or Connect
Head-Fi.org › Forums › Equipment Forums › Sound Science › Why 24 bit audio and anything over 48k is not only worthless, but bad for music.
New Posts  All Forums:Forum Nav:

Why 24 bit audio and anything over 48k is not only worthless, but bad for music. - Page 2  

post #16 of 2273
Quote:
Originally Posted by SilverEars View Post

The controlled condition, I really want is the most resolving audio system known to man.  Then, if it still cannot distinguished, and only then, I will side with the rebel alliance here.  biggrin.gif

tongue.gif

se
post #17 of 2273

I think people are missing the point that a sample rate only approximately twice the frequency rate (and getting close to 20K cycles is where that occurs) is low resolution.  Think of a fax machine sampling a line 1/10th an inch thick with a scanner whose resolution is 1/20th of an inch.  There would only be two samples (or dots) to represent the line.  Is the line on the fax there, yes, but it's highly pixelated.  And there have been experiments that show ultrasonic information in a recording does have an effect on fidelity.  If you're not hearing the benefits of higher resolution digital, then either the equipment is not resolving enough to allow it, the software (or downsampling) is not up to snuff, or the listener may not be cued into the benefit's of higher res files.   When I am color correcting video and I shift the hue slightly and toggle back and forth for a client, sometimes they can't see the color shift, while I can see it easily. Does that imply it's not there because some people can't see  (or hear) it?  No it doesn. But perhaps if I did that same color shift on my 15' DLP projection screen the person might be able to see it in that instance because the scale of the playback device makes things more easily discernible.  

 

When I'm mixing audio I can usually easily hear a 1db increase, and 1 db is supposed to be the threshold of human hearing.  There are several times I've only raised a clip .5 db.   

 

I also hate to bring this up but my theory is that many in the pro audio world have compromised their hearing by  exposure, so perhaps that is where some of the opinions about 44.1/16 audio quality come from, and possibly is the source of all the bad sounding recordings out there.   ALso ABX testing configs go through added passive or active components which could easily mask the fine details of higher res files.

 

I should add that I am listening though monitors with edge of the art vapor deposited beryllium concentric drivers, (TAD) that act as point sources  and so perhaps that is why small sonic differences are so easily discernible.  


Edited by emailists - 5/1/14 at 8:01pm
post #18 of 2273
Quote:
Originally Posted by emailists View Post
 

I think people are missing the point that a sample rate only approximately twice the frequency rate (and getting close to 20K cycles is where that occurs) is low resolution.  Think of a fax machine sampling a line 1/10th an inch thick with a scanner whose resolution is 1/20th of an inch.  There would only be two samples (or dots) to represent the line.  Is the line on the fax there, yes, but it's highly pixelated.

 

According to Nyquist, it only takes two points to perfectly reproduce a sound wave. The stair step and pixel analogy is not the right way to think about it. Digital audio is different than digital images. You are reconstructing the sound wave from the ground up. You aren't creating an image of it in pixels.

 

The threshold of human perception for changes in volume in direct a/b comparison is between .5 and 1 dB, depending on the sound. For tones, the lower, for sparse and varied music, the higher.

 

The resolution of 20-20 in a redbook audio file is EXACTLY the same as the resolution of the same frequencies in a 24/96 file. The only difference is the depth of the noise floor and the super audible frequencies (which have been proven to be not necessary for sound quality of music). Recorded music that has been mixed and mastered fits easily within 16/44.1 with room to spare all around.

post #19 of 2273
Quote:
Originally Posted by emailists View Post
 

I am a bit aghast at some of the opinions displayed in this thread.  On a high resolution system, the benefits of 192K over 96K is clearly audible.  I use CHesky Jen Chapin recording of the same album I bought in 96 then 192 when it was released.  Just because someone can't hear the difference in their system doesn't mean it doesn't exsist.   I guess at one time just because people couldn't see the curvature of the earth led them to believe the earth was flat.

 

Prove it please. Citations needed.

 

Why don't you take the Chesky Jen Chapin album you have, take the 192kHz and use foobar to downsample it to 44.1k. Then all you have to do is post your ABX log from foobar's ABX tool for 20 trials between the 44.1kHz and 192kHz versions of the same song. I think everybody here would be interested in seeing the results. The best part of all is that since you already have the album, you can do the test for free!

 

cheers


Edited by ab initio - 5/1/14 at 10:25pm
post #20 of 2273
I'm going to go subjective for a second here, but I've tried listening to two recordings of the same master... one Hi-Res and one not (not blindly just regular listening) on many different systems with different DACs and could not tell the difference.

