Why 24 bit audio and anything over 48k is not only worthless, but bad for music.

[eq1849]

EQ, why don't you tell me all of this?

we are waiting , Dave…

Please enlighten all of those who lack understanding of bit depth and DR

 

[bigshot]

The Realiser has head tracking, so does Apple Spatial Audio. We’ll see these kinds of DSPs sooner than we think.

 

[Davesrose]

we are waiting , Dave…

Please enlighten all of those who lack understanding of bit depth and DR

No sorry, the impetuous is on you as you keep claiming bit depth is not about DR.

 

[BS5711]

After listening to 24 bit for 15-20 minutes then switching to 16, the 16 sounds brighter and less full.

I also passed the three blind listening tests I made regarding 16 vs 24 bit.

I disagree with your opinion regarding 24 192 vs 16 44.1.

We also fortunately have brain science research where higher sampling rates with inaudible frequencies show differences, via EEG.

Our mind can play all sorts of tricks in normal listening scenarios so your first comment is not really very meaningful.

We have no idea what blind tests you passed so that isn't of great value either really.

Given the right circumstances with high volume and silence in a track I could likely pass a blind test also but that doesn't have much value in normal listening.

It is feeling like your assertion of superior sound might ultimately boil down to subjective assessment on your part and we all know how unreliable that can be.

Anyway, good luck to you with this argument, I recall you had a crack at this previously, just one in a long line of people that think they have outsmarted the science that created digital audio in the first place.

 

[bigshot]

Resolution and timing are also at play.

Resolution and timing are equal in 16/44.1 and high rate audio. The differences are frequency range and the depth of the noise floor. Both are audibly transparent and sound the same to human ears.

 

[eq1849]

No sorry, the impetuous is on you as you keep claiming bit depth is not about DR.

Hi. Bit depth does determine DR. About 96 db in 16 bit and about 144 db in 24 bit.


Now with that out of the way, back to this:

Dave how many quantization steps are there in the 23rd bit of 24 bit audio?

What decibel range does this cover?

How many quantization steps are there in the 15th bit of 16 bit audio?

What decibel range does this cover?

Mathematically, does the 23rd bit in a 24 bit signal capture smaller voltage amplitude values than the 15th bit of 16 bit audio?

What about the 24th bit of a 24 bit signal versus the 16th bit in a 16 bit signal? More or less precision?

 

[Davesrose]

Hi. Bit depth does determine DR. About 96 db in 16 bit and about 144 db in 24 bit.


Now with that out of the way, back to this:

Dave how many quantization steps are there in the 23rd bit of 24 bit audio?

What decibel range does this cover?

How many quantization steps are there in the 15th bit of 16 bit audio?

What decibel range does this cover?

Mathematically, does the 23rd bit in a 24 bit signal capture smaller voltage amplitude values than the 15th bit of 16 bit audio?

What about the 24th bit of a 24 bit signal versus the 16th bit in a 16 bit signal? More or less precision?

EQ, do you know what decibel range means??

 

[eq1849]

EQ, do you know what decibel range means??

Dave, for the third time can you answer the following or not?


Dave how many quantization steps are there in the 23rd bit of 24 bit audio?

What decibel range does this cover?

How many quantization steps are there in the 15th bit of 16 bit audio?

What decibel range does this cover?

Mathematically, does the 23rd bit in a 24 bit signal capture smaller voltage amplitude values than the 15th bit of 16 bit audio?

What about the 24th bit of a 24 bit signal versus the 16th bit in a 16 bit signal? More or less precision?

 

[Davesrose]

Dave, for the third time can you answer the following or not?


Dave how many quantization steps are there in the 23rd bit of 24 bit audio?

What decibel range does this cover?

How many quantization steps are there in the 15th bit of 16 bit audio?

What decibel range does this cover?

Mathematically, does the 23rd bit in a 24 bit signal capture smaller voltage amplitude values than the 15th bit of 16 bit audio?

What about the 24th bit of a 24 bit signal versus the 16th bit in a 16 bit signal? More or less precision?

EQ, again what is decibel range? You yourself tried to answer DR of 16bit vs 24bit, so it really shouldn't be a hard answer. As far as quantization, the basic formula is 2^number of bits. That means 16 bits has 65,536 levels. That's more than enough for reproducing a sine wave.

 

[eq1849]

EQ, again what is decibel range? You yourself tried to answer DR of 16bit vs 24bit, so it really shouldn't be a hard answer. As far as quantization, the basic formula is 2^number of bits. That means 16 bits has 65,536 levels.

dave, how many levels are in the 24th bit of 24 bit audio?

How many levels are in the 23rd bit of 24 bit audio?

Does the sampling theorem call for 65,536 quantization steps? If not, how many does ST call for?

 

[Davesrose]

dave, how many levels are in the 24th bit of 24 bit audio?

How many levels are in the 23rd bit of 24 bit audio?

Does the sampling theorem call for 65,536 quantization steps? If not, how many does ST call for?

