24bit vs 16bit, the myth exploded!
Aug 20, 2023 at 3:49 AM Post #7,036 of 7,175
By the way, if anyone is interested, here is a link to a transfer I did of a blue wax French Columbia release from 1927 of Igor Stravinsky conducting The Firebird Suite.
That is indeed interesting.
It's not a particularly great performance,
TBH, it’s a poor performance. Most of it isn’t great, as you state, but in some places it’s really pretty terrible, sounding more like a mediocre school band than a professional orchestra, particularly in regards to timing precision. What orchestra was it?
nor is it a great recording …
Being a wax cylinder I presume it was an acoustic/mechanical recording? It certainly has the balance and distance problems of a horn recording but also has a freq response in places I’d more associate with an electrical (microphone) recording, so I’m not sure what it is. If it was an electrical recording, then it was pretty shoddy mic/mic placement. Don’t suppose you know which?
but it's a document of Stravinsky's first recordings as a conductor.
That’s particularly interesting in Stravinsky’s case. He famously disliked not only individual conductors but pretty much all conductors in general and likewise, he was often criticised as a conductor. He certainly didn’t appear to have a particularly good technique in the later videos I’ve seen and sometimes (though certainly not always) a somewhat emotionless interpretation style. So it’s interesting to hear an early example of him conducting.

G
 
Aug 20, 2023 at 11:06 AM Post #7,038 of 7,175
The topic has been covered a dozen times at least. What else is there to talk about now?
 
Aug 20, 2023 at 11:18 AM Post #7,039 of 7,175
Aug 20, 2023 at 3:17 PM Post #7,040 of 7,175
At this point, "I have lost all ability to can".
 
Aug 20, 2023 at 7:04 PM Post #7,041 of 7,175
The topic has been covered a dozen times at least. What else is there to talk about now?
Well, make a new thread. Most people that come to this thread is because of the topic heading.
 
Aug 20, 2023 at 7:13 PM Post #7,042 of 7,175
Everything they need is in the first post.
 
Aug 21, 2023 at 7:40 AM Post #7,044 of 7,175
460+ pages on 24 vs 16 bit which can be returned to the original topic at any time if anyone has anything new to add.
Even if it was 4600 pages, a lot of people in the World still believed 24 bit offers better sound quality compared to 16 bit in consumer audio. Myths die hard and this 24 bit 16 bit thing isn't really a myth, but something intuition says when knowledge and understanding doesn't say otherwise.

The way to make more people listen to what you have to say is to hone your messaging, something people who know what they are talking about are often neglecting. How to explain this 16 bit vs 24 bit thing to people effectively? What are our "talking points?"

A --- Resolution is a picture thing, not a sound thing.
B --- Fidelity is a sound thing.
C --- Sampling rate dictates highest possible frequency (Nyquist-Shannon sampling theorem)
D --- Bit depth dictates how low the noise floor is (dynamic range) and nothing else.
E --- Dithering adds noise, but makes digital audio 100 % distortion free!
F --- "Staircases" are just an abstract way to illustrate digital signals.
G --- Reconstruction filters in DACs smooth out the signal.
H --- In band-limited signals the signal can take only one route between sample points.
I --- In any given situation the dynamic range of human hearing is about 70 dB (~12 bits)
J --- Children can hear frequencies up to 20 kHz. Adults less than that.
K --- 44.1 kHz sampling rate is enough for the ears of children let alone adults.
L --- 80 dB (~13 bits) of dynamic range is enough in the most demanding consumer audio.
M --- Shaped dithering can increase perceptual dynamic range as much as 20 dB.
N --- 24 bit offers benefits in audio production for technical and practical reasons.
O --- 16 bit is overkill by about 3 bits in consumer audio.
P --- Sound quality comes mostly from recording, producing, mixing and mastering.
Q --- If 24 bit sounds different from 16 bit, it is different master or/and Placebo effect.
R --- 24 bit is used as snake oil in consumer audio as a business model.
S --- 24 bit truncated to 16 bit sounds exactly the same, even without dither!

