24bit/96/192KHz vs 16bit/44.1/48KHz for music playback - A scientific look.

I was Googling for how Google Music handles lossless audio and amidst the search results my eyes focused on this article.

Having read it I'm now more aware of the misconceptions people make (myself included) and will no doubt have this in mind when getting new music.

The TL;DR version is explained in this extract:

Unfortunately, there is no point to distributing music in 24-bit/192kHz format. Its playback fidelity is slightly inferior to 16/44.1 or 16/48, and it takes up 6 times the space.

There are a few real problems with the audio quality and 'experience' of digitally distributed music today. 24/192 solves none of them. While everyone fixates on 24/192 as a magic bullet, we're not going to see any actual improvement.

...

The ear hears via hair cells that sit on the resonant basilar membrane in the cochlea. Each hair cell is effectively tuned to a narrow frequency band determined by its position on the membrane. Sensitivity peaks in the middle of the band and falls off to either side in a lopsided cone shape overlapping the bands of other nearby hair cells. A sound is inaudible if there are no hair cells tuned to hear it.

It's worth mentioning briefly that the ear's S/N ratio is smaller than its absolute dynamic range. Within a given critical band, typical S/N is estimated to only be about 30dB. Relative S/N does not reach the full dynamic range even when considering widely spaced bands. This assures that linear 16 bit PCM offers higher resolution than is actually required.

It is also worth mentioning that increasing the bit depth of the audio representation from 16 to 24 bits does not increase the perceptible resolution or 'fineness' of the audio. It only increases the dynamic range, the range between the softest possible and the loudest possible sound, by lowering the noise floor. However, a 16-bit noise floor is already below what we can hear.

...

16 bits is enough to store all we can hear, and will be enough forever.

Click to expand...

So what do you guys think of this? I have several albums in 24/96 and my DAC/Amp supports that over Asynchronous USB. Maybe I should drop the output from Windows to 16bit and 44.1 or 48KHz with this new knowledge in mind considering the bulk of my music is 16/44.1 anyway.

Think a re-read in the evening is in order as some of it went over my head as well this morning

another snippet for the TL;DR crew.

Empirical evidence from listening tests backs up the assertion that 44.1kHz/16 bit provides highest-possible fidelity playback. There are numerous controlled tests confirming this, but I'll plug a recent paper...<snip>

BAS conducted the test using high-end professional equipment in noise-isolated studio listening environments with both amateur and trained professional listeners.

In 554 trials, listeners chose correctly 49.8% of the time. In other words, they were guessing. Not one listener throughout the entire test was able to identify which was 16/44.1 and which was high rate [15], and the 16-bit signal wasn't even dithered!

Click to expand...

Nice read
I always have kept mine @16bit 44.1 because I could never tell the difference between them

I think 24/96 does sound much better

But why does it sound better to me, not because its 24bits96khz, because its mastered differently. the 24bit reissues from original analog tapes have a lot less dynamic compression or none at all so i think thats why people always think 24bit music is the "magic bullet" they dont realise that most cds these days have been clipped to **** even if they are 16/41 flac files.

So yeah its not the bits or khz making it sound better its the no clipping. Get the 24bit knock it down to 16 flac or 320kps i say



(mrk, try and listen to 'one world' in the best quality you can find. i assume you have good equipment your ears will be in for a treat

That's the point of the article Samsara That if you converted the 24/96 Mastered re-release to CD format then the sonics are identical.

And aye the final parts of the article go into detail about this that marketing depts have a field day promoting 24/96/192 when the only real difference is that a higher quality master was used. Had the same master been used for a CD re-release then it would sound identical to the 24/96 version!

With this in mind I don't think it matters a great deal whether the Windows DAC output is set to default 16/41 or 24/96 considering any corrections are being done by the amp internally?

Which artist does "one world" btw? There are a few on Google

Not a bad article but he basically jumps from:

"Here's how hearing works" to "192khz is useless" with literally no external references, padding it out with gumpf about light and seeing other ranges of light.

In our hypothetical Wide Spectrum Video craze, consider a fervent group of Spectrophiles who believe these limits aren't generous enough. They propose that video represent not only the visible spectrum, but also infrared and ultraviolet.

Click to expand...

We can all see ultraviolet light, it just gets filtered out because it's damaging.
http://en.wikipedia.org/wiki/Aphakia
So basically, he hasn't done his research at all on other aspects and expects us to take his word for it? Nope. No thanks. I appreciate this guy knows a lot about sound though.

In my experience the mastering of a track matters a great deal more than anything else. Perfectly good songs are ruined by overcompression (as already mentioned) and generally poor balance - Simon and Garfunkel's classic Bridge Over Troubled Water has horrendous mastering. If someone thinks they can hear a difference betwen 44.1/16 and 192/24 then that's up to them. I'd argue that the way harmonics work, even if you're capturing frequencies beyond human hearing, they will interact and create different sounds - the same as playing an instrument in a small room and a concert hall will create completely different experiences. The difference between /16 and /24 is more of a technical one, with extremely clean power input there may be some benefit.

