24bit/96khz - computer language unlocked

So this will show how they relate (see the link at the bottom)

1 2 4 8 | 16 32 64 128 | 256 512 1024 2048 | 4096 8192 16384 32768 | 65,536 131,072 262,144 524,288 | 1,048,576 2,097,152 4,194.304 8,388,608

This is 24 bits in binary. Binary lets you represent any number using 1’s and 0’s (actually it works RIGHT to left with the RIGHTMOST number being 1, then 2, etc.

So let’s just use 8 bits for a moment – If I wanted to represent the number 34, in binary

128 64 32 16 8 4 2 1 – is what is represented

0 0 1 0 0 0 1 0 ( a 1 for 32 and 1 for 2 =34) – computer code – 1’s and 0’s - = On or OFF

This is how computer code is generated..Binary is converted to hexadecimal and you get your data representations

This came as a result of someone asking about 24bit/96khz with a great explanation here:

http://tweakheadz.com/16-bit-vs-24-bit-audio/

24bit/96khz is PLENTY

Since there is NO AUDIO EQUIPMENT IN EXISTENCE that can successfully take advantage of all 24 bits, yes, 24 bits is "good enough".

tblob said:

Since there is NO AUDIO EQUIPMENT IN EXISTENCE that can successfully take advantage of all 24 bits, yes, 24 bits is "good enough".

Click to expand...

Bingo, remember kids... you're only as good as your noise floor

Thats not how it works...LOL

eyekon218 said:

I would have to guess that 99% of the music I have has been ripped from 16 bit 44.1khz encoded cd's.

https://www.youtube.com/watch?v=4nUkNvl8mQw

^ my hearing drops off at about 48 seconds in.

I am seeing these sites out there like hdtracks.com that are encoding in 24/192. Why? I dont see the point of covering an audio range that goes up to 192khz when human hearing doesnt go above 20-22khz. The Nyquist frequency tells us for waveform reconstruction double is all that is required which is why compact discs are recorded at 44.1khz.

Click to expand...

Its is not frequency response
24bit - is the DEPTH (think of it as handling tracks) - originally CD's were recorded at 16 bit that is 16 levels of audio capture or 65,536 levels. AT 24bit that number jumps to 16,777,216 levels..PER SECOND !!!.Yeah

Details here
http://tweakheadz.com/16-bit-vs-24-bit-audio/

arcorob said:

Its is not frequency response
24bit - is the DEPTH (think of it as handling tracks) - originally CD's were recorded at 16 bit that is 16 levels of audio capture or 65,536 levels. AT 24bit that number jumps to 16,777,216 levels..PER SECOND !!!.Yeah

Details here
http://tweakheadz.com/16-bit-vs-24-bit-audio/

Click to expand...

Actually it's not "per second" it's per sample. (more impressive)

Apologies to the OP, none of this has much to do with the original post.

My bad...

tblob said:

Actually it's not "per second" it's per sample. (more impressive)

Apologies to the OP, none of this has much to do with the original post.

Click to expand...

ooops...Well, the link explains it MUCH better than I could..LOL

Thanks for the correction !!!

Bit depth determines the SN ratio.

Now you need to investigate sampling rate. Sampling rate is per second. Sampling rate determines the bandwidth.

Bandwidth relates to the highest frequency you can capture. It's actually independent of bit depth. This is why you can have 16, 24 or even 32 bit signals at a 44.1KHz, 48KHz, 88.2KHz, 96KHz, 176.4KHz or 192KHz sampling rate. Nyquist theory says that the sampling rate (frequency) has to be twice as high as the highest frequency signal you wish to capture.

Lastly there is 2,822,400Hz (2.8MHz) which is the sampling rate for SACD. SACD is a 1 bit Delta sigma type system. Delta sigma encoding does not use word lengths (bit depth). It uses only a single 1 or single zero for each sample. For SACD delta sigma modulation is known as DSD (direct stream digital)

eyekon218 said:

I am seeing these sites out there like hdtracks.com that are encoding in 24/192. Why?

Click to expand...

Why? Because it sounds better! Simple as that.

