Member
Join Date: Jun 2006
|
What does "sample rate" mean? What's the difference between 44.1kHz and 48.0kHz? The music sold on iTune Music Store are all in 44.1KHz.
What does it mean, and how does it affect the quality of sound? |
quote |
Mac Mini Maniac
Join Date: Sep 2005
|
It's the number of times per second that the sound level is sampled.
normally, you have 16-bit 44.1kHz sound. Pros use 24-bit 48kHz sound (moden equipment also does 96 and 192kHz). The bits is the amplitude of the signal, which is measured 44.1 or 48 thousand times per second. In normal 16-bit audio, you can have 2^16 (65536) different amplitude levels. The pros need more precision, so they use 24 bits to get 16.7 million different sound pressure levels. Note that we're not talking frequencey here. The microfone has no concept of frequency, only how much force is pushing on it at any given time. The result of measuring that pressure (with 16 or 24 bits) every 44100 or 48000 times per second does not make frequency either, only the data put together creates frequency as the signal increases and decreases in amplitude. Frequency is simply a function of the change in amplitude over time. Whoops. This might be a bit over the mark. Hope it helps at least. What you should focus on is not the sample rate, since it's the same as CDs. What you should worry about is the bitrate used for encoding the file. iTunes use 128kbit AAC which is less than CD quality and you're therefore buying an inferior product as compared to a CD. Converted 07/2005. |
quote |
Veteran Member
Join Date: Oct 2004
Location: Leiden, the Netherlands
|
Sample rate has to do with the frequency with which an original (analog) sounds source has been converted to a digital stream. The higher this frequence, the higher the highest pitch which can be reproduced reliably once the datastream is converted back to sound.
Because normal audio CD's use 44.1 KHz, most music on line uses that frequency as well. The difference between 44.1 and 48 KHz is small enough that I suspect nobody will be able to hear the difference. |
quote |
Veteran Member
Join Date: May 2005
Location: Paris, France
|
Yonzie and julesstoop have this nicely covered, but one thing you should note too is that resampling at a higher sample rate than the original will not magically improve the quality: in fact, due to complicated reasons relating to a very annoying gentleman called Nyquist, it may well cause a deterioration in the sound quality, especially if the result is compressed at a given bitrate (say, 128 kbps), because more data must be squeezed into a file of the same size. In other words, don't rip CDs (44.1 kHz) at 48 kHz.
Here's the first Google image result I found, which might help you visualise this better: http://www.innova-organs.com/images/SampleRate.jpg In the graph, time is on the x-axis with amplitude on the y-axis. … engrossed in such factional acts as dreaming different dreams. Last edited by Dorian Gray : 2006-11-06 at 14:16. |
quote |
*AD SPACE FOR SALE*
Join Date: Apr 2005
Location: Cleveland-ish, OH
|
I know the cd we just finished recording at the studio is done in 48kHz 24 bit and the kbps is something crazy like 2304 or some shit. I was blown away when I saw that.
Die young and save yourself.... @yontsey |
quote |
Veteran Member
Join Date: Oct 2004
Location: Leiden, the Netherlands
|
That's uncompressed audio for you.
A normal 16 bit, 44.1 KHz CD is 1411 Kbps. |
quote |
Member
Join Date: Jun 2006
|
What about the 16-bit and 24-bit? There is no such option to choose from in iTune.
So, normal CDs are recorded in 16-bit 44.1kHz, and I should rip those CDs in 44.1KHz as well, am I correct? When should I use 48KHz? |
quote |
Veteran Member
Join Date: Oct 2004
Location: Leiden, the Netherlands
|
Although most modern Mac have got a ADC (analog to digital converter) behind their line in (audio input) that allows sampling at higher bitrates (than 16 bits), it's obviously still senseless to rip a CD on a higher quality. A nomal audio CD is 16 bits audio, sampling it in 24 bits will not make it sound better. Besides that it would take 50% extra bandwith resulting in larger files sounding no better.
