Sampling Rates and Bits

[FunkDaddyBrown]

Ok........REALLY n00b question, but, what is with all this talk i see about sampling and bit-rates in recording. What so important about them and what exactly are they? Any help would be greatly appreciated.

 

[fuzzrhythm]

Very simply....

Bit-depth is related to dynamic range. The bit depths used in most music you hear is either 16-bit or 24-bit. The higher the bit depth, the lower the noise floor.

Sample rate ultimately represents the frequencies you'll hear.

Research Nyquist Theorem, dither, and digital noise floor. Also analog-to-digital and digital-to-analog converters. (ADC/DAC)

When it comes to recording/mixing, you want to be using 16-bit or 24-bit and 44.1 KHz sample rate to start out.

Your final mixes should be 16-bit 44.1Khz for CD quality.

 

[Waffleness]

The above poster made good points, I would like to add an example:

An audio signal is continuous. A digital signal isnt - its chopped up. The number of chops, and their accuracy is dictated by sample rate and bit depth. For example:

A 16 bit bit-depth will give you 2^16 = 65536 'levels' that your audio signal is divided into. The signal is 'quantised' into these levels. Lets call this peice of info a sample.

The number of times a sample is played a second is the sample rate. I.e. every second 44,100 samples are taken when a sample rate of 44.1k is used. The human ears freq range is 20Hz to 20kHz. To accurately sample a 20kHz signal in the digital world, it must be sampled with a rate of 40kHz due to 'Nyquist', otherwise you get nasty freqencies appearing in the low end of your audio signal. Its the same concept as to why car tyres appear to be rotating in reverse.

Thats it all very simply.

 

[suprstar]

It's kinda like the difference in scanning a photo at 72 dpi or 300 dpi or 1200 dpi. Photos/sounds are continuous, digital images/wav files are chopped up. dpi/sample rate is how many chops. The more chops, the finer more accurate representation you have of the original photo/sound.

 

[FunkDaddyBrown]

Thanks, that helped a lot. One more quick question though fuzzy, by lowering the noise floor do you mean there'll be less noise or more noise

 

[fuzzrhythm]

Lower noise floor means there is less noise. With 24-bit digital audio, the noise floor created by the analog-to-digital conversion is below that of any other analog piece in your chain. At this point, digital noise floor is a non-factor.