16 bit to 24 bit.....What’s the other 8 bits?

Fact: 24-bits allows more than enough dynamic range for any musical application. 16-bits doesn't allow for enough dynamic range, depending on the musical application. It's that simple.

The increased dynamic range is vital in a DAW when processing, and can subjectively improve the sound quality of properly-dithered 16-bit audio.

I don't know why you guys are arguing over all of this.

 

English may not be your first language, but you seem to have a high command of it. Read your statement again and think about it. It is an empty one, and has no meaning. You are merely attempting to give it an objective meaning where there can't be.

 

Again, you realize that you've made a very general, blanket statement that cannot be accurate.

 

Old men yell at clouds.....

 

Audio "arguments", seem mostly based on....
Take something that is a given, and try to convince everyone that is it not true, but with no real proof, other than "I think so, it must be"...

 

And it's wrong. There is no real world recording with that much dynamic range, and you wouldn't want one because the volume required to appreciate the loudest and the faintest sound at the same time would be earsplitting. Before CD was standardised, 14 bits were considered for a while. Even that is better than any analogue medium of the day and would have been absolutely fine.

Because it gives mixing engineers a lot of error margin for boosting tracks that were erroneously recorded far too quiet, without also pushing their noise floor up to intolerable levels, and without the need to rerecord. That doesn't mean it's needed in a consumer product.

 

And it isn't. People in their 60s can hear maybe up to 10 kHz instead of 20 kHz, and that seems like a huge loss. Half of the frequency range gone, 10.000 Hz! But in audio you have to think in ratios. It's only an octave, and only in a realm where no musical notes exist, just the highest of highest overtones. Even 10 kHz is a painfully high beep.

 

Yes I agree. As I said, for most it's no big problem.

 

And that's not even taking into consideration the dynamic range of even the highest level of professional microphones.

16/44.1 is already overkill

 

Okay, that's fair. I just think that statistics are not really the reality, not that they're untrue, just that people sometimes think that a majority means "pretty much everyone". I'm 23, and male, and whenever I take a hearing test, the aide who's giving the test is like, accusing me of hitting the button when I couldn't hear anything, when I clearly can, and makes me take the test again. (Which is honestly hilarious to me most times)
And yeah, admittedly, it's rare to have 96 decibels of dynamic range in music, but a higher bit depth doesn't really affect the dynamic range much, mostly what it affects is the vertical "resolution", if you will, of the music. Don't think about the peaks, think about the sound waveform as a whole. When it's in a digital file, really the overall volume doesn't change across bit depths, not in practice, anyway.
But now I realize that I'm arguing against my own post now, as well. Sorry, I need to stop posting when I'm half asleep.
Also sorry if I'm seeming argumentative in these posts, I'm not trying to be. I'm just naturally an intense debater, sometimes people think I'm angry when I'm just excited, even in real life.

 

It's not wrong. It doesn't matter if there are no real world recordings with the full dynamic range that 24-bits will afford. That's what 24-bits will give you. That's the math. 24-bits is overkill, but that's the standard. 20-bits is more realistic, but no one is using that these days.

 

Another reason for higher resolution 24 and 32 bit is that editing at that resolution is less destructive than editing a 16 bit file.

 

I like 20-bits myself. Certainly for DVD-A writing, which is something I like to do, 20-bits would give more options of dealing with the fixed maximum data extraction rate from DVD-A.

I think the situation is that with our standard 8-bit computer words, a 20-bit rate doesn't save any storage over a 24-bit rate...so 24-bits are what we get.

 

Fact: 8 bits is more than enough for anybody*. Listen:

brothers-16bit.flac
brothers-8bit.flac

*not really - this is just a demo that the benefits of higher bit depth, especially in "HD audio", are often overstated.

(the 8 bit FLAC is actually a larger file, because the algorithm isn't optimized for 8-bit)

A 56KHz sample rate would exceed the possible hearing limits of anyone that has ever lived, while allowing zero-impact aliasing filters. Sadly, like bit depths above ~20 that exceed any other audio hardware, the extra data of the high-resolution format specifications is wasteful unless you are making music for dogs.

