dB/octave arrive at a 1dB difference @20/20kHz..

View attachment 143140
Yeah, that's cool, I understand that (already). But I don't see how it relates to amp specs for the other points I highlighted in my previous post, ie +/-0.05dB in an amp spec is gonna mean random variation across the whole spectrum from 20Hz-20000Hz - so that's not a "tilt" phenomenon (and certainly not in reference to over an octave). So @S34MH1 is comparing apples to oranges.
 

Lambda

Major Contributor
Joined
Mar 22, 2020
Messages
1,785
Likes
1,519
Anything below -120 dBFS is inaudible. Amir wrote an in-depth article about the subject.
Not relay sure how you come to this conclusion/assumption?
Fielder shows measurements in his paper for the other extreme for how loud music passages can get in live venues, registering values as high as 130 dB SPL.
And this is not even true peak!


If i just uses this graphic and go from the lowest to the highest spot it is almost 140dB?
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,835
Likes
16,497
Location
Monument, CO
At 1 kHz the dynamic range on the chart is 100 dB. We are less sensitive to deep bass. An average very quiet room is probably about 40 dB; my room is around 30 dB (little lower on a good day) but is isolated (floating walls and ceiling heavily insulated). The other rooms are more like 50 dB during the day dropping to maybe 40 dB at night. The threshold of pain is around 120 dB at 1 kHz, depending upon who you talk to, so a dynamic range of 80 dB or so is plenty for most of us (40 dB to 120 dB, or 30 dB to 110 dB).

This is a site I found with a quick search: http://www.sengpielaudio.com/TableOfSoundPressureLevels.htm
 
Last edited:

Lambda

Major Contributor
Joined
Mar 22, 2020
Messages
1,785
Likes
1,519
At 1 kHz the dynamic range on the chart is 100 dB
Yeah but we are not only listening to 1khz?!
The threshold of pain is around 120 dB at 1 kHz
But not on 20Hz and not for short pulses.

AN average very quiet room is probably about 40 dB
And yet we can hear signals well below 40dB.

Have you ever done a hearing test?
they don’t do it any special isolated room and you i can easily hear signals in the singe digit dB SPL range.
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
And yet we can hear signals well below 40dB.

40dB noise levels is the total noise over the entire audible band.
It may be 10 to 15dB in a narrower and medium to higher frequency band so we can hear below the noise floor of a room because there is also a frequency component in that noise floor which does not get masked by noise in bands away from the signal.
 

Lambda

Major Contributor
Joined
Mar 22, 2020
Messages
1,785
Likes
1,519
40dB noise levels is total over the entire audible band.
It may be 10 to 15dB in a narrower and medium to higher frequency band so we can hear below the noise floor of a room because there is also a frequency component in that noise floor.
Yes and this is why it makes no sense to just substrate the ambient noises from the threshold of pant to get available dynamic.

And this is not even true peak!
The values are averaged and can differ about ±10 dB. With sound pressure p is
always meant the root mean square value (RMS) of the sound pressure, without
extra announcement.
This SPL values are normally RMS and so we can't just translate RMS to peak without knowing the Crest.

for very low frequency and short pulses:
The ear operates as an energy detector that samples the amount of energy present within a certain time frame. A certain amount of energy is needed within a time frame to reach the threshold. This can be done by using a higher intensity for less time or by using a lower intensity for more time. Sensitivity to sound improves as the signal duration increases up to about 200 to 300 ms, after that the threshold remains constant.[2]
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,835
Likes
16,497
Location
Monument, CO
Yeah but we are not only listening to 1khz?!
It's a common reference frequency for the midband. I simply stated conditions at which dynamic range is ~100 dB; the curves show that level spans a rather broad range of frequencies through the midrange.

But not on 20Hz and not for short pulses.
Did not claim it was. See links to OSHA site. It's your hearing, do what you want.

And yet we can hear signals well below 40dB.
Sometimes, depending upon the signal and the room. I often have my AVP set to -40 dBFS for background music in the media room but that would be too quiet in the family room.

Have you ever done a hearing test?
they don’t do it any special isolated room and you i can easily hear signals in the singe digit dB SPL range.

I have had many hearing tests and it was always in a special room and with headphones. Haven't had an audiologist come to my family room to administer one. I really doubt I could hear single-digit SPL in my family room without headphones but I am sure my hearing acuity and technical knowledge of the subject is far inferior to yours.

Sorry I stepped into this, have better things to do. This seems like another one of those endless Internet debates all too common these days that are simply a waste of time.
 

Lambda

Major Contributor
Joined
Mar 22, 2020
Messages
1,785
Likes
1,519
Did not claim it was. See links to OSHA site. It's your hearing, do what you want.
"All OSHA" is dB(A)!
According to your source:

150 dBA​
Hammer stroke in a smithy at 5 m distance (greatest level)
130 dBA​
Loud hand clapping at 1 m distance (greatest level)



I really doubt I could hear single-digit SPL in my family room without headphones
No one is stopping you from using Headphones?
as well as no one is stopping you from going into a quiet room.
If we want to talk about limits of hearing dynamic it shuld be kind of obvious to assume ideal listening conditions.

