24bit vs 16bit, the myth exploded!
Nov 7, 2019 at 10:51 AM Post #5,101 of 6,198

ALRAINBOW

Headphoneus Supremus
Joined
Oct 5, 2012
Posts
4,139
Likes
257
Aside from pages of math, at what point Do we use our ears to allow for subjective listening? I can and most people with good ears can pick out 24 over 16 bit depth. I’m not Alone in this. I follow all the math and it’s conjecture with it but what do we hear ? If you don’t hear it there are two reasons
One being the system used is subpar and can’t show what many can hear.
Second is your not noticing the change.
to fill up pages of math and all of its own issues is futile.
 
Nov 7, 2019 at 11:38 AM Post #5,102 of 6,198

castleofargh

Sound Science Forum Moderator
Joined
Jul 2, 2011
Posts
9,394
Likes
4,780
Aside from pages of math, at what point Do we use our ears to allow for subjective listening? I can and most people with good ears can pick out 24 over 16 bit depth. I’m not Alone in this.
indeed you're not alone claiming to be able to. among you guys, those who produce evidence of being able to consistently pick up the 24bit file represent a handful at best. add a criteria where those people have enough control over their rig to ensure that the difference is caused by the bit depth and not some crap going on in the computer, DAC, or the masters being literally different ones, and now we're lucky if we can find 1 guy on the entire forum who can tell anything consistently at normal listening level and a few of his favorite albums.
that intriguing disparity seems to suggest that most people saying they can tell the difference are full of crap and/or too lazy to verify that they actually can do something before bragging about doing it all over the web.
 
Nov 7, 2019 at 11:51 AM Post #5,103 of 6,198

ALRAINBOW

Headphoneus Supremus
Joined
Oct 5, 2012
Posts
4,139
Likes
257
A well said reply really. I’m not special except maybe special Ed lol. But I do wonder is it hard. I have had many in my room to show and pretty fast its obvious to them. Now is my setup a causing it is a good Question. but it’s not thst Simple I’m was a headphone guy and even up til now can use them or even good CIEMS ona good DAP. But I do wonder what I’m hearing.
to me it’s a blacker background and a seemingly lower noise in the track. Less air but still has presence. I thank you for not going right into bashing lol. These forums are my first home in audio.
thanks all.
 
Nov 7, 2019 at 12:17 PM Post #5,104 of 6,198

bigshot

Headphoneus Supremus
Joined
Nov 16, 2004
Posts
22,697
Likes
4,524
Location
Hollywood USA
I'd suggest getting an independent person who understands the concept to administer the test to you. I'd also check to see if you can tell the difference on different equipment. You may just be hearing some kind of artifacting going on. It might be worthwhile to take a 16 bit recording and bump it up to 24 to see if you're hearing a difference in the file or the recording. When you get unexpected results, that's when you want to test further- tightening controls, checking variables, getting independent verification...
 
Last edited:
Nov 7, 2019 at 12:24 PM Post #5,105 of 6,198

71 dB

Headphoneus Supremus
Joined
Sep 17, 2017
Posts
1,554
Likes
522
Location
Helsinki, Finland
I can and most people with good ears can pick out 24 over 16 bit depth.

Should be pretty much impossible in blind tests using reasonable listening levels (so that we are not listening to the 16 bit dither amplified so much you can hear it). If the master is the same, 24 bit and 16 bit should sound the same. Make yourself a properly dithered 16 bit version of a 24 bit file (so the master is the same).Use an audio editor to produce the difference of 16 bit and 24 files. That's the 16 bit dither, because proper dither removes quantization error completely and replaces it with dither noise. (noise that correlates with the signal is replaced with a little louder noise which doesn't correlate with the signal). Can you hear the this dither (difference signal) using reasonable listening levels? Should be pretty hard to hear anything, because the level of 16 bit dither noise without shaping is something like -95 dBFS and even if shaped dither can go up to -70 dBFS it's perceived quieter because of the properties of human hearing, shape of equal loudness levels. Even if you hear something (you should not), this is the ideal situation to hear the difference of 24 bit and 16 bit, because there is nothing else, just the difference. When you listen to a 16 bit file, the signal itself masks the dither making it totally impossible to hear in any listening scenario that makes sense.

