Digital audio stream terms?

Status
Not open for further replies.

DaveC

Distinguished
May 8, 2004
204
0
18,680
Archived from groups: rec.audio.tech,rec.audio.misc,rec.audio.pro,uk.rec.audio (More info?)

I'm listening to a stream via WMP. The Get Info box says:

Bit Rate: 24 Kbps
Audio Codec: Windows Media Audio 9
20 kbps, 32kHz, mono 1-pass CBR

Can someone please clarify what the bit rate means and what the audio codec
numbers (20 kbps & 32 KHz) mean?

Thanks,
--
DaveC
me@privacy.net
This is an invalid return address
Please reply in the news group
 
G

Guest

Guest
Archived from groups: rec.audio.tech,rec.audio.misc,rec.audio.pro,uk.rec.audio (More info?)

"DaveC" <me@privacy.net> wrote in message
news:0001HW.BCD28C33003D04CAF04885B0@news.individual.net...
> I'm listening to a stream via WMP. The Get Info box says:
>
> Bit Rate: 24 Kbps
> Audio Codec: Windows Media Audio 9
> 20 kbps, 32kHz, mono 1-pass CBR
>
> Can someone please clarify what the bit rate means and what the audio
codec
> numbers (20 kbps & 32 KHz) mean?

I believe they mean...

20 kbps = 20,000 (or 20480?) bits per second of data
through the network into your computer for this stream.

32kHz = the (original?) sample rate implying absolute
maximum 16KHz high frequency limit (likely lower).

CBR = constant (vs. variable/dynamic) bit-rate.
 

DaveC

Distinguished
May 8, 2004
204
0
18,680
Archived from groups: rec.audio.tech,rec.audio.misc,rec.audio.pro,uk.rec.audio (More info?)

On Thu, 20 May 2004 20:26:29 -0700, Richard Crowley wrote
(in article <10aqtn512ccu029@corp.supernews.com>):

>
> "DaveC" <me@privacy.net> wrote in message
> news:0001HW.BCD28C33003D04CAF04885B0@news.individual.net...
>> I'm listening to a stream via WMP. The Get Info box says:
>>
>> Bit Rate: 24 Kbps
>> Audio Codec: Windows Media Audio 9
>> 20 kbps, 32kHz, mono 1-pass CBR
>>
>> Can someone please clarify what the bit rate means and what the audio
>> codec numbers (20 kbps & 32 KHz) mean?

> I believe they mean...
>
> 20 kbps = 20,000 (or 20480?) bits per second of data
> through the network into your computer for this stream.

My guess would be that the 24 Kbps is the network stream speed...

> 32kHz = the (original?) sample rate implying absolute
> maximum 16KHz high frequency limit (likely lower).
>
> CBR = constant (vs. variable/dynamic) bit-rate.

Anyone else?
--
DaveC
me@privacy.net
This is an invalid return address
Please reply in the news group
 

Stimpy

Distinguished
Jan 15, 2001
138
0
18,680
Archived from groups: rec.audio.tech,rec.audio.misc,rec.audio.pro,uk.rec.audio (More info?)

DaveC wrote:
> On Thu, 20 May 2004 20:26:29 -0700, Richard Crowley wrote
> (in article <10aqtn512ccu029@corp.supernews.com>):
>
>>
>> "DaveC" <me@privacy.net> wrote in message
>> news:0001HW.BCD28C33003D04CAF04885B0@news.individual.net...
>>> I'm listening to a stream via WMP. The Get Info box says:
>>>
>>> Bit Rate: 24 Kbps
>>> Audio Codec: Windows Media Audio 9
>>> 20 kbps, 32kHz, mono 1-pass CBR
>>>
>>> Can someone please clarify what the bit rate means and what the
>>> audio codec numbers (20 kbps & 32 KHz) mean?
>
>> I believe they mean...
>>
>> 20 kbps = 20,000 (or 20480?) bits per second of data
>> through the network into your computer for this stream.
>
> My guess would be that the 24 Kbps is the network stream speed...
>
>> 32kHz = the (original?) sample rate implying absolute
>> maximum 16KHz high frequency limit (likely lower).
>>
>> CBR = constant (vs. variable/dynamic) bit-rate.
>
> Anyone else?

