Poll - how important IS 96 freq. in 24 bit system?

  • Thread starter Thread starter zip
  • Start date Start date

Just how important is the 96 freq. in a 24bit digital system???

  • Must HAVE!!

    Votes: 5 7.4%
  • Will definitely help sound

    Votes: 19 27.9%
  • Will slightly or somewhat help sound

    Votes: 20 29.4%
  • No real or substantial difference

    Votes: 24 35.3%

  • Total voters
    68
eyeteeth said:
Vox... I have never seen rates broken down that way.

My understanding, is in brief... The highest record able frequency, is half the sample rate... 44.1, 22khz; 88.1, 44khz... etc. which as mentioned is still far beyond a humans hearing range, especially as we get older. The beginning of the problem, is what the A/D converters do with the frequencies... harmonics we can't here. Think of all the harmonic overtones present in a cymbal crash. Even though we can't "hear" those frequencies, they all enhance, and effect the frequencies we CAN hear. So at 44.1, these harmonics get chopped at 22khz... the result however, is the D/A converters manifest these frequencies out of the range as digital noise at a corresponding frequency in the bottom end of the spectrum.

So the higher the sample rate, the truer sound capture you will have as there is a more accurate representation of harmonic overtones, and the less digital noise created from cutting frequencies.

I have, it's the easier way of looking of how it works to the simplest form. What your saying is correct, but also, since the humans hear 20-20khz, and this is all digital, the actual recorded rates wouldn't be doubles 44.1, 88.2 etc. the extra they laid on top (instead of going 20, 40, 80, etc) were for basically error corrections, aliasing sounds & others if I remember correctly.
 
Also the 'brick wall' filter which is neccessary for 44.1 can be much more benign in a 96k sampling rate and a lot of people feel that those filters are a big contributer to harsh sound.

And there are a lot of studies that indicate that high freq's above human hearing range do affect how the equipment reproduces the lower freq's we can hear and also how we hear them.
 
Blue Bear Sound said:

By George! I think that dumbass Bruce thinks that 88.2 being a multiple of 44.1 somehow offers a smoother transition back down to 44.1 than 96 would! LOL

I knew he was a dumbass, but this is the ultimate proof! :)
 
I only use 96 if I'm doing a solo guitar recording or something with just 1 to 2 tracks. Otherwise I use 44.1/24
 
OLLLLLLLLLLLLLLLLLLLLLLLLLLLLLD THREEAADDDDDDDDD


check the post date guys
 
Yea... I noticed that... afterwards...

What's with all the old threads popping up? I noticed many are several years old... but this one trapped me anyway.
 
bennychico11 said:
OLLLLLLLLLLLLLLLLLLLLLLLLLLLLLD THREEAADDDDDDDDD


check the post date guys


What is your point? People STILL ask the question!
 
eyeteeth said:
Yea... I noticed that... afterwards...

What's with all the old threads popping up? I noticed many are several years old... but this one trapped me anyway.

people doing searches and coming across polls. When you vote on the poll it automatically puts it at the top of the forum. And there it stays alive for awhile....


I say as I just bump it back to the top :p
 
Ford Van said:
What is your point? People STILL ask the question!

I know...but people above us are also trying to respond to a poster who may not even be active on this site anymore. Many people just don't notice the date and start posting away.
It was probably just a polling fault that this got bumped to the top again...
 
how does the number of times a second you sample something effect the magnitude of what you can sample anyway? i thought that was stored in the bit rate... just shows what i know :o
 
Cazzbar said:
how does the number of times a second you sample something effect the magnitude of what you can sample anyway? i thought that was stored in the bit rate... just shows what i know :o

:confused" Can you ask a question in the english language now please? ;)

You might stay away from terms like "bit rate", as the reality is it is the "big depth".

Define what you mean by magnitude.
 
If I remember correctly the bit depth refers to the number of amplitude values a given sample can have.

The sample rate refers to the highest frequency that can be recorded.

Someone correct me if I'm wrong.
 
Ford Van said:
:confused" Can you ask a question in the english language now please? ;)

You might stay away from terms like "bit rate", as the reality is it is the "big depth".

Define what you mean by magnitude.

my mind's completely boggled at the moment, through staring at other things! so i don't think i'm gonna make sense, i should have definitely stayed away from this thread anyway :)

yep, i did indeed mean bit depth...

ok, you have a frequency/rate at which you sample something, and then you need to store the result of that sample... and the more bit depth you have to store the result of the sample, the closer your approximation to the real thing will be? (jeesh this is going badly isn't it)

so i was just wondering how the rate at which you sample something effects the result of a particular sample (although i appreciate things will sound bad if you're only capturing once every 10 minutes!)