Im not claiming i have Golden Ears. they are "normal"
And i have been to live Concerts with way over 135dB.
i have been clapping a lot 130dB(A)@1m and my arms al not 1m long.
I also hammered a lot and pooed a view balloons.
Screen-Shot-2017-08-28-at-2.29.32-PM.png
 
Last edited:

Beershaun

Major Contributor
Forum Donor
Joined
Oct 3, 2019
Messages
1,864
Likes
1,910
Amir did a YouTube video on "how loud is loud" to answer the question. Please watch it.

 

Lambda

Major Contributor
Joined
Mar 22, 2020
Messages
1,785
Likes
1,519
Well the paper he is quoting says in the abstract
A dynamic range of over 120 dB is found to be necessary in the most demanding of circumstances, requiring the reproduction of sound levels of up to 129 dB SPL. Present
A Jazz concert was measured at 127dB SPL
According to the same Paper just audible can be as low as -15dBSPL @ about 4Khz for golden ears.

Amir also says he is sure 1ms of 140dB is not gonna hurt your hearing.
He also says SPL meter don't measure Peak and peak is a good 10-20dB over what the SPL meter shows.


answer the question.
So he answers a lot of things but not how and why it is often quoted.
"Anything below -120 dBFS is inaudible"

In fact he talks 40min about how we need way over 120dB peak dynamic.

Best case from his numbers.
The golden Ear subject from the paper can hear -15dB SPL the same papers upper limit of 129dB SPL for a concert.
would give a required dynamic -144dB Dynamic.
So a DAC/AMP would need to have -144dB SPL to be truly transparent if the golden ear subjekt from the paper wants to listen at original live volume.

Normal Hearing person hears -4dB SPL but lets say 0 dB SPL.
To faithfully reproduce a unaplified jazz concert for a normal person at original volume would need -127dB dynamic.

So how and why is -120dB call a "strict limits" and an Absolute limit?
 
Last edited:

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,835
Likes
16,497
Location
Monument, CO
OSHA is based upon average, not peak, SPL levels. Endless sources cite 20~30 dB peaks above average. The average level at which people listen has to be high enough to be well above the noise floor so soft sounds are heard. Needing 144 dB dynamic range does not mean you can hear sounds at -144 dB SPL; you are mixing up listening thresholds, peaks, and dynamic range.

Readings in dBA are used to emulate the way we hear and thus louder low-frequency sounds are attenuated (weighted) by the curve in a manner similar to the loudness curves to reflect our decreased sensitivity to those low-frequency sounds.

I have never heard 120 dB being called any kind of strict limit. It has been used as the threshold of pain, the point at which midrange sounds start to actually hurt, but of course there are a myriad examples of peaks far above that. And duration matters, of course a short burst is less likely to cause damage than sustained sounds.

For inaudibility, I thought the discussion had morphed into practical limits. I do not think I can hear a sound that is say 40 dB below the average noise level of my room but I have never tried that specific experiment. Most quiet residential rooms are in the 40~50 dB range, well above the absolute threshold of audibility. How far below the noise floor one can hear depends upon things like the exact frequency and how long the sound lasts. In the midrange, 10 dB is about half the loudness level, so if you can just hear a 40 dB sound in your room and you cut the volume in half that is 30 dB. Maybe you can pull another 10 dB out, which puts you at 20 dB, still 20 dB above the threshold of audibility. If your room's noise is 40 dB, then a sound at 0 dB is only 1/16 the volume.

I am not sure what all the fuss is about, but have no interest in carrying on this train wreck.
 

Lambda

Major Contributor
Joined
Mar 22, 2020
Messages
1,785
Likes
1,519
OSHA is based upon average, not peak, SPL levels. Endless sources cite 20~30 dB peaks above average.
This is what i’m saying.
Endless sources cite 20~30 dB peaks above average
i did not claim its otherwise.

The average level at which people listen has to be high enough to be well above the noise floor so soft sounds are heard.
And if you Want to hear at Original volume same as you would hear in a live concert you need a system that can produce the same SPL peak level
and on the other side has noise lower than your hearing threshold.

Readings in dBA are used to emulate the way we hear and thus louder low-frequency sounds are attenuated (weighted) by the curve in a manner similar to the loudness curves to reflect our decreased sensitivity to those low-frequency sounds.
Yes.
I have never heard 120 dB being called any kind of strict limit. It has been used as the threshold of pain, the point at which midrange sounds start to actually hurt, but of course there are a myriad examples of peaks far above that. And duration matters, of course a short burst is less likely to cause damage than sustained sounds.
This is also all correct.

I have never heard 120 dB being called any kind of strict limit.
You might have not but have you read the first post of this thread?
 

Beershaun

Major Contributor
Forum Donor
Joined
Oct 3, 2019
Messages
1,864
Likes
1,910
Another point that is often not considered is why the target for a DAC is more strict than for an amplifier or a speaker. The reason is the Amplifier will Amplify all the DAC signals by 26-29db depending on it's gain. So if the DAC noise level is 90db then the Amplifier will Amplify that to -61db noise floor. And I the 2nd harmonic is -80db then the Amplifier will Amplify that to -51db. So upstream systems defects get amplified along with the signal we want and therefore need to more strictly controlled.
 
Top Bottom