I’m not Alone in this. I follow all the math and it’s conjecture with it but what do we hear ? If you don’t hear it there are two reasons One being the system used is subpar and can’t show what many can hear. Second is your not noticing the change. to fill up pages of math and all of its own issues is futile.

The "critical limit" of bit depth in digital audio is about 13 bits, about 20 dB more dynamic range than vinyl at best. That's how much dynamic range you want in consumer audio and what you need to serve all reasonable listening scenarios. 16 bit audio gives you 3 bit (18 dB) worth of safety margin over the critical limit so anyone who understands digital audio can sleep well at night knowing their 16 bit music has enough dynamic range (well, the recording itself may not have, but the media certainly has!). Placebo-effect plays a huge role in audio. You have also the issue of different masters for different bit depths and even your hardware can handle different file formats differently. A 24/44.1 kHz version may sound the best becasue it is the best master, but it sounds just as good when converted properly into 16 bit and played on gear that treats 24 bit and 16 bit the same way and the listener doesn't know about the bits so that placebo effect can't function. Then we are in a situation the math describes.
 
Nov 7, 2019 at 2:12 PM Post #5,106 of 6,198

gregorio

Headphoneus Supremus
Joined
Feb 14, 2008
Posts
4,058
Likes
2,172
[1] Aside from pages of math, at what point Do we use our ears to allow for subjective listening?
[2] I can and most people with good ears can pick out 24 over 16 bit depth. I’m not Alone in this.
[3] I follow all the math and it’s conjecture with it but what do we hear ?
[4] If you don’t hear it there are two reasons: One being the system used is subpar and can’t show what many can hear. Second is your not noticing the change.
[5] to fill up pages of math and all of its own issues is futile.

1. At the point that the reproduced sound enters your ears.

2. This is the Sound Science sub-forum. If you are going to make factual assertions, they either need to agree with the science or contradict/disagree with it but ONLY if you have reliable evidence to support your assertion. Unfortunately, your assertion both contradicts the demonstrated science and you have presented no reliable supporting evidence! The demonstrated science shows that at normal/reasonable listening levels no one, good ears or not, can "pick out 24 over 16 bit depth" with commercial (competently made) music audio products. Unless, you can provide some reliable evidence to the contrary, the ONLY logical conclusion is that you arrived at this assertion simply because it has been suggested to you (in false marketing for example) or you have performed some seriously flawed test yourself. Furthermore, this conclusion has been demonstrated countless times: Audiophiles often claim they can hear the difference but without exception, on those occasions when they "put their money where their mouth is" and are reliably tested, they can't. Most commonly, it turns out they were actually comparing different masters and falsely attributing the audible difference to 24 vs 16bit.

3. This statement is also false unfortunately. The math was conjecture over 90 years ago but over 70 years ago was proven (by Claude Shannon) and therefore was no longer a conjecture. Furthermore, the entire digital age depends on that proof and it it were wrong, the digital age would not exist.

4. This too is a common falsehood peddled by some/many audiophiles. It is easily disproven by the fact that the very best systems (top commercial studio systems) and highly trained, experienced engineers can't hear the difference (given the conditions above). How then is it possible that untrained amateurs, using audiophile systems that are subpar (compared to top pro studios) are hearing a difference?