I've taken several nice 192 cuts, down sampled them to have identical versions at 96 and 44.1. Then burned those to disc. Now I can set down and play the same exact material that's only at the different rates. It's not hard at all to hear the quality change. It's just a smoother sound at higher data rate and just a little rougher character as you get down to 44.1.

Once you hear the changes you get it. At the higher data rate it's starting to get more and more like good analog. And I don't listen loud enough to think I'm hearing down to a noise floor nor do I think I have dog hearing. So I don't buy the idea that all higher data rate and depth do is extend above the hearing range nor just lower floor that's below your equipment noise or surrounds natural ambediant level. The character of the notes are more natural.

That's not to say their isn't some really good material at 44.1. I've some that are tremendously moving music on redbook. But over time I've found myself preferring, if playing digital, to do it from something that's at one of the faster data rates just because it's more real performers in front of you. I don't have a whole lot of listening time and I want to get as much pleasure out of it as I can. More realistic recreations of the recorded event give me that!

I'd suggest spending time seeking out a chance to hear more material and less time spent reading others, like mine, thoughts. Once you hear it yourself and know, it's a whole lot easier to know what stuff is just junk science and coming from a theoretical position and not from the real thing. Listening instead of just yapping about it!

EV3

evsentry3 said:

Why? Because it sounds better! Simple as that.

I've taken several nice 192 cuts, down sampled them to have identical versions at 96 and 44.1. Then burned those to disc. Now I can set down and play the same exact material that's only at the different rates. It's not hard at all to hear the quality change.

Click to expand...

Are you saying you put both 96K and 44.1K on a redbook audio CD? Or some other kind of disc?

A DVD.

I use a program from Cirlinca which will burn DVD-A's. My Oppo player doesn't care if one track is at 192/24 and the very next something else.

EV3

evsentry3 said:

I've taken several nice 192 cuts, down sampled them to have identical versions at 96 and 44.1. Then burned those to disc. Now I can set down and play the same exact material that's only at the different rates. It's not hard at all to hear the quality change.EV3

Click to expand...

Use a DAW tool like Waves, Audition or Pro-Tools to process the audio. If you are just using some freeware like Audacity then I'm not surprised that you might hear a difference.

No

tblob said:

Are you saying you put both 96K and 44.1K on a redbook audio CD? Or some other kind of disc?

Click to expand...

what he said was he downsampled..meaning made two versions....think of if he made an mp3 copy...it would be a lesser (compressed) version of the original.

Let me dumb this down a bit

In the early days of digital, people said it sounded cold, flat, not as good as analog. Why?

Analog is represented as a continuous wave. Think of a line.

Digital is a series of data blocks. Think of the music as - - - - - - - - -
Not a continuous wave but blocks representing the sound with gaps from one byte to the next.

Now the less data you have to describe the music the more it will NOT sound like analog. But imagine if you will, so much data per block that the gaps are imperceptible.

Look at it this way. If I described a bit of music like this
--------
vs
---------------------------------------------------------------------------

Which is going to give me more sound ?

I know this is a dumbed down version but the point is, todays digital files and DACs can rival analog because of the higher quality files and units to play them..

I know there are folks who can explain the exact process (sort of like my link) but too often the reality of the why and what it means gets lost in the how technical details....LOL

Digital critics are a little too fond of citing the "block" shape of digital music recreation. It may be read from sources like that, but there are component response times as voltage control circuits are turned on and off to take a little of the "edge" off the wave transitions. Again, "a little", though not enough to completely silence critics. I'm no electronics expert, but I don't think it would be, or it as not been, impossible to design in ways to smooth out those hard corners digital critics use in their kindergarten drawings of music signals from a recording format toward whose extinction they are so energetically biased.

LOL...good post

ec1st said:

Digital critics are a little too fond of citing the "block" shape of digital music recreation. It may be read from sources like that, but there are component response times as voltage control circuits are turned on and off to take a little of the "edge" off the wave transitions. Again, "a little", though not enough to completely silence critics. I'm no electronics expert, but I don't think it would be, or it as not been, impossible to design in ways to smooth out those hard corners digital critics use in their kindergarten drawings of music signals from a recording format toward whose extinction they are so energetically biased.

Click to expand...

I hear ya...

