24bit & 96khz?

[mkathery]

Hi All,
I know this is a silly question but I really want to know:
What is the meaning of 24bit & 96khz?
Do you get higher quality by recording at (20bit @ 96khz) or by (24bit @ 48khz)?

Best Regards,

[mindnoise]

24 bit refer to the dynamic range a signal can have without distorting/clipping
or how quite a signal can be before it disappears in the backgroundnoise level,
basically.
e.g. 16bit mean 96db range (CD Player) while 24bit mean 144db this is where you get deaf or at least do some irreversible damage to your ears.
for every additiona bit (e.g. 16 bit, 17 bit...) the volume is doubled which refers to an volume gain of 6db in the real world.
where in the digital domain 0 dbFS (FULLSCALE e.g.24bit used) is translate to +18dbU in the analog level, depends on what measurement system you use or your preference.

while 96k samples per second just relates to the accuracy by which an analog signal is translated into a digital and how accurate signal´s placement is on the timeline.
A digitized signal will be the more a accurate representation of it´s analog "counterpart" the faster it is sampled, since you also get the more spectral components of a signal more accurately.
The benefits and whys of higher samplerates are also a quite diversive discussion today, still.

I guess that´s the basics.

for more info go here: (esp. Levels)
http://www.theprojectstudiohandbook.com/directory.htm


regs

[mkathery]

thank you very much mindnoise for the info, and that was very usefull link.

regards,

[Duardo]

e.g. 16bit mean 96db range (CD Player) while 24bit mean 144db this is where you get deaf or at least do some irreversible damage to your ears.
for every additiona bit (e.g. 16 bit, 17 bit...) the volume is doubled which refers to an volume gain of 6db in the real world.

That's not actually quite right...24 bits doesn't get any "louder" than 16, it actually lets you get "quieter"...while it's true that 24 bits theoretically give you 144 dB of dynamic range, but actually with today's converters it's more like 100-120 dB...you can't get any louder than 0 dB FS, so what you get with more bits is more detail at lower levels.

while 96k samples per second just relates to the accuracy by which an analog signal is translated into a digital and how accurate signal´s placement is on the timeline.

Higher sampling rates have nothing to do with the accuracy of a signal on a timeline...they let us capture higher frequencies, period. Recording at 96 kHz is more accurate in that it captures more frequencies, but 48 kHz is perfectly capable of capturing all of the frequencies we can hear (although depending on the quality of the converter, a 96 kHz converter may sound better due to filter designs, etc). But the signals are no more accurately "placed" on the timeline at 96 kHz than they are at 48 kHz.

A digitized signal will be the more a accurate representation of it´s analog "counterpart" the faster it is sampled, since you also get the more spectral components of a signal more accurately.

It's true that it will be more accurate, but that's simply because it's capturing more frequencies. As to whether it will sound any better, it depends on the converter...but with proper filter design, there's no reason that 48 kHz can't sound just as good as 96 kHz.

-Duardo

[mindnoise]

Of course you right about the details of my rough overview. I just wanted to give direction and not spoil all the learning success. Therefore the link!

as I said:

The benefits and whys of higher samplerates are also a quite diversive discussion today, still.

regs

[BenTerry]

Just remember that in the digital realm whenever you start to increase sample rates you are creating larger files and have less available tracks. Also remember that everything is going down to 44.1 and 16 bits in your CD player, so make sure that you don't leave out that 15th vocal harmony because you think that the higher sample rate is going to make everything sound that much better because in the end it's not going to matter all that much, especially with the converters on the 002 or MBox. 48K or 44.1K are fine and until you have an HD3 system or some other great converters with more tracking capabilities then just stick to 48 and do that 15th harmony.

[where02190]

To give an analog analogy, think of bit depth as tape width, and sample rate as tape speed.

[kraster]

The Nyquist theory states that the sampled data must be double the range of the maximum frequency the brain is capable of decoding and then some. One of the drawbacks of sampling is that at half the sampling rate a subharmonic frequency is generated so at 44.1k a low pass filter is applied to anything above 20k so this harmonic will be inaudible. Doubling the sampling rate to 88.2 or 96k means this subharmonic is at 44.1k and the filter is applie to any siganl above 40k which is well beyond the human range of hearing. Bit rate and sample rate or anlagous to a film strip. Bit rate being the size of the frame being captured and the sample rate being the frame rate. Larger Bits increase the the lower threshold of the dynamic range allowing you to record quieter signals with reduced noise.

[Paver]

Ok, I'm still after all these years of using pro tools wondering if it's worth tracking in 24 bit and using all my plug ins and dithering down with L2 or would I be just as well off tracking in good ol' 16 bit. At the end of the day, does it sound any different?

[drthfdr]

Jeeesus..

... no more beer for you...

24 bit refer to the dynamic range a signal can have without distorting/clipping
or how quite a signal can be before it disappears in the backgroundnoise level,
basically.
e.g. 16bit mean 96db range (CD Player) while 24bit mean 144db this is where you get deaf or at least do some irreversible damage to your ears......

<snip>
regs

Husky