Can someone please explain to me why and how a 24bit/44.1kHz file is considered high resolution

I don’t mean to start a flame war nor a long contentious discussion about the merits of high resolution versus CD resolution or, god forbid, MQA rather I’m trying to find out if my understanding of the function of bit depth is somehow wrong or flawed. I’ve also done online research and the internet is just filled with flat out incorrect information so please double check all your information sources before posting a link. Thanks!

So here is how I understand bit depth, as opposed to sampling frequency - which I think I understand correctly. Please do not nit pick my choice of words but rather focus on the overall concept that I’m trying to explain and grasp.

The bit depth is used to enable a range of values for the amplitude of a given sampled frequency - the higher the bit depth, the greater the range of values available for the frequency. And this in turn gives the different dynamic range available for a different bit depths. For 16 bit the maximum dynamic range is 96dbB and for 24bit the maximum dynamic range increases to 144dB.

Now comes the trick part - dynamic range is a function of the audio signal being recorded and not the bit depth of the file. What I mean is that if the music itself is very quiet like a classical piano trio then that music may only have a dynamic range of 70dB or less (the range from silence to loudest passage in the music). I am also over simplifying since I’m leaving out things like noise floor, to name but one.

So in the case of a quiet piano trio going from 16bit/44.1kHz to 24bit/44.1kHz is not going to increase the dynamic range of the music.

Now I also understand that there other reasons for using 24bit versus 16bit, especially during the recording, mixing and mastering of the music, but I’m strictly talking about the end result, aka the file that we actually listen to.

My very simple analogy is to that of a shipping crate - a bigger crate means that one can fit more items or a bigger item in the crate. So bit depth is the size of the crate. And in this analogy the audio or music would be the item that goes in the crate. Now just as putting a given item into a bigger crate does not increase the size of the item, so putting a given audio/music signal/recording into a file with a greater bit depth does not increase the dynamic range of the music.

All of the above leads me to believe that calling a 24bit/44.1kHz file “high resolution” is simple marketing BS. Notice I’m not daring to go near the sampling frequency discussion since I believe that area is not as cut and dried as the bit depth area.

Please feel free to correct me but if you reference any article that shows anything like this and claims that it’s true, I will simply ignore it since I’m trying to understand what actually going on and I’m not interested in marketing BS:

Thanks in advance and please let’s try and keep the discussion on a civil tone.


Isn’t calling anything “high resolution” simply marketing BS? I mean, what’s the objective definition of “high resolution”?


While you are probably correct, that’s exactly the type of discussion I’m trying to avoid. But that may well be impossible on this community board.

1 Like

You could perhaps think of bit depth as more like the ratio of the overall size of a sieve to the size of its holes: the greater the bit depth, the greater the ratio between the largest and smallest items the sieve can hold.

Agree with you on some music being well within the capacity of 16 bit. But I had read a very serious and technical report some time ago (will post it here if I can find it) which claimed human ear can literally NOT distinguish the difference from 96kHz onwards but 24 bit dynamic range makes a little audible difference to 16 bits. So he basically claimed nothing can technically ever sound better than 24 bit/96 kHz to us. Hence your point about high res music being a bit BS!

High-resolution audio [Wikipedia]:

The [industry groups] formulated the following definition of high-resolution audio in 2014: “lossless audio capable of reproducing the full spectrum of sound from recordings which have been mastered from better than CD quality (48 kHz/20-bit or higher) music sources which represent what the artists, producers and engineers originally intended.”

1 Like

In practice, -144 DBfs is way below the noise floor of even the best audio electronics available.

At 16 bits (-96 DBfs), you (the recording engineer) typically want to dither (randomly flip the least-significant bit of) the output to push the quantization noise up to frequencies that humans are less sensitive to (this is also known as noise-shaping). At 24 bits, there’s no need to dither, as other sources of noise swamp the quantization noise.

Recording in 24 bit lets you:

  1. be slopp(ier) in setting the recording level, without risk of setting it too high and having the signal clip, or too low and having the quietest passages lost in the noise.
  2. do all your digital editing, DSP, etc, without loss of real resolution.

At the end of the day, you can adjust the level and reduce to 16 bits, with no loss of detail. As a distribution format, 24 bit is just a waste of space.

That waste of space is exploited by MQA to hide some additional (higher sample-rate) information in the 8 least-significant bits of the 24 bit samples (reducing the actual bit depth to 16 bits).

So you can believe at most 1 of two things

  1. A bit depth of 24 bits is audible/necessary for “high res” or
  2. MQA works (i.e. delivers something resembling a “high res” experience).

(The smart money is on neither being true.)


