24 bit music question


And finally, the "ear can't listen more than" reminds me those endless debates around vinyl vs CD, or tube vs transistor amps.
Problem is,it's been tested in blinded conditions and effectively, people can't tell the difference. Again and again and again. So yeah, the ear can't listen to stuff that it cannot detect.  
Certainly not vinyl vs CD, or tube vs transistor amps. Especially at high volume it's damn easy to tell the difference.
People can grow attached to particular types of distortion/noise, but if it is sound fidelity you want, there is no reason whatsoever to use outdated analog media like vinyl or tape cassettes, regardless of the volume level.
 
No, there isn't. That is completely and utterly untrue. However, both the tube amplifier and the vinyl player adds particular distorsion that is rather pleasing to our auditory apparatus. For a tube amplifier that is even-order harmonic distorsion, for the vinyl player it is a kind of chorusing effect that depends, basically, on the minute variations in distance of the grooves to the center of the record. Both of these are noticeble (so it is indeed quite easy to pick out which is which), and a lot of people find these artifacts to be nice sounding.

However, it is not in any meaningful way a more truthful representation of the original, and it most certainly does not contain "more information". 
 
There's simply more information.

-Sigh-
More noise, mostly. Even analog is not really continuous: everything is discrete at the Planck scale, or when dealing with objects made out of normal solid matter, already at the molecular scale. Even if we would be technologically capable of producing vinyl with grooves whose depth is accurate down to the molecular scale, and players who can play it back perfectly, it would still be only discrete, digital information that can be encoded. It would be an interesting exercise to calculate what bit depth and sampling rate corresponds to this "molecular-perfect vinyl".

But of course we cannot make perfect vinyl grooves (or at least it would be way too expensive), and we cannot make perfect vinyl players (that would also be way too expensive and probably impossible to do in real-time with current technology). So what you get is introduction of noise at various stages.

So yes, strictly speaking, a vinyl record contains more information than a CD, if you count the noise as information. But it contains less correct information. The noises, distortions and artifacts introduced by vinyl outweigh by orders of magnitude the theoretical advantage of being able to represent the sound wave at a higher resolution because there is no sampling and quantization, only "analog".

I just did a quick Google search and found out that as usual, I'm not the first one to make such an observation. I'm now reading this: http://www.st-andrews.ac.uk/~www_pa/Scots_Guide/iandm/part12/page2.html
 
It has been shown, repeatedly, that people cannot tell the difference between CD quality (or perhaps very slightly above) recordings and higher quality ones.

This means that there is no point in recording more information than 16/44.1 or so for playback.

It is true that an analogue recording, if perfect, contains more information than a digitisation of the sound - but the human ear can't detect the difference anyway.

Analogue recording media are generally not perfect and introduce distortions of their own, which can be picked up by human ears.

By comparison, CDs have heaps of error-correction, and are digital, meaning that exactly all the sound information on the disk reaches the playback electronics intact.

(at this point, a good DAC is needed)

Then at the amplification stage, silicon-based kit is just straight-off more accurate  - and will do exactly what its supposed to do, with a minimum of distortion to the signal.

The bottom line is that people may like the distortions and artefacts of old audio equipment, but modern digital systems do a better reproduction of the original sound.
 
Last edited by a moderator:
This whole topic reminds me a bit of the old "the more megapixels the better" discussion in digital photography. Like "my new smartphone has a 30megapixel camera so it _has_ to be better than your 18 megapixel dslr", completely neglecting that there is a _lot_ more to it... good quality lenses for starters.

In the audio context: stating that an audio recording is 24/96 simply doesn't say anything about the mixing or the mastering process. I doubt "Death Magnetic" will sound any better in 24/96 than in 16/44 if the same kind of mastering chain would be applied.

And as a hobby musician I can say: I've gone back to recording in 16/44,1 some years ago because I couldn't make out a big difference in audio quality. It only inflates the project folder. Yeah, I _do_ prefer buying vinyl records, but that's mostly because of a certain sentimental value.
 
This whole topic reminds me a bit of the old "the more megapixels the better" discussion in digital photography. Like "my new smartphone has a 30megapixel camera so it _has_ to be better than your 18 megapixel dslr", completely neglecting that there is a _lot_ more to it... good quality lenses for starters.
You don't even need to look at other components to argue against this: more pixels on the same tiny sensor area means more distortions and interferences, resulting in a worse image - that's why the professional digital reflex cameras have huge sensors.
 
Last edited by a moderator:
This whole topic reminds me a bit of the old "the more megapixels the better" discussion in digital photography. Like "my new smartphone has a 30megapixel camera so it _has_ to be better than your 18 megapixel dslr", completely neglecting that there is a _lot_ more to it... good quality lenses for starters.
You don't even need to look at other components to argue against this: more pixels on the same tiny sensor area means more distortions and interferences, resulting in a worse image - that's why the professional digital reflex cameras have huge sensors.
Yeah... you're right. Lenses just were the first thing that came to my mind when thinking about digital cameras.