5. It would be futile if both: A. The math wasn't proven and demonstrated in practice in every digital device on the planet and B. Someone chooses to believe the marketing hype and audiophile myth and dismiss the demonstrated science. This isn't the "marketing hype and audiophile myth" sub-forum though, it's the sound science sub-forum and therefore it isn't futile, except of course to those who come here by error, thinking it is the "marketing hype and audiophile myth" forum!!
But I do wonder what I’m hearing ... I thank you for not going right into bashing lol.
Don't you think you should find out what you're hearing BEFORE making the assertion (in a sound Science forum) that what you're hearing is a difference between 24bit and 16bit AND (falsely) "bashing" others' equipment and/or listening skills?

G
 
Nov 7, 2019 at 2:46 PM Post #5,107 of 6,198

ALRAINBOW

Headphoneus Supremus
Joined
Oct 5, 2012
Posts
4,139
Likes
257
I’m not bashing and my comment on conjecture is based on math and then hearing it. If this thread is purely math then why is it a myth exploded lol.
even after your comment on it being a fact for 70 years. Geezzz just play and comment then post of math to comment on what’s heard !
 
Nov 7, 2019 at 2:48 PM Post #5,108 of 6,198

ALRAINBOW

Headphoneus Supremus
Joined
Oct 5, 2012
Posts
4,139
Likes
257
1. At the point that the reproduced sound enters your ears.

2. This is the Sound Science sub-forum. If you are going to make factual assertions, they either need to agree with the science or contradict/disagree with it but ONLY if you have reliable evidence to support your assertion. Unfortunately, your assertion both contradicts the demonstrated science and you have presented no reliable supporting evidence! The demonstrated science shows that at normal/reasonable listening levels no one, good ears or not, can "pick out 24 over 16 bit depth" with commercial (competently made) music audio products. Unless, you can provide some reliable evidence to the contrary, the ONLY logical conclusion is that you arrived at this assertion simply because it has been suggested to you (in false marketing for example) or you have performed some seriously flawed test yourself. Furthermore, this conclusion has been demonstrated countless times: Audiophiles often claim they can hear the difference but without exception, on those occasions when they "put their money where their mouth is" and are reliably tested, they can't. Most commonly, it turns out they were actually comparing different masters and falsely attributing the audible difference to 24 vs 16bit.

3. This statement is also false unfortunately. The math was conjecture over 90 years ago but over 70 years ago was proven (by Claude Shannon) and therefore was no longer a conjecture. Furthermore, the entire digital age depends on that proof and it it were wrong, the digital age would not exist.

4. This too is a common falsehood peddled by some/many audiophiles. It is easily disproven by the fact that the very best systems (top commercial studio systems) and highly trained, experienced engineers can't hear the difference (given the conditions above). How then is it possible that untrained amateurs, using audiophile systems that are subpar (compared to top pro studios) are hearing a difference?

5. It would be futile if both: A. The math wasn't proven and demonstrated in practice in every digital device on the planet and B. Someone chooses to believe the marketing hype and audiophile myth and dismiss the demonstrated science. This isn't the "marketing hype and audiophile myth" sub-forum though, it's the sound science sub-forum and therefore it isn't futile, except of course to those who come here by error, thinking it is the "marketing hype and audiophile myth" forum!!

Don't you think you should find out what you're hearing BEFORE making the assertion (in a sound Science forum) that what you're hearing is a difference between 24bit and 16bit AND (falsely) "bashing" others' equipment and/or listening skills?

G
I don’t need to know first I can easily discern them Hahahaha. But joking aside it’s easy and can be duplicated like math but math and it’s effects can’t be duplicated always
 
Nov 7, 2019 at 3:21 PM Post #5,109 of 6,198

gregorio

Headphoneus Supremus
Joined
Feb 14, 2008
Posts
4,058
Likes
2,172
[1] I’m not bashing and
[2] my comment on conjecture is based on math and then hearing it.
[3] If this thread is purely math then why is it a myth exploded lol. [3a] even after your comment on it being a fact for 70 years.
[4] Geezzz just play and comment then post of math to comment on what’s heard !