It was discussed on another forum a few times:
http://theartofsound.net/forum/archive/index.php/t-16720.html
http://theartofsound.net/forum/archive/index.php/t-16828.html
http://theartofsound.net/forum/archive/index.php/t-20474.html

Tl;dr Mastering matters more than format, generally.

mrk said:

Which artist does "one world" btw? There are a few on Google

Click to expand...

dire straits

Here's a simple explanation..

Every sound can be broken down into sine waves. A perfect square wave forms can be made up of sine waves with infinitely progressing smaller sine waves on the (even I think?) harmonics.



Adverts often show sample rate like this, a block ridden waveform which barely resembles the original waveform :-



This is incorrect.

PCM (pulse code modulation) as used in audio, is similar to FM (frequency modulation), where the data shows points and creates a pure sine wave between the two points based on the frequency and amplitude.

Like this :-



Now the higher the sample rate (Frequency) the more harmonics a waveform can have (remember how the square wave is made at the start), but your ears only hear to around 20khz. Any frequency higher than 20khz and your speaker is producing sine waves which are too small for your ears to detect.

Now higher sample rates have a use. When working with digital audio in a recording environment you sometimes have audio effects which work by altering pitch (chorus, phasers etc). When you use these effects on high frequency sounds these effects pitch down the sample and in turn bring aliasing distortion (filtered out by the DAC filter) into the audible range. By using a higher sample rate you are moving the aliasing and filter further above the audible range so this pitching down never becomes an issue.

Now on to bit rate.

16 bit has a dynamic range of 120db. That means that it can go from silent to deafening (literally) in 65536 steps. Which is once again the full level a human can hear.
24 bit has a dynamic range of 144db which is beyond the dynamic range of human hearing.
When recording audio at 16 bit, to get the full dynamic range people generally set the loudest part of the music to 0db. But if any transient sound accidentally goes over that level it goes into clipping, which is digital distortion.
24bit dynamic range allows you to still record audio at 0db but also gives you a headroom allowing any transients to go higher without leading to clipping/distortion.

So.. in short.

You only need 16/44.1 for audio playback. Any more than that and you're going beyond the level of what your ears can handle. Take zero notice of the marketing fluff as it's all loblox

Interesting read thanks. The "loudness war" is a bit pants but interestingly i have been using Creative soundcards for the past 8 maybe 10 years and have had a decent enough experience whilst listening to music and gaming, my favourite sounds always come from my midi music sessions and using a usb midi keyboard controller and messing around with different free demo's of daws and the vst's instruments they all come with. probably the best music experience is live bands with decent pa systems, but going back to the creative x-fi soundcards i have used in the past (x-fi extreme music PCI and x-fi titanium PCI-E) whilst they sounded ok and the experience was good enough but interestingly the new soundblaster z card i am testing at the moment is so much louder but not only loud but it sounds much better than the 2 x-fi cards i had especially in midi music and games sound better too, so what is causing that?, software?, the sound card auto sets iteslf up to 24-Bit, 48000Hz (studio quality) in windows 7 x64, although i am sure i can tell a very slight differnce in midi audio live playback with the controller and audio set at 96000 sample rate (Hz) creative asio device at 2ms buffer length, i wonder why games sound louder?, placebo effect perhaps? because it is a new soundcard or software manipulation?. But whatever the creative engineers did a good job, perhaps slightly better onboard components on the soundcard? is it the soundcore3d chip?, i was quite impressed if i'm honest, only downside is that the sbz oem basic soundcard might melt sandwhiched in between two hot gtx780;s, hahaha, classic stuff, haha.

I see what you guys are saying above yeah and the last few days I've been playing around. There's certainly no quality difference listening to 24/96 Flacs at 16/44.1 but in Windows I have noticed if I set the amp (USB connection) to 16/44.1 then at very high volume when there's nothing playing you can hear the ambient hiss faintly like you get on an analogue amp. There is no distortion or interference from anything else but the ambient hiss is there. That kind of volume level is ear deafening though so would never be that high when playing stuff.

Set Windows to any 24 bit mode and it's dead silence even at max volume. So it appears that Windows 7 outputs the cleanest audio at 24bit when using a DAC but there's no audible difference in musical performance.

I'll leave it at 24/96 purely for the clean background

Samsara said:

dire straits

Click to expand...

Cheers!

Recording at 44.1kHz limits the highest frequency sound to not much over 20kHz.

In testing my own hearing with test tones, it drops off at around 17 kHz (like most people my age) but picks up again in the low to mid 20s kHz before dropping off to silence again).