The good news is, the more data and processing ability, the better the sound. I started this only because of my recent (new) DAC experience..

ec1st said:

Digital critics are a little too fond of citing the "block" shape of digital music recreation. It may be read from sources like that, but there are component response times as voltage control circuits are turned on and off to take a little of the "edge" off the wave transitions. Again, "a little", though not enough to completely silence critics. I'm no electronics expert, but I don't think it would be, or it as not been, impossible to design in ways to smooth out those hard corners digital critics use in their kindergarten drawings of music signals from a recording format toward whose extinction they are so energetically biased.

Click to expand...

Any competently-designed DAC includes a way of smoothing those edges. It's called an analog reconstruction filter, and its job is to get rid of signal energy above the Nyquist frequency (half the sampling frequency).

There are incompetently-designed DACs out there, most of them non-oversampling and some at very fancy prices, that lack the reconstruction filter and send the signal and all of the out-of-band noise on to your preamp, amp, and speakers. IMO DACs of this type have no place in a high-fidelity music system.

The easiest way I have ever found to explain sampling rates to people is to simply ask them to imagine a ramp, like a handicapped ramp. It is a long, smooth, gradual incline or decline, with no individual stops along the way. You gradually go up or down on it, right?

Now imagine a set of stair steps next to the ramp. Each step is a digital sampling point. You still go up or down, but there is a distinct 'bump' every so often (the sampling rate). If you have fewer steps, you have to step higher with each step up you take. If you have more steps, you have a lower step to take, but you take more of them to get to the same place the ramp next to you goes.

Since music is made up of waves that go up and down, it is difficult to accurately recreate that waveform with a small number of samples per second. The higher the number of samples, the more 'steps' but the smaller each step will be. In other words, higher sample rates 'smooth out' the curve until it starts to resemble the original analog waveform.

Or imagine if you had to draw a circle using only straight lines. If you used only four straight lines, you'd have a square - not very circle-like. If you used thousands of very short straight lines, it still would not be a circle, but it would look more-or-less like one from a distance. The smaller you make the straight lines, and the more of them you have, the more the final drawing resembles a circle.

When you go to the movie theater, you being fooled into thinking the pictures are moving. In fact, each frame of a (film) movie is a single, non-moving photo. But they stream them onto the screen so fast (24 to 72 frames per second) that the human brain is fooled into thinking the picture is actually moving. The phenomena is called "persistence of vision."

The human ear is likewise fooled by digital representation of analog sound. At a certain sampling rate, the digital ones and zeroes can approximate the curved wave of an analog sound well enough that the human ear believes that is what it is hearing.

What remains is the question as to where this takes place, at what sampling rate.

I will not presume to give an answer. It's far too angry-making of a question. Some say that lower sampling rates, such as those used by the audio CD, are well beyond what is necessary for accurate reproduction of music, and that anything above that is simply wasted. Others believe that the higher the sampling rate, the better it sounds, period. Attempts to perform objective testing to come to a conclusion that all can live with always end in bitter acrimony and harsh language.

So let us say that there is a sampling rate, call it "X," which is satisfactory for listening to digital reproduction of analog music. You will have to decide the value of X for yourself.

What you may call analog is not really free of discrete steps. Magnetic tape used for recording has finite size magnetic domains. Thus it is not possible to record fully continuous signal on tape. 24 bits, as was mentioned above is more than enough - best audio ADCs and DACs have only 20 bits of resolution. The rest is noise.192K/second sampling is enough to successfully record CD4 encoded vinyl and play it back. So there is no loss here either.

botrytis said:

Wrong - just need to have the proper DAC, is all. If you believe that then your ANALOG system wouldn't work either (although there is more headroom with digital than with analog).

Click to expand...

Which DAC has a noise floor < -144dB?

arcorob said:

actually it works RIGHT to left with the RIGHTMOST number being 1, then 2, etc.

Click to expand...

That's how all numbers work.

208 decimal: that's 8 ones, 0 tens, 2 hundreds.

in binary:
11010000: that's 0 ones, 0 twos, 0 fours, 0 eights, 1 sixteen, 0 thirty-twos, 1 sixty-four, and 1 one-hundred-twenty-eight = 208