Audio sampled on 48 KHz is to be found on DAT-recorders and also often used with DV (the video standard, that is). |
quote |
BANNED
I am worthless beyond hope. Join Date: Jul 2004
Location: Washington, DC
|
|
quote |
Less than Stellar Member
|
|
quote |
Veteran Member
Join Date: Oct 2004
Location: Leiden, the Netherlands
|
Quote:
44,100 (Hz) * 16 (bits) * 2(channels) = 1,411.2 (kbps) So it doesn't even use RLE. This is completely logical by the way, since audio CD's historically simply contain a linear data stream. The player just reads the data in a linear fashion, very much like the analog media that predated it. Unless you count the moment of analog to digital conversion as a compression moment, of course. Which is technically justified, but (most of) the actual compression in that stage takes place before conversion to a digital signal. |
|
quote |
BANNED
I am worthless beyond hope. Join Date: Jul 2004
Location: Washington, DC
|
Quote:
Each wave is still represented in steps rather than a smooth single line. Those steps are really, really small, a tenth of the coarseness of mp3 at 140-ish kbps, but they're still a stepped approximation of sound. |
|
quote |
Student extraordinaire
Join Date: May 2004
Location: Canberra, Australia
|
BarracksSi, I think you're confusing compression with fidelity. CD audio is not compressed.
|
quote |
BANNED
I am worthless beyond hope. Join Date: Jul 2004
Location: Washington, DC
|
Alright, I'll concede that CD audio doesn't have weird mathematical crap going on behind the scenes. But still, it's not a truly accurate representation of sound. "Zero" is quieter than background magnetic noise of tape or a smooth record groove, and a CD's dynamic range is only dictated by the numbers represented by each bit and not physical limitations, but it's still not the real thing.
1411 kilobits per second is "compressed" when compared to analog. It's less accurate than 2304 kilobits per second. But, it's finer than what a human can reasonably hear, and it's still slow enough to have been encoded by processors a quarter-century ago. Mp3's and such do some other weird stuff to make 120-plus kbps sound like a reasonable facsimile. People say that CD audio is "uncompressed", but I don't see it that way. It's still a chopped-up representation of the original sound. Mp3 and AAC are merely further re-compressions of CD audio. |
quote |
Veteran Member
Join Date: Oct 2004
Location: Leiden, the Netherlands
|
Because you can only call something "compressed" if you measure it's true dynamic range first: compression is always relative.
So if the original multitrack recording was done with a higher quality, the mixed and mastered CD could be regarded as compressed compared to the original (and this is of course generally spoken the fact). Most of the time though when talking about compression with regard to computers we mean relatively computationally intensive mathematical tricks to make something (much) smaller without losing (most of) the original fidelity. |
quote |
OK Mr. Sunshine!
Join Date: Oct 2006
Location: Toronto
|
Short answer: If your source is 44.1k/16bit, there is no point ever converting it to anything higher.
Here's the explanation that I was given in my recording engineering course: When converting an analogue signal (i.e. microphone input) into a digital recording, the Digital to Analogue Converter takes snapshots of the sound wave. The sample rate (i.e. 44.1k) is the number of pictures it takes, the bit rate (i.e. 16 bit) is the number of bits they use to describe each snapshot Think of it this way: If you have a person posing for a portrait and you take a shot of them with a camera that can only replicate 256 colours, you're going to get a grainy picture. If you then take that picture into photoshop and convert it to millions of colours, it's not going to look any nicer - photoshop can't add detail that's not present in the original - you'll just have a much larger file size. The same rule applies to audio. If you have a source that's 44.1k/16 bit, there is no reason to ever convert it to anything higher, unless you need to fill up your HD for some reason. Once you convert a file into an mp3, there's no point converting it back to CD quality (unless you want to burn it to an audio CD for CD player listening - I still use a portable CD player) - you won't reproduce the original fidelity. So, when you buy a 128kbs AAC file from iTunes, you will never be able to get that file to sound like the CD. That being said, unless you're on a decent audio system, you probably won't be able to notice the difference. Once you're on a high quality audio system, you can hear the limitations of 16 bit. It doesn't sound "live". If you put a 24 bit recording onto that high end system, it sounds pretty frickin' sweet, pretty much like a live performance to my ears (though nothing will ever beat actually being there...). I hope that helps explain it . |
quote |
Member
Join Date: Jun 2006
|
Quote:
If the CD is 24-bit, is it possible to rip it as 24-bit lossless on iTune? I don' see any option to choose 16-bit or 24-bit in iTune. |
|
quote |
Veteran Member
Join Date: May 2004
Location: Promise Land of Trustafarians
|
Red Book, the standard for audio CDs, specifies that audio be 16-bit. If there is a 24-bit CD out there (doubtful), it isn't *technically* a CD and would probably not be recognized by your computer as an audio CD but rather data. I don't even think it would work in any standalone CD player.
edit: FYI, it's iTunes, not iTune. |
quote |
Member
Join Date: May 2005
Location: Australia
|
Normal CD's are 16-bit, however....