EQ, I gave you the formula for figuring out number of levels for every bit depth.....why are you continuing with this asinine questioning of how many levels for whatever bit depth? When it comes to the Nyquist Theorem, the maximum sample rate is 40kHz. Do you think if I give the number for 24 bits....it will be higher, and higher is always better? Yes, we must have the ability to raise the knob to 11!

 

[71 dB]

There is nothing wrong with stating your opinion; however,I think you should make it clear this is your opinion.

Except this is sound science forum and we are supposed to state scientific facts rather than opinions. I admit that the 13 bit is an "opinion", but it is based on science of human hearing and practicalities. Anyway, the limit is close to 13 bit and the point is 16 bit digital audio provides enough dynamic range in consumer audio. In studios, in music production, 24 bit gives flexibility and beneficial safety margin. As for sample rates go, it is math/science that once your sample rate is at least twice the highest frequencies of your signal you are good and anything above that is redundant. This is not just my opinion in any way. It is a fact I learned in university.

I do think you tend to be falling into the mindset of bit equaling DR and sample rates equalling frequency.

Huh?

Resolution and timing are also at play.

Are you sure you understand these things yourself? Someone without deeper understanding of the topic falls easily into the intuitive assumption that larger numbers in digital audio must mean better resolution and in many ways that is indees true, but without understanding one can't know how much resolution is even needed.

Digital audio at 22050 Hz sampling frequency and 8 bit of dynamic range does NOT have enough resolution for transparency.
Digital audio at 44100 Hz sampling frequency and 16 bit of dynamic range does have enough resolution for transparency.
Digital audio at 96000 Hz sampling frequency and 24 bit of dynamic range is overkill.

These come from the properties of human hearing and the practicalities of music consumption. 24 bits of dynamic range (144 dB) is the difference of a cannon nearby and a mosquito.

When it comes to the Nyquist Theorem, the maximum sample rate is 40kHz.

You must mean minimum.

Does the sampling theorem call for 65,536 quantization steps? If not, how many does ST call for?

No. It doesn't call for any quantization steps. It only tells us how high the sampling frequency has to be. The properties of human hearing and music consumption practicalities call for about 8,000-10,000 steps (13 bit).

 

[gregorio]

How much better can sound reproduction get, are we almost at the pinnacle or is there still some way to go, can we expect major improvements in the future?

Still quite a long way, although that’s mainly due to transducers and acoustics, and binaural processing in the case of HPs.

I test for effects of the 16 bit format(brighter, shallower soundstage depth, grey background) and 24 bit(warmer fuller, dark background) rather than hearing a difference, imo.

How do you test for those things if you don’t hear a difference?

I also test for longer periods of time, say listening to 24 bit for 20 minutes casually then switching down to 16, where the brightness, lack of depth can be more easily perceived.

There is no difference in brightness between 24bit and 16bit to be perceived.

Why is it so hard to grasp for so many that but depth is not just DR.

For the same reason it’s so hard for us to grasp that the Earth is flat and that 1+1=3.

There are around 65,000 levels to capture voltage amplitude in 16bit.
Almost 17,000,000 levels in 24 bit.

And just 2 levels in the case of SACD, so what?

If only they would measure the brain itself with, say, Electroencephalography(EEG), to see what’s happening in the brain while listening to various sampling rates to see if there are effects or not…

An EEG measures the electrical activity of the brain, not “what’s happening in the brain”! There is no way currently to measure what is happening; what we’re hearing/perceiving. It is irrelevant whether there are any electrical effects or not, what’s relevant is whether there are any perceivable effects.

[1] Dave how many quantization steps are there in the 23rd bit of 24 bit audio? [1a] What decibel range does this cover?
[2] How many quantization steps are there in the 15th bit of 16 bit audio? [2a] What decibel range does this cover?
[3] Mathematically, does the 23rd bit in a 24 bit signal capture smaller voltage amplitude values than the 15th bit of 16 bit audio?
[4] What about the 24th bit of a 24 bit signal versus the 16th bit in a 16 bit signal? More or less precision?

As you seem so determined, I’ll answer:
1. Two. 1a. ~6.02dB
2. Two. 2a. ~6.02dB
3. Are you talking about voltage amplitudes or audio? Theoretically, the 23rd bit could capture a smaller voltage amplitude than the 15th bit but as far as audio (music) is concerned, then the answer is: “The same”.
4. Same precision.

You don’t seem to have any understanding of how digital audio works or indeed of what voltage amplitudes or sound/audio actually exist. Bit depth defines the theoretical dynamic range of what could be captured WITHOUT considering dither. It is therefore somewhat misleading from a practical standpoint because digital audio is always dithered.

G

 

[sander99]

@eq1849: You don't remember we had this same discussion with you at least 5 times?

 

[71 dB]

@eq1849: You don't remember we had this same discussion with you at least 5 times?

That explains the Déjà vu I have been experiencing!