But I think my point is still valid. Companies just try to find the highest number they can find in their product and center the whole marketing strategy around it. And slowly but surely these "facts" trickle into people's brains... so the next time they buy a camera they grab the one with the highest number of megapixels they can find, not for the one with a bigger sensor or a better lens.
 
Last edited by a moderator:
so the next time they buy a camera they grab the one with the highest number of megapixels they can find,
Yeah, but that push for the highest numbers has limits. Once you realize that you 18M camera on your smartphone takes crappy pictures most of the time you understand you've been had. Plus, the density of pixels per square inch tends to increase more slowly nowadays, so they really have to focus on what actually matters in the end: getting good pictures.
 
so the next time they buy a camera they grab the one with the highest number of megapixels they can find,
Yeah, but that push for the highest numbers has limits. Once you realize that you 18M camera on your smartphone takes crappy pictures most of the time you understand you've been had. Plus, the density of pixels per square inch tends to increase more slowly nowadays, so they really have to focus on what actually matters in the end: getting good pictures.
In terms of technological development I fear that, once the whole megapixel thing has worn off, companies like Samsung, Sony or Apple will find a new hot thing to sell their stuff. Let's wait for 400€ smartphones that play 24/192 soundfiles, have a shiny 4k display and let you shoot 4k video footage.
 
96dB is not the extent of human dynamic range. It is however pushing the limits of what you can accurately record on a microhone.

It is a bit more difficult to design with more bandwidth dacs, but if 5 million dollars say its worth it in the form of pono, its certainly a huge extra market for the pyra.

Seeing as dedicated mediaplayers arent all the rage nowadays.
 
But more megapixels means bigger JPEGs which means you need a new computer to store them all on, and the cycle continues. Likewise the progress from 120kbps MP3 to 360kbps MP3 to 44.1k/16 FLAC to 96k/24 and beyond.
 
Ultrasound could be useful if you're a biologist doing experiments on dogs or other animals, just like sensors that can capture EM waves outside the visible spectrum can be useful to astronomers. Any self-proclaimed audiophile who insists on encoding and playing back ultrasound for humans is being just as silly as a self-proclaimed photographer who insists on a camera that captures not just visible light, but also UV light and X rays -- not to later manipulate the images to give false colors to the invisible frequencies, but just because, you know, they have very sensitive eyes. And of course they need a screen that can display the UV light and the X rays too.

Actually that could be fun. You could then make a picture of the sun and get a sunburn and skin cancer by looking at it B) .
 
Contact your local mastering engineer for more information about why 96/24 is better than 44/16.

I'm tired to hear the same "scientific arguments" over and over since the CD debuts. I was already playing music, using synths back then. A while after I worked in pro recording studios when i played guitar for a living, talked to experienced audio guys etc. so i give up on this thread, it's all yours.

For french readers, here's one classic lecture :

http://cec.sonus.ca/pdf/La_pratique_du_Mastering.pdf
 
Contact your local mastering engineer for more information about why 96/24 is better than 44/16.
During processing / manipulation / recording / production, sure. For storing samples that will be transposed etc, of course. For distribution to humans that only want to listen to the music? No. It's not better for that.

It's like a supermarket that advertises its prices with an accuracy of 8 decimal digits: yes, I suppose you could say that it is "better", that it is more accurate, but no, it does not really make a difference as long as the price you pay as a consumer is rounded to the nearest cent anyway. For professionals who may need to buy items in very large quantities, it makes sense to have such weirdly accurate prices (because those prices will get multiplied by huge quantities), but for an average consumer who typically only buys one of an item, or maybe occasionally two or three, it more than suffices to round the prices to 2 or 3 digits.
 
Actually that could be fun. You could then make a picture of the sun and get a sunburn and skin cancer by looking at it B) .
Digital sensors, and I assume film too does capture infra red, due to the red filters/chemicals not being precisely tuned to the same limit as our red cones. You just need to filter out all the visible light so the sensor gets activated/chemical gets converted only due to the IR. I assume they pick up UV too by a similar process. Looking at an IR picture doesn't actually feel hot though, just as the silver from an old silver oxide positive doesn't reflect the original light that formed it.

OT in the extreme - I suspect a camera sensor would work better as an IR sensor if it didn't have the RGB filter in front of it. And it wasn't your point, though I suspect this thread has gone to seed anyway, so I may as well.
 
Your orders of magnitude comparison is off by orders of magnitude.
Yes, you're right: in the supermarket analogy I was introducing 5-6 additional decimal digits (8 instead of 2-3), while in the case of audio we are talking about doubling the sampling frequency and 8 more bits per sample, so what used to be one 16 bit sample become two 24 bit samples, so the accuracy improves by 32 bits or about 10-11 decimal digits. So it would be a more accurate analogy if the supermarket would express its prices with 13 decimal digits, e.g. on the Google playstore you would probably pay $0.9999999999999 instead of just $0.99.

If you're talking about 192kHz sampling frequency with 24 bit samples, you go from one 16 bit sample to four 24 bit samples, so 80 extra bits of accuracy or about 24 extra decimal digits. So on the Google playstore you would get prices like $0.99999999999999999999999999.
 
Back
Top