I remember Bob Stuart saying 16 bit is barley enough but 24 bit is overkill. 20 bit will allow for full dynamic range but it’s easy for computers to do 24 bit.
You are correct, you only need the box you need and a bigger box with nothing in it won’t make the music any better.

1 Like

I believe that bit depth is not about dynamic range only. It is also about the “precision” of digital wave versus the original analog wave. With higher bit rate, digital approximation of the amplitude becomes closer to the original analog wave.
In fact, sample rate and bit rate are both about “precision” and approximation. Sample rate in working on sound wave length, and bit rate is working on sound wave amplitude.
Check graphs form “Bit depth” section here:

It is very similar with picture/image:

  • picture resolution (size of the smallest square; e.g. 300dpi) is the equivalent of sample rate
  • picture color space deep (maximum color range for each square; e.g. 8-bit, 16-bit etc.) is the equivalent of bit depth.

Depending on picture / music content, some pieces may not have benefits from higher resolution and/or higher deep. Extreme example: digital version of an empty picture or a quiet track will be the same with the original analog piece, no mater the sample rate and bit rate.

I think 16/44.1 is “CD quality.” I guess anything better than that is “high resolution.” IDK.

Exactly the kind of reference I was hoping to avoid - nothing of any value or truth to be found in Wikipedia’s articles on digital audio.

@Jacques_Distler - I believe I mentioned that there are reasons to record, edit and master using 24bits but I’m concerned with the finished product - the file used for actual listening by the consumer.

I don’t quite understand what it is that you are trying to say but what I do understand is correct then I don’t see how this makes any difference to the sound of a 24bit file.

@DanielAvasilichioaei the linked article comes awfully close to showing a stair step :grin: However the article completely ignores the limits and abilities of human hearing. And I’m not sure that bit depth really means more amplitude samples of a given analog audio signal rather than just greater range or samples. So this article could just be marketing BS in more technical clothing. Hard to say for sure so perhaps with a stronger technical background can weigh in.

This is exactly what I’m talking about - 24bit audio is overkill for music with limited dynamic range.

“anything better” - not necessarily
“anything greater” would be more correct.

The marketing BS for high resolution audio has very effective :smiley:

Oh, OK, sorry.

Sounds to me as if you already have a pretty good grip on what’s going on, though. Not sure there’s much to add.

1 Like

“anything better” meaning anything more than 16/44.1. Personally, it has to be better even if not humanly detectable. Otherwise, let’s all cancel our Tidal and Qobuz subscriptions and listen to Apple Music or Pandora using a transistor radio or crystal radio.

1 Like

I do not understand why 24 bits needs to mean 16 bit plus 4 bits, why could it not mean 96db in 24 bits? Would that not be better? Can anybody explain that?

look at our oscilloscopes.

James a reductio ad absurdum argument is not going to get us anywhere. I did not say that high resolution wasn’t worthwhile but rather that 24bit does not make the file high resolution.

First of all 16 + 4 = 20, not 24

And as I said earlier i would like to keep the discussion about the merits of 24bit versus 16bit and not about sampling frequency, which is another whole can of worms.

1 Like

I don’t mean to be flippant but I’m compelled to ask, ‘so what?’

Select the best compromise between required computing resources and accurate audio reproduction. It’s a one size fits all but it’s a decent compromise.

I don’t have any problems with that approach and I’ve long felt that perhaps 24bit/48kHz would be a good compromise. The 24bits would allow for the use of DSP and digital volume control and the 48kHz sampling rate would make the bats happy, not as happy as 96kHz would but happy nonetheless :grin:


I really don’t understand what you’re going on about here, Ralph. “Resolution” is typically all about sampling frequency, so how can we discuss “high resolution” without bringing it up?

So do I have post some links to all the sites that sell music downloads and show you that they, not me, call a 24bit/44.1kHz file “high resolution”? That is exactly my point: 24bits does not make a file high resolution.

I don’t know how long you have been at this audiophile game but I’ve been at it for over 30 years and the one thing that I’ve found to be true is that marketing BS is piled very high and very deep in this hobby. If a 10 gauge wire is “better” than a 12 gauge wire, why then a 4 guage wire must be better still. If 200 watts is “better” than 100 watts then 2,000 watts must be better still. The list of nonsense and marketing BS goes on and on, with the rush to high resolution audio being just the latest one. And I left out jitter, one of my all time favorites!

I read lots of posts on this community by very well meaning individuals with whom I have no quarrel that clearly show that they have swallowed the marketing BS hook, line and sinker. And I generally say nothing since it’s their time and money and they can do what they like with it. However I do feel that the real information, free of marketing BS, should be made available so that others can make more informed buying decisions .

So that’s what I’m going on about. :wink:

And now for my sins I will listen to some mp3 files and hopefully I won’t bleed out through my ears.