Both definitions are correct...

With streaming audio, the kbps figure represents the number of Kilo Bits Per
Second being streamed down into your computer. The kHz figure represents
the sample rate, i.e. the number of samples per second when the source was
sampled.

CBR = Constant bit rate. That is to say the sample/rip was taken at a fixed
kbps value. Some encoders can calculate the 'best' (a-hem) bit rate on the
fly, this is known as VBR (Variable Bit Rate).
 

DaveC

Distinguished
May 8, 2004
204
0
18,680
Archived from groups: rec.audio.tech,rec.audio.misc,rec.audio.pro,uk.rec.audio (More info?)

On Fri, 21 May 2004 10:44:41 -0700, Stimpy wrote
(in article <2h6tidF9nduqU1@uni-berlin.de>):

>>>> Bit Rate: 24 Kbps
>>>> Audio Codec: Windows Media Audio 9
>>>> 20 kbps, 32kHz, mono 1-pass CBR
>>>>
>>>> Can someone please clarify what the bit rate means and what the
>>>> audio codec numbers (20 kbps & 32 KHz) mean?

> With streaming audio, the kbps figure represents the number of Kilo Bits Per
> Second being streamed down into your computer. The kHz figure represents
> the sample rate, i.e. the number of samples per second when the source was
> sampled.
>
> CBR = Constant bit rate. That is to say the sample/rip was taken at a fixed
> kbps value. Some encoders can calculate the 'best' (a-hem) bit rate on the
> fly, this is known as VBR (Variable Bit Rate).

So the 24 Kbps is how fast it's being delivered over the 'net;
32 KHz is the sample rate it was digitized at the source;
and 20 Kbps is ... hmm, I'm getting a bit lost here.

An audio CD is digitized at 44.1 KHz, but there's no Kbps rating associated
with the digitizing, that I'm aware of.

Clarification?

Thanks,
--
DaveC
me@privacy.net
This is an invalid return address
Please reply in the news group
 
G

Guest

Guest
Archived from groups: rec.audio.tech,rec.audio.misc,rec.audio.pro,uk.rec.audio (More info?)

DaveC wrote:

> An audio CD is digitized at 44.1 KHz, but there's no Kbps rating
> associated with the digitizing, that I'm aware of.

16 bits per sample, two seperate channels no compression therefore
1411.2Kbps. The data rate from the disk though is higher as each byte is
encoded as 14 bits for resilience, so 2469.6Kbps.

You did ask.
 
G

Guest

Guest
Archived from groups: rec.audio.tech,rec.audio.misc,rec.audio.pro,uk.rec.audio (More info?)

DaveC wrote:

> On Fri, 21 May 2004 10:44:41 -0700, Stimpy wrote
> (in article <2h6tidF9nduqU1@uni-berlin.de>):
>
>
>>>>>Bit Rate: 24 Kbps
>>>>>Audio Codec: Windows Media Audio 9
>>>>>20 kbps, 32kHz, mono 1-pass CBR
>>>>>
>>>>>Can someone please clarify what the bit rate means and what the
>>>>>audio codec numbers (20 kbps & 32 KHz) mean?
>
>
>>With streaming audio, the kbps figure represents the number of Kilo Bits Per
>>Second being streamed down into your computer. The kHz figure represents
>>the sample rate, i.e. the number of samples per second when the source was
>>sampled.
>>
>>CBR = Constant bit rate. That is to say the sample/rip was taken at a fixed
>>kbps value. Some encoders can calculate the 'best' (a-hem) bit rate on the
>>fly, this is known as VBR (Variable Bit Rate).
>
>
> So the 24 Kbps is how fast it's being delivered over the 'net;
> 32 KHz is the sample rate it was digitized at the source;
> and 20 Kbps is ... hmm, I'm getting a bit lost here.
>

I think 20 kpbs is the encodde rate, and 24 is the
delivery/streaming rate - 4 kbps of overhead.

I *think*.

> An audio CD is digitized at 44.1 KHz, but there's no Kbps rating associated
> with the digitizing, that I'm aware of.

