• 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!

Can you define me a threshold for "audible - hearable"?

mehmethand

Member
Joined
Mar 19, 2021
Messages
10
Likes
4
I want to ask a question. Let's say you have 2 audio files to compare and they sound very similar. You can't spot and hear the difference with ABX blind test. "You can't hear the difference" doesn't mean that other people can't hear it,too. Hearing depends on where you listen, how you listen, how loud you listen, your hearing skills... etc and all these change from person to person. So here is my question :

You have 2 very similar audio files. You can't hear the sound difference between them and you want to know if the sound difference between these files have a chance to be heard by some other people. Is there anyway to make an assumption about the difference between 2 compared files is hearable or not ? Is there some kind of threshold any kind of analytic data or scientfic research where you can say this difference is hearable between 2 compared audio files?

With audio analyzer softwares you can have rms, peak,lufs delta... like values or have spectrograms, spectrum analysis graphs... Is there any kind of information hidden in there to define a threshold for hearblity betwwen 2 compared audio files?

Note : I'm using deltawave audio null comparator software for this kind of comparison. https://deltaw.org/
 
Last edited:

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
7,424
Likes
7,940
Location
Brussels, Belgium
Amir thinks that 70 SINAD is transparent for untrained ears (like actually being trained by Harman and participating in their experiments). So there is that I guess.
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,202
Likes
16,982
Location
Riverview FL
For me -70dB is the difference between normal loudish and barely hear anything if at all.

Using Audacity, you could take a track and attenuate parts of it by whatever dB you like, to see the difference between that and your normal listening levels.

You'd have to attenuate by 50 and then another 20dB - it limits the change to -50dB at a time.

Example:

Some music with 70dB attenuation in the middle of it.

1616188528550.png


If you think nothing is there in the attenuated part, look closer:

(see linear scale on left)

1616188616787.png
 

AdamG

Proving your point makes it “Science”.
Moderator
Forum Donor
Joined
Jan 3, 2021
Messages
4,636
Likes
14,918
Location
Reality
I use a calibrated umk and some app on my iPad to pretty much monitor volume spl in my room. So what I am going to say is obviously location dependent. With nothing playing music or tv. My room noise floor hovers between 50db and 60db. As I increase volume I don’t really hear what’s being played until I hit about 62>db. What I’m saying is the answer to your question may be specific to your listening room/location and personal hearing capabilities. Until you measure your room, it’s all a guess on our part.

I realize that this may not be the answer you were looking for. Just trying to help add context to the answer to your inquiry.
 

Chrispy

Master Contributor
Forum Donor
Joined
Feb 7, 2020
Messages
7,751
Likes
5,910
Location
PNW
Doesn't seem like a worthy use of time to try and figure out what others might or might not hear differently due all the variables involved. Looking at an analyis of the file might give you some ideas if the differences are great enough (or if the difference isn't just response rather than say a different mix).
 
OP
M

mehmethand

Member
Joined
Mar 19, 2021
Messages
10
Likes
4
I asked the same question to the developer of Deltawave software in another forum. I'm copying his answer here just to show his opinion. I would really like to hear other approaches.
Here is his answer to the same question :
"
There are a few things you can use to judge whether the difference is audible or not.
1. The delta file, the difference between the reference and comparison files. You can play this file at a loud level (use the volume control if it's too low, but be careful not to blow your speakers!) If you have to add 50dB+ to the volume of the delta file to hear it, then the differences will not be audible. You can also listen to what remains in the delta file. If it's all noise and no music, or music is well below the level of noise, you will not hear the difference.
2. The RMS Null dB value, and especially dbA value can be useful. If these are below -70dB, you will not hear the difference. These are average values, so this doesn't mean there can't be one or a few spots where the differences are larger, but overall, this means the level of differences is below audibility.
3. Look at the delta waveform and you can judge by eye if there's sufficiently large differences that might be audible. For example, this delta file represents something that just barely at the edge of being audible, possibly when playing really loud or in a very, very quiet environment, and with very young ears :)
"
 

dc655321

Major Contributor
Joined
Mar 4, 2018
Messages
1,597
Likes
2,235
I asked the same question to the developer of Deltawave software in another forum. I'm copying his answer here just to show his opinion. I would really like to hear other approaches.
Here is his answer to the same question :
"
There are a few things you can use to judge whether the difference is audible or not.
1. The delta file, the difference between the reference and comparison files. You can play this file at a loud level (use the volume control if it's too low, but be careful not to blow your speakers!) If you have to add 50dB+ to the volume of the delta file to hear it, then the differences will not be audible. You can also listen to what remains in the delta file. If it's all noise and no music, or music is well below the level of noise, you will not hear the difference.
2. The RMS Null dB value, and especially dbA value can be useful. If these are below -70dB, you will not hear the difference. These are average values, so this doesn't mean there can't be one or a few spots where the differences are larger, but overall, this means the level of differences is below audibility.
3. Look at the delta waveform and you can judge by eye if there's sufficiently large differences that might be audible. For example, this delta file represents something that just barely at the edge of being audible, possibly when playing really loud or in a very, very quiet environment, and with very young ears :)
"

FYI - @pkane is the author of DeltaWave and a member here.