Those cat scarer things in people's gardens literally do my head in (the one we have claims to do a 27-21 kHz sweep). I wouldn't say it's audible as such but it's detectable (both very quiet and unpleasant somehow).

Now supposedly, you need over double the sampling rate for the frequency you're trying to reproduce. Therefore 44.1 kHz has a max reproduction limit in the low 20s kHz. Therefore it's not capable of reproducing my hearing range.

I wouldn't claim to be able hear the difference between CD quality and higher recordings though. That higher range is barely audible for me and would probably just get drowned out by my sub-17 kHz normal hearing range.

Edit: Googling, it seems that the 20Hz to 20kHz range gets quoted all over the place, with the odd acknowledgment that babies can sometimes hear up to 22 kHz. And then there's stuff like this: http://recordinghacks.com/articles/the-world-beyond-20khz/

It seems it's very similar to the whole speaker cable debate.

A while ago, I just decided to buy, rip and keep audio (music really) in the highest grade audio possible.

I once tested recording mp3s at different bit rates and decided that after 192 kb/s I couldn't hear any difference. Years later I upgraded speakers / soundcard / CD player / amp / headphones and noticed a harshness to a bunch of my old rips that I didn't hear at the time. I don't want that to happen again, and if the price premium isn't massive, why not (theory be damned) buy music in the highest quality available?

Because a bought lossless file (say FLAC) @ 16/44.1 takes up less space than @ 24/96 or 24/192 and has no audible difference, this is the whole point of this thread

This is important for those who like to keep their lossless music on portable devices.

As far as PCM is concerned then it's safe to assume that 16-bit depth is enough for playback. Human ears degrade too much to make use of the fidelity. I know there's a whole argument where some say that 24-bit sounds better due to distortion or something (much like the whole vinyl vs. CD argument) but that's some crazy audiophile thing where it sounds more 'real or 'crispy' in their heads.

I'd been wondering about this stuff recently while exploring the VIA HD audio options on my new notebook, offering various different format options. Think I changed it to 48KHz and 24-bit just to see what difference it made if any, can't say there was any discernable one so I've changed it back to 16-bit/44.1KHz after reading this thread.

TheVoice said:

I'd been wondering about this stuff recently while exploring the VIA HD audio options on my new notebook, offering various different format options. Think I changed it to 48KHz and 24-bit just to see what difference it made if any, can't say there was any discernable one so I've changed it back to 16-bit/44.1KHz after reading this thread.

Click to expand...

The noise and crosstalk is pretty bad on onboard sound anyway, it's the worst example of a 24-bit device. It's possible 24-bit has manufacturing benefits however - it may be cheaper to make a "bad" 24-bit chip than an "excellent" 16-bit one.

spoffle said:

It seems it's very similar to the whole speaker cable debate.

Click to expand...

probably because it's all subjective to the end user, i once tested some basic cheap bell wire speaker cables against some expensive qed brand cables connected to the same seperate hi-fi system and i couldn't really tell a difference except that one set of speaker cable looked better than the other cables, so was it aesthetics over sound quality?, metaphysics and the laws of perception is a curious thing . By the looks and sounds of things you might as well stick with 16/44 for playback/listening, but another question is "cheaper DACS v expensive DACS whilst listening in 16/44.1?, ha .

You wont enjoy the next step then haha, the industry is moving over to blu ray audio cd's soon, all of us in the industry thinks its a waste of time but hey ho. the bit wars are just aimed at misinformed people like you say to give a perceived advantage by marketing departments. At our studio we record at 44khz/24bit as well, cd's are produced at that with the exception of using 24bit as it makes a large difference with some of our plugins. Although i can say after sitting in on some mastering sessions at metropolis studios that what makes the largest difference in quality is using the best of every component that the signal passes through, this makes more of a difference long before bit depth is considered lol. prism converters and speaker systems costing more than my house haha

Re: DAC/Amp - It doesn't matter if it's expensive or cheap though, 16/44.1 envelopes the full range for playback humans can hear

The difference that is subjective is exactly what a person has trained their ears to pick up on. Person A may be more technical than person B and over the years has gradually upgraded speakers/headphones/amps and at each step has noticed new details in music they are familiar with (usual occurrence with most people from communities like this one and others where we share a common interest) whereas person B will not know any different because music is just music, they just want to hear a song - And so on

Really thin speaker cable may warm up and fade with repeated use on large speakers as well due to resistance - I had this on my old Tannoy speakers and NAD amp where I got thin cabling that looked nice (clear sheath) and was cheap but after long sessions the cable was warm and soft. Over time the sheath started to go yellow from the repeated warm up/cool down.

Since then I've always had thick gauge cables on my speakers, not necessarily expensive cable but thick cable that won't get bogged down with heavy use.