[Danz03]

You are using the Softube Trident A-Range EQ plug-in for you media player?

 

I'm talking about a VST plugin that I'm using in my everyday media player.

 

[Ham Sandwich]

I'm using VST plug-in EQs as well.  Both Foobar and J River Media Center can use VST plug-ins.  So can some other media players.
 
 
There is no reason to stick with the stock 5 to 10 band EQs that come with most media players.  Those EQs don't have enough control to be able to do corrective EQ fixes and they can sound poor as well.

 

[PelPix]

It's true that you can't hear the difference, but it makes a big difference when you're actually doing mixing and you may, in what we call a "worst case scenario," have to increase the gain until the noise is audible.  In these cases having a bit depth higher than 16 bit becomes practical.  Not for listening, but for processing.
I master in "Overkill mode" (192khz/64-bit float).  It has no purpose whatsoever beyond compatibility.

 

[kwkarth]

Quote:

It's true that you can't hear the difference, but it makes a big difference when you're actually doing mixing and you may, in what we call a "worst case scenario," have to increase the gain until the noise is audible.  In these cases having a bit depth higher than 16 bit becomes practical.  Not for listening, but for processing.
I master in "Overkill mode" (192khz/64-bit float).  It has no purpose whatsoever beyond compatibility.

And your final product after post, is no doubt, of higher quality as a result.  
 

 

[PelPix]

That too, because it allows more smooth sound after high quality dithering than something natively 16-bit.  It's the same principle as video game anti-aliasing.

 

[aspenx]

Quote:

You are using the Softube Trident A-Range EQ plug-in for you media player?

I want to know exactly what EQ too. Currently I'm using Electri-'s parametric EQ with foobar2k through a VST wrapper from George Yohng.

 

[xabu]

Quote:

2 = .... It is in fact the fundamental tenet of the Nyquist-Shannon Sampling Theorem on which the very existence and invention of digital audio is based. From WIKI: “In essence the theorem shows that an analog signal that has been sampled can be perfectly reconstructed from the samples”....
 

(I did not read the whole thread ... so if the following was already mentioned ... just ignore )
 
Well, I'm somewhat with you regarding the whole bit mathematics but regarding Nyquist your cite is a little bit symplifying:
 
"In practice, neither of the two statements of the sampling theorem described above can be completely satisfied, and neither can the reconstruction formula be precisely implemented. The reconstruction process that involves scaled and delayed sinc functions can be described as ideal. It cannot be realized in practice since it implies that each sample contributes to the reconstructed signal at almost all time points, requiring summing an infinite number of terms. Instead, some type of approximation of the sinc functions, finite in length, has to be used. The error that corresponds to the sinc-function approximation is referred to as interpolation error. Practical digital-to-analog converters produce neither scaled and delayed sinc functions nor ideal impulses (that if ideally low-pass filtered would yield the original signal), but a sequence of scaled and delayed rectangular pulses. This practical piecewise-constant output can be modeled as a zero-order hold filter driven by the sequence of scaled and delayed dirac impulses referred to in the mathematical basis section below. A shaping filter is sometimes used after the DAC with zero-order hold to make a better overall approximation."
 
http://en.wikipedia.org/wiki/Nyquist-Shannon_sampling_theorem
 
Well, and there is also something more about the bit mathematics ...
 
Quote:

... each bit of data provides 6dB of dynamic range ...

 
So lets say
 
with 1 bit we would resolve 6 dB into 2 values (e.g. 0 dB and 6 dB, nothing in between)
with 2 bit we would resolve 12 dB into 4 values (e.g. 0,4,8,12 dB, nothing in between)
with 3 bit we would resolve 18 dB into 8 values (more values, finer resolution)
with 4 bit we would resolve 24 dB into 16 values (more values, finer resolution)
.
.
.
with 16 bit we would resolve 96 dB into 65536 values (that means 1 dB would resolve into 684 values, much finer resolution)
.
.
.
with 24 bit we would resolve 144 dB into 16777216 values (that means 1 dB would resolve into 116508 values, much much finer resolution, perhaps almost overkill

)
 
 
I would call that a step rise in volume resolution ...

 

[khaos974]

Or rather think of it this way, what is the tiniest difference of air pressure that you can hear, any difference of air pressure below that cannot be head? That will be the 2nd step on a scale from 0 to 65635.
On 16 bit, it means that we hear a continuous variation of volume level from that single step to 96 dB higher.
On 24 bit, it means that we hear a continuous variation of volume level from that single step to 144 dB higher.
 
=> The steps are not more detailed, you don't get more resolution, you get more dynamic range.

 

[xabu]

Quote:

Or rather think of it this way, what is the tiniest difference of air pressure that you can hear, any difference of air pressure below that cannot be head? That will be the 2nd step on a scale from 0 to 65635.
On 16 bit, it means that we hear a continuous variation of volume level from that single step to 96 dB higher.
On 24 bit, it means that we hear a continuous variation of volume level from that single step to 144 dB higher.
 
=> The steps are not more detailed, you don't get more resolution, you get more dynamic range.

Correct, there is a limitation of what difference you can hear.
 
If the tiniest difference that could be heard would be 2 * (6 dB / 65635) = 0,0002 dB ...
 
you would be right, with this picture, there were some amount of non audible steps in bigger resolution ... but bigger resolution it were nonetheless ... (... and more dynamic range, no question about that ...)
 