 

You are restating your own claim without addressing any of the arguments against it, and you misrepresent what I said. I said that 16 bits is already overkill for consumer audio, even according to the most extreme audiophile standard. Bit depths above 16 arent even being discussed. And just to nail the point home:

• Even 14 bits is better than any analogue recording device ever designed, and used to be considered audiophile!
• 12 bits is roughly on par with the best studio quality reel to reel recorders and would still be excellent sounding for all but the most esoteric demands.
• Even 10 bits would be fine for most consumer demands, but some would begin to question what's the point of digital at that level, when you can have the same performance with a good consumer grade analogue tape deck.
• 8 bits is really the point when we can all agree is too little, because the noise floor will become noticeable in the majority of recordings. It is rougly on par with a mid-to-low-end cassette player without NR.

This shows how good 16 bits are, there really shouldn't be any discussion. 18 bits or more is ridiculous over over over over over over overkill.

 

You cleverly picked and example where the difference becomes unnoticeable . It has low dynamic range because the bass guitar is playing all the time. A lot of current pop music is like that due to the "loudness war". The other day I found an example on Youtube where someone converted "Gangnam Style" down to 4 bits, the difference to the 16 bits original is almost undiscernable.

 

I've been wondering something for sometime and maybe around here knows the answer. There has been for some time audio players (CD and DVD players, music file players...) that state that do upsampling to 88.2/96/192 Khz and 24 bit od regular red book resolution. My Sony UBP X-800 UHD BD player has a mode that upsamples non HiRes material, sampling frequency x2 and 16 bit to 24 bit. I wonder if these "upsampling players" do something more than add "empty" samples and padding a 16 bit sample to 24 bit padding with zeros. Maybe they also add some kind of dither or noise shaping? I know dither is needed when going downin resolution like in 24 to 16 bit conversion, but is the opposite also true?
I must say that I like how my Sony player sounds with its upsampling on. It has no D/A converters by the way, only two HDMI outs and one digital coaxial out.

 

I picked this because it is definitely not a "loudness war" song or album mastering, and was analog instruments recorded digitally in 1985 (and the first CD to sell 1M copies).



Brothers in Arms (album) - Wikipedia

If I wanted to make it easy on myself, I'd take something acoustic that seems musical, like Jack Johnson's "Banana Pancakes":

 

Your comment made me think some!
True an example with very low dynamic range could be easily done in 4 bit or 8 bit, but going through some of music with what I thought was a lot of dynamic range, I was kind of surprised, that even that stuff, really is maybe 25-35db or so of range at best.

 

Many people get surprised when I tell them that Madonna's Like A Virgin was track to 16 bit digital but mixed down to 12 bit digital that was what Nile Rodgers had around for mixing down the album at Power Station. I've been told or read somwhere that he brought his own Sony Digital recording equipment to Power Station Studios and being mixed down at only 12 bit in the mids 80's it's not the worst early digital recording I've listened to, on the contrary, it sounds quite well.

 

Remember The Art of Noise?



Anything wrong with the audio? Sounds fine? Those Bah and Boh voices were sampled at 8 bits.

 

Even so, it still has its flaws. There is a lack of depth that I always notice. George Michael's "Faith", and INXS's "Kick" albums are examples of when digital recordings started to get good, IMO.

 

You have to take Like A Virgin for what it is, an 80's Pop album by a new artist. It doesn't sound bad, Asia's Astra or the soundtrack for Giorgio Moroder's Metropolis (also tracked and mixed down to Sony digital but at full 16 bit) sound worse IMO.
I agree that by the late 1980's digital recordings started to sound good, at least as good as an early technology with equipment made 30 years ago can sound. The first time I heard a good use of digital equipment was on Van Halen's 1991's album For Unlawful Carnal Knowledge, but digital was only used for mix down, the album was tracked to analog but the end result at least for my taste is a great sounding album. The first full digital recording I heard I really liked and surprised me by it's sound it's the soundtrack from 1994's movie Stargate which was recorded at Air Studios.

 

Not disagreeing but is it when "Digital got really good", or when how they "Used digital to get a good result"?

 

The thing was brickwalled to death!