1. Clearly you were "bashing", you stated that anyone who can't hear the difference had either subpar equipment or listening skills. You think maybe that stating others have subpar equipment or listening skills is a compliment?
2. No it's NOT, your comment is false, the math is not a conjecture, it's a proven theorem!
3. Because many audiophiles believe the (false) myth that 24bit is audibly different to 16bit under the conditions mentioned.
3a. Yes, because many audiophiles don't know or don't understand the proven math and instead believe the marketing hype that contradicts it!
4. Why? The "math" dictates what is recorded and reproduced with 24bit vs 16bit, what you think you hear or "wonder what you're hearing" is not covered by the math, as I already effectively stated in my point #1 previously!
[1] I don’t need to know first I can easily discern them Hahahaha.
[2] But joking aside it’s easy and can be duplicated like math but math and it’s effects can’t be duplicated always

1. You stated "But I do wonder what I’m hearing" and now you're stating you know exactly what you're hearing, which is it? Hahahaha.
2. Sure, it's trivially easy to play (for example) two different masters and discern a difference, that's the whole point! What would be the point of paying studio time and an engineer to make another different master which sounded exactly the same? Unless you can eliminate this (and other potential audible differences), then you cannot truthfully assert the difference you are hearing/perceiving is due to the difference in bit depth.

G
 
Nov 7, 2019 at 4:03 PM Post #5,110 of 6,198

bfreedma

The Hornet!
Joined
Feb 3, 2012
Posts
2,631
Likes
1,715
I don’t need to know first I can easily discern them Hahahaha. But joking aside it’s easy and can be duplicated like math but math and it’s effects can’t be duplicated always

Short answer: In this forum, sighted subjective opinions are not an acceptable substitute for valid testing protocols. As suggested before, try assessing the difference again ensuring the masters are the same and the test is properly proctored.
 
Nov 7, 2019 at 5:25 PM Post #5,111 of 6,198

bigshot

Headphoneus Supremus
Joined
Nov 16, 2004
Posts
22,697
Likes
4,524
Location
Hollywood USA
I don’t need to know first I can easily discern them Hahahaha.

Well, your posts don't exactly inspire confidence in the truth of what you say... I think you don't know what you're talking about, and you're in the wrong forum to try to bluff your way through.
 
Nov 7, 2019 at 6:20 PM Post #5,112 of 6,198

old tech

500+ Head-Fier
Joined
Jul 8, 2015
Posts
773
Likes
378
Location
Sydney, Australia
Ever wonder why it is necessary for the manufacturer to put a hi-res light or indicator on your playback device if there is a clear difference in sound between 16 and 24 bits? Perhaps the visual is what you are hearing?
 
Nov 8, 2019 at 12:51 AM Post #5,113 of 6,198

castleofargh

Sound Science Forum Moderator
Joined
Jul 2, 2011
Posts
9,394
Likes
4,780
I don’t need to know first I can easily discern them Hahahaha. But joking aside it’s easy and can be duplicated like math but math and it’s effects can’t be duplicated always
me too, I just read the name of the file I'm going to play and instantly can tell that the 24/96 tag means hires. easy.
 
Nov 15, 2019 at 10:49 AM Post #5,114 of 6,198

billbishere

100+ Head-Fier
Joined
Apr 3, 2015
Posts
485
Likes
316
Location
Columbus, Ohio
Surprised to see this thread still going strong for 10 years . I was recently trying to get a better grasp on 24bit since qobuzz is cheaper than tidal now
 
Nov 15, 2019 at 11:58 AM Post #5,115 of 6,198

castleofargh

Sound Science Forum Moderator
Joined
Jul 2, 2011
Posts
9,394
Likes
4,780
Surprised to see this thread still going strong for 10 years . I was recently trying to get a better grasp on 24bit since qobuzz is cheaper than tidal now
the short summary goes like this:
there is nothing wrong with 24bit files, but there are many issues with the beliefs triggered by reading "24bit" somewhere.
 

Users who are viewing this thread

Top