So be nice
 
OP
M

mehmethand

Member
Joined
Mar 19, 2021
Messages
10
Likes
4
FYI - @pkane is the author of DeltaWave and a member here.

So be nice
It is great to hear that. Deltawave is my main tool for my study even I don't understand many things on this great software. I have limited knowledge about digital audio. pkane is a genius for me and very helpful. His knowledge and experience on this subject is far beyond mine. I'm a student at a univercity at Istanbul/Turkey and I'm working on a project where I need to compare a lot of audio files and I need to show if the sound difference between these files are audible or not. At least to find a probability for being audible. So I'm just searching for a shortcut without doing hundreds of blind abx tests and asking for help from professional guys here.
 

dc655321

Major Contributor
Joined
Mar 4, 2018
Messages
1,597
Likes
2,235
It is great to hear that. Deltawave is my main tool for my study even I don't understand many things on this great software. I have limited knowledge about digital audio. pkane is a genius for me and very helpful. His knowledge and experience on this subject is far beyond mine. I'm a student at a univercity at Istanbul/Turkey and I'm working on a project where I need to compare a lot of audio files and I need to show if the sound difference between these files are audible or not. At least to find a probability for being audible. So I'm just searching for a shortcut without doing hundreds of blind abx tests and asking for help from professional guys here.

Maybe you could write a wrapper around DeltaWave to batch process a list of files?
Detect those files with audibility metrics as recommended and use that much smaller set in blind testing?

Dunno... Just a thought.
 

Wes

Major Contributor
Forum Donor
Joined
Dec 5, 2019
Messages
3,843
Likes
3,788
It is great to hear that. Deltawave is my main tool for my study even I don't understand many things on this great software. I have limited knowledge about digital audio. pkane is a genius for me and very helpful. His knowledge and experience on this subject is far beyond mine. I'm a student at a univercity at Istanbul/Turkey and I'm working on a project where I need to compare a lot of audio files and I need to show if the sound difference between these files are audible or not. At least to find a probability for being audible. So I'm just searching for a shortcut without doing hundreds of blind abx tests and asking for help from professional guys here.

you might want to read about using sparse matrix theory to "search" thru a factor space with a reduced number of points

I forget the title, but if you start here:
https://williamghunter.net/statisticsforexperimenters/

you'll run into it.
 

CMOT

Active Member
Joined
Feb 21, 2021
Messages
147
Likes
114
I don't know, I think some of the above answers kind of missed the point (or maybe I did). Since there is a great deal individual variability in hearing (not just in "skills" but in the actual, physical sensory systems), it is hard to specify exactly what differences might never be detectable by any human being. There are super face recognizers, there are super tasters (slightly different, but not really - a physiological difference leads to the effect), there seems to be a small population of people that can differential many more colors (probably due to a fourth cone type). So one can imagine that some humans have some physiological differences that all "bat-like" hearing - or something like that. Perception is all in the head in one sense, so if there is a signal difference, it might be detectable. Instead you just have say, given the hearing abilities of a good chunk of the human species (where good chunk is defined at whatever mean and distribution you decide), this difference won't be detectable. But there is always a chance that my differences run aground on the particular individual on the tail of the distribution or with genuine physiological differences... [there is wonderful work on why some monkeys have two cone types and others have three cone types, plus why humans have three cone types]
 

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,632
Likes
10,205
Location
North-East
Maybe you could write a wrapper around DeltaWave to batch process a list of files?
Detect those files with audibility metrics as recommended and use that much smaller set in blind testing?

Dunno... Just a thought.

In the next version, DeltaWave will accept command line arguments that can be used to pass in the reference and comparison files. It will then run the comparison, write the results to a file, and exit. This you can use in a batch file or in a script to run comparisons automatically. This is what I did to process multiple files from Gearslutz loopback thread: https://deltaw.org/gearslutz.html
 

dc655321

Major Contributor
Joined
Mar 4, 2018
Messages
1,597
Likes
2,235
so if there is a signal difference, it might be detectable. Instead you just have say, given the hearing abilities of a good chunk of the human species (where good chunk is defined at whatever mean and distribution you decide), this difference won't be detectable.

