Sample Rates, Bits, what is it all about?
Posted on Aug 15, 2008 08:58 pm
Member Since: Jan 29, 2008
Hey guys im having trouble understanding sample rates and bits and all that lingo. I have grasped the fact that changing them and adjusting them create much better audio quality and have something to do with how much your computer can handle and latency and stuff (I think anyhow). I basically want to know this. What sample rate should I set on my Fireface 800 interface to get the best audio quality. I do have a quiet fast computer BTW so it should be able to handle it. The Fireface also has a fine tune and active thingo that has to do with sample rates. These are the things measured in kHz and bits. I hope im not confusing people but some light on the subject would be great.
[ Back to Top ]
Noize2uCzar of MidiAdministrator
Aug 15, 2008 09:32 pm The sample rate on CD is set to 44.1 kHz which is the general default and is very good for almost every use. If you need a little more headroom in the digital world for editing and effects I would suggest going up to 48 kHz which is what I use for most recordings most of the time.
Since: Apr 04, 2002
I still do a lot of sound design and sample and loop work. For that depending on the need I will go to 96 kHz or have even gone as high as 192 kHz which is absolutely unnecessary for general home recording.
On the bit depth thing, which is different then the bit rate by the way you can use 16 bit and get good quality. But I prefer using 24 bit and then dithering down to the 16 bit bit depth which again is the default for Cd quality sound.
Aug 15, 2008 09:34 pm What do you mean in terms of headroom? And all this fine tune and active controls for sample rate, what are they there for. Will I ever have to really use them?
Aug 15, 2008 09:35 pm Dumbing this down - Not the last word...
The sampling rate will dictate the frequency response. The Nyquist frequency is half the sample rate. 44.1kHz (the audio standard) will reproduce 22,050Hz (slightly beyond what most gear and human hearing can handle). 48kHz (the video standard) can reproduce slightly higher frequencies -- but when recording for an audio target, that gain is usually lost - and then some - by downsampling to the target rate (which is why you'll find that around 80% of professional engineers simply use 44.1kHz for recording audio projects and 48kHz for video -- I won't get into "high res" at this point).
The bit depth (a.k.a. word length, never to be confused with "bit rate" which is entirely unrelated) dictates the resolution and "accuracy" of the sample - The dynamic range.
16-bit (as on a CD) allows for approximately 65,000 possible points of resolution from the bottom (-96dBFS) to the top (-0dBFS) of the scale.
24-bit word lengths allow for approximately 16.7 million possible points along with 48dB more downward headroom -- with a theoretical dynamic range of -144dBFS. FAR beyond ANY piece of gear you'll ever encounter.
The important part is that resolution... 24-bit tracking, calculations probably in 32 bit (or more), etc., etc. until the final delivery format (assuming that will be a 16-bit CD).
[EDIT] This thread got busy all of a sudden!
As far as headroom is concerned, 24-bit allows for far more downward headroom -- You can be much more gentle on the front end without the worry of the noise floor. "In theory" a signal at -0.0dBFS in 16-bit would be identical to the same signal in 24-bit at -48dBFS.
So you can *keep* more headroom *and* ridiculously high resolution - It's a good thing.
It's important to know that raising your word length does NOT allow for higher levels -- as some might say... "All 1's" is "all 1's" no matter what the sample rate or bit depth is. The fact that the dynamic range is so much wider (the additional 48dB in 24-bit) and the resolution is so much higher (256 times that of 16-bit) allows you to record lower level signals with better resolution.
And 90% of the time, lower (within reason) is better (no matter what the "hot as you can without clipping" paragraphs in the gear manuals say).
Fine-tuning controls for sample rates... You probably won't ever need them.
Noize2uCzar of MidiAdministrator
Aug 15, 2008 09:55 pm The sample rate itself is the nuber of sampler per second that are recorded into the wave form. So the higher the sample rate the more room you have to work with and the finer editing you can do.
Since: Apr 04, 2002
So at 48 kHz you will have 48000 samples per second to work with. at 96 kHz you will have 96000 samples per second to work with. It makes for less degrading of the audio when using DSP processing or effects if you prefer.
As I stated above, most work is done at 48 kHz 24 bit for normal recording. I will however bump it up to 96 kHz if I know I am goind to be doing a lot of fine editing or using a large bit of effects on a track. Sonar 7 allows for mixed sample rates in my tracks so if I know I am going to be twisting a guitar track or such with a lot of effects I will bump only that track up to allow for less quality loss when editing.
And yes, some will argue there is no loss in the digital world. I can attest to the fact that some pluggins will degrade the audio a little each time you process the track.
But in the end if it is something you are not going to be adding a lot of effects to then there is no need to eat up hard drive space to only dumb the file down to 16 bit 44.1 kHz in the end to burn it to CD.
Nice description on the bit rate Massive.
If you would like to participate in the forum discussions, feel free to register for your free membership.