But:
 
Edit: Don't get the meaning of what I wrote above anymore ... I think somehow I tried to make sense of what you wrote
 
------------------------- snip
 
Due to the fact that the dB scala (sound pressure level) is not linear but logarithmical based on the sound pressure measured in Pa we had to calculate the whole stuff better in Pa ... because it's not dB values what is stored ...
 
So in fact 1 dB is the smallest amount of difference in sound pressure level the human ear can recognize (all related to 1khz),
 
that would be 0,00002(244) Pa at 1 dB above 0 dB and e.g.  0,02244 Pa at 1 dB above 60 dB, etc.
 
But to hold/represent these values which range from 0,00002 Pa to 20 Pa which translates to (0 - 120 dB) in a binary format, you would definetely need more than 16 bit.
 
Even for 0 - 96 dB (0,00002 Pa - 1,26191 Pa) you would need more than 16 bit.

 

[xnor]

No you don't, and from the Handbook of Recording Engineering the dynamic range of music as normally perceived in a concert hall doesn't even exceed 80 dB...
With popular music you're in the range of 10 dB..

 

[xabu]

Quote:

No you don't, and from the Handbook of Recording Engineering the dynamic range of music as normally perceived in a concert hall doesn't even exceed 80 dB...
With popular music you're in the range of 10 dB..

 
I don't know exactly what of my statements you refer to with "No you don't", if it is the last two you're wrong.
That's just mathematics.
You need more than 16 bits to hold all values between 0,00002 and 1,26191.
There's nothing to discuss.
 
Regarding the rest I've to admit I must pass until I learned more about the whole mastering stuff ... very interesting in any case!
 
But nevertheless with 16 bit you have a pot of max. 65536 values available per sample.
If that's enough for current mastering standards and digital domain specifics than it's enough.
Though the audible range human ears can perceive exceeds this pot of values clearly ...

 

[nick_charles]

Quote:

Quote:

No you don't, and from the Handbook of Recording Engineering the dynamic range of music as normally perceived in a concert hall doesn't even exceed 80 dB...
With popular music you're in the range of 10 dB..

 
I don't know exactly what of my statements you refer to with "No you don't", if it is the last two you're wrong.
That's just mathematics.
You need more than 16 bits to hold all values between 0,00002 and 1,26191.
There's nothing to discuss.
 
Regarding the rest I've to admit I must pass until I learned more about the whole mastering stuff ... very interesting in any case!
 
But nevertheless with 16 bit you have a pot of max. 65536 values available per sample.
If that's enough for current mastering standards and digital domain specifics than it's enough.
Though the audible range human ears can perceive exceeds this pot of values clearly ...

Mathemetically you are correct, as a matter of min and max levels perceived and we are talking about zero background noise then sure 16 bits is inusfficient to capture that range noiselessly. Pragmatically though in any environment there is so much background noise that several low order bits of resolution get lost in noise. Unless you can listen in an anechoic chamber the low level signals are drowned out and the difference between 16 bits and 24 bits is academic rather than essential.
 

 

[D. Lundberg]

 
Quote:

So lets say
 
with 1 bit we would resolve 6 dB into 2 values (e.g. 0 dB and 6 dB, nothing in between)
with 2 bit we would resolve 12 dB into 4 values (e.g. 0,4,8,12 dB, nothing in between)
with 3 bit we would resolve 18 dB into 8 values (more values, finer resolution)
with 4 bit we would resolve 24 dB into 16 values (more values, finer resolution)

 
You'll actually get everything inbetween, it's the noise floor that limits the dynamic range of a properly dithered digital signal.
The amount of bits you have represents the amount of error signal you'll get from quantization (rounding to steps), and the error signal will (if the signal is properly dithered) be part of the noise floor.
More bits will just lower the noise floor and give you more dynamic range. Nothing is lost unless the signal goes below the noise floor of the system.
 
And you can shape the noise floor to greatly extend the dynamic range at certain frequencies. You can, for example, have 120dB of dynamic range (20Hz-20kHz) in a 1 bit system (like SACD).
So when it comes to sound quality it doesn't matter if you have 16,7 Million or 2 qauntization steps, just that the noise floor is low enough to not be audible or mask the signal.

 

[xnor]

^ Precisely.
 
xabu, here are some files I created to play around with: (all 16 bit, 44.1 kHz)
http://www.mediafire.com/?hy4n6e9cg1jt7s0 - 100, 1k and 10 kHz sine at below -80 dB
http://www.mediafire.com/?zq8t1ovnx7dn4q1 - the same, but below -100 dB
 
And the spectrum of the 2nd file:
 

 
As mentioned before, dithering (1 of the 16 bits) was used to keep the noise floor down. And there are a lot of options to shape that noise to make it less audible than white noise.
 
Also (try to) listen to those files. 

 

[xabu]

Quote:

And you can shape the noise floor to greatly extend the dynamic range at certain frequencies. You can, for example, have 120dB of dynamic range (20Hz-20kHz) in a 1 bit system (like SACD).
 

 
Because you have only 1 bit you have more samples to "emulate" the missing bits (2.8224 MHz)  ... DSD and Sigma Delta DACs work completely differently from Ladder DACs and PCM.
And and as far as I know they use now up to 5 bits in modern Sigma Delta DACs.