(my ignorance on the subject should be clear enough for you to put me straight now anyway lol)

edit: it's something to do with the fact waves go up and down isn't it :p and that you'll need to sample at twice the frequency to work out a wave's amplitude... or something, i'll sit at the back
 
Last edited:
Ford Van said:
By George! I think that dumbass Bruce thinks that 88.2 being a multiple of 44.1 somehow offers a smoother transition back down to 44.1 than 96 would! LOL

I knew he was a dumbass, but this is the ultimate proof! :)


well, it does really.



A simple halving of samples is a much easier, cleaner and smoother transition than having to compute 96khz--->44.1khz.


If you think of it, at some point the computer is going to have to "make up" samples.


I think (as do many other professionals i've talked to) that the benefits of the smoother transition back down to 44.1 outweigh the tiny difference than the extra 8khz sample rate.
 
after reading thru all of this n trying to grasp everything i want to ask somthing just to make shure

im using acid studio right now n i could only record my vocals at 16/44
i bought the audiphile 2496 n im about to buy the m-powered PT

i juss wanted to know how big of a difference the jump from 16/44 to 24/44 or even 24/96 would be

oh n im using an mxl 990 for now...
 
Cazzbar...

Let me try an analogy... may help... or make things worse.

Lets take the computer monitor you're staring at, and a really high resolution picture...

Make the picture your desktop...

Let's make Frequency your desktop resolution, and Bit rate, the amount of colors you can display. (See where this is going yet)

If you turn down the resolution, the picture starts to get blocky, and looses detail of the image.

If you take the picture in full resolution, but decrease the number of colors, you loose definition and clarity...

The higher the sample frequency, the more accurate representation of the audio wave, the higher the bit rate, the better definition the system has of the sample.

They go hand in hand to an extent.

Crap... now I think I just confused myself. :D
 
MessianicDreams said:
well, it does really.



A simple halving of samples is a much easier, cleaner and smoother transition than having to compute 96khz--->44.1khz.


If you think of it, at some point the computer is going to have to "make up" samples.


I think (as do many other professionals i've talked to) that the benefits of the smoother transition back down to 44.1 outweigh the tiny difference than the extra 8khz sample rate.

I am sorry, but you have absolutely NO idea how sample rate conversion is done, and that is obvious from this post!

You can talk to all the professionals (who also seem to misunderstand how it is done too!) you want who agree with your totally wrong premise. That does not make your point any more valid.

When audio is sample rate converted, it is totally re-sampled at the new sample rate. The new samples are based upon algorithms that predicts how the original audio will make a wave form that can be re-sampled by an algorithm that predicts the new wave form of the audio.

The only "efficiency" that you might gain is that there are less samples to process at 88.2, but in NO WAY OR FORM is the process any more "simple" because the original sample rate is some number dividable by the target sample rate.

Actually, the source quality rule comes into effect, and quite the opposite is true! By having more samples available for re-sampling, and better re-sample is possible!

The biggest deciding factor on a sample rate conversion is the quality of the algorithm that is used, and the original quality of the file. If you can argue that 96khz potentially SOUNDS better than 88.2, then you can successfully argue that a SRC from 96 to 44.1 will sound better than a 88.2 to 44.1 will!

If you just think of it like this, it makes a lot more sense: The sample rate conversion is only as good as the source quality.

Get it now?
 
eyeteeth said:
Cazzbar...

Let me try an analogy... may help... or make things worse.

Lets take the computer monitor you're staring at, and a really high resolution picture...

Make the picture your desktop...

Let's make Frequency your desktop resolution, and Bit rate, the amount of colors you can display. (See where this is going yet)

If you turn down the resolution, the picture starts to get blocky, and looses detail of the image.

If you take the picture in full resolution, but decrease the number of colors, you loose definition and clarity...

The higher the sample frequency, the more accurate representation of the audio wave, the higher the bit rate, the better definition the system has of the sample.

They go hand in hand to an extent.

Crap... now I think I just confused myself. :D

Please quit calling it "bit rate". There is no such thing as "bit rate" in digital audio. There is "big depth".

Sample rate.

Bit depth.

One is how many "pictures" you can take per second.

The other is how detailed each picture can be!

There are 1 bit coverters out their that take pictures at obscene sample rates!

Get it?
 
Last edited:
Back
Top