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

May 28, 2025 at 2:24 PM Post #3,722 of 3,803
I’m glad we agree.
We didn’t agree as that is NOT the quote I agreed with, which I made entirely clear. You’re just deflecting because you couldn’t answer the question that you quoted or my last question: “So, either you’ve mathematically proven that digital audio doesn’t exist or your math must be completely wrong. Which of these two options are you going for?

However, your response does answer my other question; “You already asked about precision and bit depth dB values numerous times in this thread and I answered (here, in this post) but you simply ignored the answer and made up completely false values, why would you do that unless you’re trolling?” - You’re deflection is a tacit admission that you’re trolling!

You pushed for answers from others and now it’s your turn, answer the questions!

G
 
May 28, 2025 at 10:51 PM Post #3,723 of 3,803
let’s just lay the math out, from a math standpoint;

If the the top ~6dB of voltage, to an AD converter, is measured in:

0.0002dB increments(16 bit)
0.0000007dB increments(24 bit)

Do you agree that:
0.0000007dB <0.0002dB



feel free to check the math and please feel free to correct anything incorrect:
((2^16)/2) / (6/32,536)= 0.0002
((2^24)/2) / (6/8,388,608)=0.0000007

Just my 2 cents.

I have a fun question for you: (don't cheat with AI or searching), if you listen between 7KHz square wave and 7 KHz sine wave, will they sound the same or not?
 
May 29, 2025 at 3:35 AM Post #3,724 of 3,803
feel free to check the math and please feel free to correct anything incorrect:
((2^16)/2) / (6/32,536)= 0.0002
((2^24)/2) / (6/8,388,608)=0.0000007

Just my 2 cents.
That math looks nonsense to me. I don't even know what you are trying to do here. The increment is constant in linear scale, not in dBs.
 
May 29, 2025 at 1:09 PM Post #3,726 of 3,803
feel free to check the math and please feel free to correct anything incorrect:
((2^16)/2) / (6/32,536)= 0.0002
((2^24)/2) / (6/8,388,608)=0.0000007
I'm not sure what you're trying to calculate here (though I have my guesses), but it's wrong. When you enter those operations into a calculator, you get:
Code:
((2^16)/2) / (6/32536) =   177689941.33...
((2^24)/2) / (6/8388608) = 11728124029610.78...
 
Last edited:
May 29, 2025 at 2:10 PM Post #3,727 of 3,803
I'm not sure what you're trying to calculate here (though I have my guesses), but it's wrong. When you enter those operations into a calculator, you get:
Code:
((2^16)/2) / (6/32536) =   177689941.33...
((2^24)/2) / (6/8388608) = 11728124029610.78...
Thanks for catching, the answers still remain 0.0002 and 0.0000007, as the math should now line up as to how those were numbers were obtained.
 
Last edited:
May 29, 2025 at 4:02 PM Post #3,728 of 3,803
I have a fun question for you: (don't cheat with AI or searching), if you listen between 7KHz square wave and 7 KHz sine wave, will they sound the same or not?
Oh wow. Cutting edge stuff here, but you still don't get it. You fundamentally do NOT understand how digital audio is converted to analog (or vice versa for that matter) and you have absolutely no interest in learning either because you can't or you don't want to break the illusion of more bits = sounds better.
 
May 29, 2025 at 4:04 PM Post #3,729 of 3,803
JavaScript:
const doYouAgree = (a, b) => {
  if (a < b) {
    console.log('What does this have to do with dynamic range in audio?')
  } else {
    console.log('What does this have to do with resolution in audio?')
  }
};

doYouAgree(0.0002, 0.0000007);

document.getElementById('answer')?.innerHTML = "Glad we agree";

const whatEq1849Thinks = `${imageDirectory}${imageFileName}`;

eq1849 callback:

mimosa-pudica.gif
 
Last edited:
May 29, 2025 at 5:09 PM Post #3,730 of 3,803
Thanks for catching, the answers still remain 0.0002 and 0.0000007, as the math should now line up as to how those were numbers were obtained.
feel free to check the math and please feel free to correct anything incorrect:
((2^16)/2)=32,536.
(6dB/32,536)= 0.0002dB per quantization integer.

((2^24)/2)=8,388,608
(6dB/8,388,608)=0.0000007dB per quantization integer.
It may be how you got the numbers (still, (2^16)/2 = 32768), but it doesn't mean that what you did has any sense. You cannot just divide decibels like that.

At 0 dBFS a change of a sample's value by one quantization level is about:
Code:
16-bit: 20 * log( 32767 / 32768 ) = -0.000'27 dB (by accident about the same what you got)
24-bit: 20 * log( 8388607 / 8388608 ) = -0.000'001 dB
At -6.02 dBFS a change of a sample's value by one quantization level is about:
Code:
16-bit: 20 * log( 16383 / 16384 ) = -0.000'53 dB
24-bit: 20 * log( 4194303 / 4194304 ) = -0.000'002 dB

And still, all this doesn't mean what (it seems) you think it means.

The error created by quantization of each sample becomes an additional signal that is added to the original signal. With dither this additional signal is white noise.
So the quantization changes our original signal into: the original signal + white noise (or shaped noise). There's nothing more.
 
May 29, 2025 at 6:17 PM Post #3,731 of 3,803

This discussion hasn’t generated much in the way of content; but if nothing else, we got this charming little penguin out of it.
 
May 29, 2025 at 6:33 PM Post #3,732 of 3,803
Oh wow. Cutting edge stuff here, but you still don't get it. You fundamentally do NOT understand how digital audio is converted to analog (or vice versa for that matter) and you have absolutely no interest in learning either because you can't or you don't want to break the illusion of more bits = sounds better.

You're far off from what I'm getting at. I didn't say bits whatsoever. So 16 bit 7KHz square wave against 16 bit 7KHz sine wave. Would these two sound the same or not. You should know the answer if you know how digital audio is converted to analog
 
May 29, 2025 at 6:47 PM Post #3,734 of 3,803
And if you tell him how the square wave is made because the answer obviously depends on that.

Digitally of course. DAC, NOT ADC such as analog synths!
 

Users who are viewing this thread

Back
Top