SACD's (super audio cd's) are 24-bit, and often come in 5.1 as well. I only know one person that actually uses them, and it's a very limited market. You'll also need a device that can support them. I would never bother because most music that I listen to (or most people listen to) isn't supported. Interesting though. http://en.wikipedia.org/wiki/Super_Audio_CD |
quote |
Veteran Member
Join Date: May 2004
Location: Promise Land of Trustafarians
|
DVD-Audio is a similar format. I have The Eagle's Hotel California on DVD-A and it sounds *awesome*. Tchaikovsky's 1812 Overture is also very well done. If you're in the market for a DVD player and you have a good speaker setup (not something that all came in one box from Best Buy) I'd definitely pick one up that supports hi-fi formats. Even budget players like the Oppo 970 can handle them now.
|
quote |
BANNED
I am worthless beyond hope. Join Date: Jul 2004
Location: Washington, DC
|
Quote:
A scratchy LP recorded onto SACD sounds like... a scratchy LP recorded onto SACD. "Downward-fidelity" conversions (CD to AAC/MP3, CD to cassette tape, 24-bit to 16-bit, whatever) always end up requiring effort to not lose too much sound quality. I used to record CDs onto tape to play in my car, and after a while, they started to sound pretty darned close, at least until they got old. |
|
quote |
careful with axes
Join Date: May 2004
Location: Hillsborough, CA
|
Quote:
|
|
quote |
careful with axes
Join Date: May 2004
Location: Hillsborough, CA
|
Compression implies an attempt to reduce in size...In this case CD audio is the pristine original. It cannot be compared to analog. That would be like implying RAW digital photograph is compressed just because there a finite number of pixels in it.
|
quote |
Mac Mini Maniac
Join Date: Sep 2005
|
Quote:
As for "Orthophonic", it harks back from the 20's where it was used to label some new hi-fi technology (serious milestone here), but now it's quite irrelevant and not even in my dictionary... It seems to me they use it to make it seem fancy when it isn't. As for "A new high fidelity recording", they just say that it's been recorded in high fidelity, they make no claims as to how little (or rather, much) it has been compressed afterwards to make it as hot as possible. What you got is a bag of hot air, sorry. Converted 07/2005. |
|
quote |
Member
Join Date: Jul 2006
Location: Manchester, England.
|
Good thread.
In Logic Pro there are options to record from 44khz right up to 192khz...read the Nyquist theory as to why it can end up a little silly. I used to master onto DAT at 48khz and that's pretty high enough - you can 'percieve' the difference in quality over CD but it's very slight - it's almost a high-end polish. I don't really have such a top notch audio system to appreciate anything higher...waste of memory to record so high also (unless I'm suddenly recording music for the local Halle orchestra here in Manchester). However, if there's a niche for music for dogs, well that's the time to grab a hold of the higher sampling rates! Here's a good article on the Nyquist theory if anyone's interested: http://www.lavryengineering.com/docu...ing_Theory.pdf |
quote |
Member
Join Date: May 2005
Location: Australia
|
|
quote |
careful with axes
Join Date: May 2004
Location: Hillsborough, CA
|
I hear researchers are actually trying to apply this technique to image capture too. Basically you take a single photodiode plus micro-mirrors and scan a landscape at ridiculous speeds to recreate a still scene.
|
quote |
BANNED
I am worthless beyond hope. Join Date: Jul 2004
Location: Washington, DC
|
Quote:
The phrase "pristine original" doesn't consider that even a portable DAT can record better than CD with a flip of a switch. |
|
quote |
careful with axes
Join Date: May 2004
Location: Hillsborough, CA
|
Quote:
It's pristine as in it was set to pick-up X amount of information. Even a digital recording sampled a 1Hz is pristine. Sound is not augmented, it is omitted. Omission and compression are different things. |
|
quote |
BANNED
I am worthless beyond hope. Join Date: Jul 2004
Location: Washington, DC
|
|
quote |
Posting Rules | Navigation |
|
Thread Tools | |