I think it is important to make clear the difference of picture resolution and sound fidelity clear. Otherwise people use picture resolution as an intuitive model for assuming greater bit depth will always translate into greater sound quality. Since Blu-ray offers much better picture resolution compared to DVD, surely 24 bit audio must be perceptually superior to 16 bit? Well no! We should not compare picture and sound like that. They are different things. Same digital signal processing principles and methods (such as dithering and filtering) can be utilised in both, but the benefits can't be compared just like that. People should be careful when sound is presented in graphical from, because the way sound "looks" is not always the way it sounds. Huge differences in how signals "look" in graphical representation may result in just small or even inaudible difference in how they sound and vice versa.

Any suggestion as to how to improve this messaging?
 
Aug 21, 2023 at 8:06 AM Post #7,045 of 7,175
Any suggestion as to how to improve this messaging?
Yep, join all those A-S points up, so that they don’t appear to be just a bunch of unrelated facts but are inter-related to create an understanding of the thread title.

G
 
Aug 21, 2023 at 9:06 AM Post #7,046 of 7,175
The topic has been covered a dozen times at least. What else is there to talk about now?

Ah yes, you are a master of understatement! :wink:

1692623071660.png
 
Aug 21, 2023 at 11:13 AM Post #7,047 of 7,175
I think all that needs to be said is that 16 bit is already overkill for the purposes of listening to music in the home. 24 bit is just a bigger file size with no added useful benefit.
 
Aug 21, 2023 at 11:36 AM Post #7,048 of 7,175
Yep, join all those A-S points up, so that they don’t appear to be just a bunch of unrelated facts but are inter-related to create an understanding of the thread title.

G
In my opinion they don't look unrelated, but I get your point.
 
Aug 21, 2023 at 11:37 AM Post #7,049 of 7,175
I think all that needs to be said is that 16 bit is already overkill for the purposes of listening to music in the home. 24 bit is just a bigger file size with no added useful benefit.
well, if you believe this:
1. decreased noisefloor doesnt matter (also useful for digital volume attenuation, tho as long the processing happens in 24 bit this doesnt matter much)
2. dac filters dont benefit from higher samplerates
3. there are sometimes different masters for 24 and 16 bit, so it also depends on taste

i think these three are the biggest benefits, also objectively "backed up" in theory

beside this hard drive capacitys are usually so big that file size doesnt matter much

tho in the end like i said sometime earlier everyone should listen and decide for themself, there is no point in debating about this, either it sounds the same to you (or/and on your system) or it doesnt
 
Aug 21, 2023 at 12:28 PM Post #7,050 of 7,175
well, if you believe this:
1. decreased noisefloor doesnt matter (also useful for digital volume attenuation, tho as long the processing happens in 24 bit this doesnt matter much)
2. dac filters dont benefit from higher samplerates
3. there are sometimes different masters for 24 and 16 bit, so it also depends on taste

i think these three are the biggest benefits, also objectively "backed up" in theory

beside this hard drive capacitys are usually so big that file size doesnt matter much

tho in the end like i said sometime earlier everyone should listen and decide for themself, there is no point in debating about this, either it sounds the same to you (or/and on your system) or it doesnt
And if you have an Full Digital Amp, it matters even less.

One thing that is forgotten too, 24bit give you an dynamic range of 144db.

The best measuring amp i've seen so far, the S.M.S.L SH-9, has an dynamic range of 137db, but, of course, A-weight with an 1kHz sine.

The real performace is around 124db with an 1kHz sine with an drop to around 110db in Multitone.

So the realistic performance is somewhere below 110db and last but not least, the Earphone/Headphone have to be able to reproduce that dynamic range. How much gets lost due to reverb and reflection in the ear canal (or even worse, with heapdhones, due to reflections in the cup, your ear, parts of your head and so on and surrounding noise)

Even with an isolating, deep inserting, All-BA (so best case) you will not even be able to reproduce the dynamic range of an Audio CD at the eardrum. And if you would be able, the amp would not be able to get beyond ~105db of Dynamic Range in an realistic listening scenario so there is not much there of the 24bit.

But why even argue, why not just test it yourself?

Here is the difference between 24bit and 16bit

https://drive.google.com/file/d/1b__JAqqWciUFHIagegN53LMNuEmHMa3_/view?usp=sharing

Enjoy the -85db quiet white noise at >=20kHz in combination with the -90db quiet white noise <=20kHz
 

Users who are viewing this thread

Back
Top