Yes, such perceptual boundaries can only be specified in probabilistic terms.
 

levimax

Major Contributor
Joined
Dec 28, 2018
Messages
2,348
Likes
3,462
Location
San Diego
I want to ask a question. Let's say you have 2 audio files to compare and they sound very similar. You can't spot and hear the difference with ABX blind test. "You can't hear the difference" doesn't mean that other people can't hear it,too. Hearing depends on where you listen, how you listen, how loud you listen, your hearing skills... etc and all these change from person to person. So here is my question :

You have 2 very similar audio files. You can't hear the sound difference between them and you want to know if the sound difference between these files have a chance to be heard by some other people. Is there anyway to make an assumption about the difference between 2 compared files is hearable or not ? Is there some kind of threshold any kind of analytic data or scientfic research where you can say this difference is hearable between 2 compared audio files?

With audio analyzer softwares you can have rms, peak,lufs delta... like values or have spectrograms, spectrum analysis graphs... Is there any kind of information hidden in there to define a threshold for hearblity betwwen 2 compared audio files?

Note : I'm using deltawave audio null comparator software for this kind of comparison. https://deltaw.org/
~ -60 dB
 

Chrispy

Master Contributor
Forum Donor
Joined
Feb 7, 2020
Messages
7,751
Likes
5,910
Location
PNW
It is great to hear that. Deltawave is my main tool for my study even I don't understand many things on this great software. I have limited knowledge about digital audio. pkane is a genius for me and very helpful. His knowledge and experience on this subject is far beyond mine. I'm a student at a univercity at Istanbul/Turkey and I'm working on a project where I need to compare a lot of audio files and I need to show if the sound difference between these files are audible or not. At least to find a probability for being audible. So I'm just searching for a shortcut without doing hundreds of blind abx tests and asking for help from professional guys here.

Just what sort of class is this project for?
 
OP
M

mehmethand

Member
Joined
Mar 19, 2021
Messages
10
Likes
4
Just what sort of class is this project for?

I live in Istanbul/Turkey. There is not enough sources in our native language about the comparison of equipments and softwares used in music production in my country.
My study is making a document about analysis and comparison of the effects of equipment and software used in music production on sound quality in my native language. I'm comparing 9 DAWs (reaper, pro tools, cubase....etc) in terms of pan law differences, automation, summing, fade scaling, 3rd part plugin behaviours, codes, sample rates... etc. I'm comparing 10 soundcards which you can buy on music stores in my country in terms of DAC, ADC, Preamp... etc. I'm comparing 4 convertors, 5 preamps, 5 different brand cables, 10mics and some more stuff. So I have tons of audio files and I'm trying to figure out a way to eliminate some of those comparisons by saying the difference is inaudible without making blind tests with 50 different producers from my country.
 

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,632
Likes
10,205
Location
North-East
I live in Istanbul/Turkey. There is not enough sources in our native language about the comparison of equipments and softwares used in music production in my country.
My study is making a document about analysis and comparison of the effects of equipment and software used in music production on sound quality in my native language. I'm comparing 9 DAWs (reaper, pro tools, cubase....etc) in terms of pan law differences, automation, summing, fade scaling, 3rd part plugin behaviours, codes, sample rates... etc. I'm comparing 10 soundcards which you can buy on music stores in my country in terms of DAC, ADC, Preamp... etc. I'm comparing 4 convertors, 5 preamps, 5 different brand cables, 10mics and some more stuff. So I have tons of audio files and I'm trying to figure out a way to eliminate some of those comparisons by saying the difference is inaudible without making blind tests with 50 different producers from my country.

As others have said, -50dB is just the start of the range. Some people will not hear distortions at this level, while others may hear to -70dBFS or lower. It's a range, a bell curve. And while DeltaWave takes into account some of the well-known perception metrics in PK Metric, it's certainly not accounting for everything.
 

Chrispy

Master Contributor
Forum Donor
Joined
Feb 7, 2020
Messages
7,751
Likes
5,910
Location
PNW
I live in Istanbul/Turkey. There is not enough sources in our native language about the comparison of equipments and softwares used in music production in my country.
My study is making a document about analysis and comparison of the effects of equipment and software used in music production on sound quality in my native language. I'm comparing 9 DAWs (reaper, pro tools, cubase....etc) in terms of pan law differences, automation, summing, fade scaling, 3rd part plugin behaviours, codes, sample rates... etc. I'm comparing 10 soundcards which you can buy on music stores in my country in terms of DAC, ADC, Preamp... etc. I'm comparing 4 convertors, 5 preamps, 5 different brand cables, 10mics and some more stuff. So I have tons of audio files and I'm trying to figure out a way to eliminate some of those comparisons by saying the difference is inaudible without making blind tests with 50 different producers from my country.

Thanks...so this is for a technically oriented audio course for recording?
 
Top Bottom