Even if there is, it is likely out of human threshold of hearing.
post #21 of 2273

Doesn't 24 bit mean there are 2^24 number of vertical levels or precision of the line-out signal?  And you use the analog volume knob which amplifies it.  This depends on how it was sampled in the first place.  If you sample the original analog 24-bit, wouldn't it make the output into 24-bit still precise?  If it was sampled at 16, then there is no point.  Also, Sampling frequency will make the signal reproduction much more accurate.  

 

I guess the analog distortion would have more affect on the signal than such high signal conversion precision.


Edited by SilverEars - 5/1/14 at 11:28pm
post #22 of 2273
Quote:
Originally Posted by SilverEars View Post
 

Doesn't 24 bit mean there are 2^24 number of vertical levels or precision of the line-out signal?  And you use the analog volume knob which amplifies it.  This depends on how it was sampled in the first place.  If you sample the original analog 24-bit, wouldn't it make the output into 24-bit still precise?  If it was sampled at 16, then there is no point.

 

Is this why we see 24-bit tracks floating around that says vinyl on it, since it was sampled 24-bit from vinyl?


Not exactly. Dither allows signals to be recreated with much finer resolution than one might expect niavely from simply the number of discrete levels. For example, see the wiki page on dither: http://en.wikipedia.org/wiki/Dither

 

 

As for the vinyl bit... yeah, folks like to record their vinyl records at absurd bitdepths and sampling rates because they like to support the harddisk industry.

 

Cheers

post #23 of 2273
Quote:
Originally Posted by ab initio View Post
 


Not exactly. Dither allows signals to be recreated with much finer resolution than one might expect niavely from simply the number of discrete levels. For example, see the wiki page on dither: http://en.wikipedia.org/wiki/Dither

 

 

As for the vinyl bit... yeah, folks like to record their vinyl records at absurd bitdepths and sampling rates because they like to support the harddisk industry.

 

Cheers

Isn't that more of interpolation, or upconverting than recreation of the original signal?  If you sample the original signal at high sampling rate, it will be more precise than sampling at lower sampling rate and up converting.  up converting is padding it with samples that are guessed(filling in the gaps), not the original samples taken.


Edited by SilverEars - 5/1/14 at 11:36pm
post #24 of 2273
Quote:
Originally Posted by SilverEars View Post
 

Doesn't 24 bit mean there are 2^24 number of vertical levels or precision of the line-out signal?  And you use the analog volume knob which amplifies it. 

 

How much dynamic range can you hear?

post #25 of 2273
Quote:
Originally Posted by SilverEars View Post
 

Isn't that more of interpolation, or upconverting than recreation of the original signal?  If you sample the original signal at high sampling rate, it will be more precise than sampling at lower sampling rate and up converting.  up converting is padding it with samples that are guessed(filling in the gaps), not the original samples taken.


No, Dither is most certainly not interpolating. Dither exchanges quantization distortion (which shows up as energy concentrated at a few frequencies) for noise (which spreads the energy over many frequencies, resulting in noise with much lower peak amplitude).

 

Cheers


Edited by ab initio - 5/1/14 at 11:50pm
post #26 of 2273

Well anyway, on second thought, I have doubts that you need this much precision because the amp is gonna distort the signal, and that's has more affect on the signal.

post #27 of 2273
Quote:
Originally Posted by SilverEars View Post

Well anyway, on second thought, I have doubts that you need this much precision because the amp is gonna distort the signal, and that's has more affect on the signal.

In general, i think the physical transducer is going to distort the waveform more than either digitization or amplification. But yes, i think fussing over high def audio formats is one of hifi's lesser concerns.

Cheers
post #28 of 2273
Quote:
Originally Posted by ab initio View Post


In general, i think the physical transducer is going to distort the waveform more than either digitization or amplification. But yes, i think fussing over high def audio formats is one of hifi's lesser concerns.

Cheers

Well, when I A/B'd my DACs, I found one that my ears seems to favor, and sounded transparent.  I wonder what it could be, maybe the DAC's analog stage is done better than the others?

post #29 of 2273

Sounds like you got stuck with some defective DACs. I always check everything before it goes into my rig, and everything has always been transparent.

post #30 of 2273
Quote:
Originally Posted by bigshot View Post
 

Sounds like you got stuck with some defective DACs. I always check everything before it goes into my rig, and everything has always been transparent.

Which DAC do you use?

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
This thread is locked  
Head-Fi.org › Forums › Equipment Forums › Sound Science › Why 24 bit audio and anything over 48k is not only worthless, but bad for music.