Sure there is. It's 1.44 M(bit)ps or something. It shows
up in Winamp when you play back 44.1 .wav files.

>
> Clarification?
>
> Thanks,


--
--
Les Cargill
 

DaveC

Distinguished
May 8, 2004
204
0
18,680
Archived from groups: rec.audio.tech,rec.audio.misc,rec.audio.pro,uk.rec.audio (More info?)

On Fri, 21 May 2004 18:52:15 -0700, Les Cargill wrote
(in article <jnyrc.42664$hH.829843@bgtnsc04-news.ops.worldnet.att.net>):

>> An audio CD is digitized at 44.1 KHz, but there's no Kbps rating
>> associated
>> with the digitizing, that I'm aware of.

> Sure there is. It's 1.44 M(bit)ps or something. It shows
> up in Winamp when you play back 44.1 .wav files.

So it's 44.1 x 8 (or whatever a byte is) x 2 (stereo) + overhead +
errorchecking = encoding kbps?
--
DaveC
me@privacy.net
This is an invalid return address
Please reply in the news group
 
G

Guest

Guest
Archived from groups: rec.audio.tech,rec.audio.misc,rec.audio.pro,uk.rec.audio (More info?)

DaveC wrote:

> On Fri, 21 May 2004 18:52:15 -0700, Les Cargill wrote
> (in article <jnyrc.42664$hH.829843@bgtnsc04-news.ops.worldnet.att.net>):
>
>
>>>An audio CD is digitized at 44.1 KHz, but there's no Kbps rating
>>>associated
>>>with the digitizing, that I'm aware of.
>
>
>>Sure there is. It's 1.44 M(bit)ps or something. It shows
>>up in Winamp when you play back 44.1 .wav files.
>
>
> So it's 44.1 x 8 (or whatever a byte is) x 2 (stereo) + overhead +
> errorchecking = encoding kbps?

44.1 kilo-samples per second x 1000 kHz/Hz * 8 bits/byte * 2 bytes/sample
* 2 channels + overhead = encoding bits/second.

So for CD-quality, it's 1,411,200 bits/second + overhead, which
could be in the neighborhood of 1.44 megabits/second if you
only have a few percent of overhead (which is feasible in
some cases).

- Logan
 

DaveC

Distinguished
May 8, 2004
204
0
18,680
Archived from groups: rec.audio.tech,rec.audio.misc,rec.audio.pro,uk.rec.audio (More info?)

On Fri, 21 May 2004 19:59:35 -0700, Logan Shaw wrote
(in article <rmzrc.13188$9f6.2820@fe2.texas.rr.com>):

> 1000 kHz/Hz

I don't understand this term.

The digitizer samples the analog signal once every 1/44,100th of a second, at
a resolution of 8 bits. Where does "1000 kHz/Hz" come into play?

Willing to learn...
--
DaveC
me@privacy.net
This is an invalid return address
Please reply in the news group
 

DaveC

Distinguished
May 8, 2004
204
0
18,680
Archived from groups: rec.audio.tech,rec.audio.misc,rec.audio.pro,uk.rec.audio,comp.sys.mac.misc (More info?)

iTunes displays data about each audio file. This data includes "sample rate"
and "bit rate". (These are MP3 audio files, converted from CD.)

Sample rates include 44.1 and 22.05. These I understand. Bit rates include
128Kbps and 56Kbps. If these are not streamed sources, but just digitized
files, why is there a bit rate associated with them. It seems just a logical
to display a bit rate for a MS Word file...

Confused (still)...
--
DaveC
me@privacy.net
This is an invalid return address
Please reply in the news group
 
G

Guest

Guest
Archived from groups: rec.audio.tech,rec.audio.misc,rec.audio.pro,uk.rec.audio,comp.sys.mac.misc (More info?)

On Sun, 23 May 2004 09:14:58 -0700, DaveC wrote:

> iTunes displays data about each audio file. This data includes "sample rate"
> and "bit rate". (These are MP3 audio files, converted from CD.)
>
> Sample rates include 44.1 and 22.05. These I understand. Bit rates include
> 128Kbps and 56Kbps. If these are not streamed sources, but just digitized
> files, why is there a bit rate associated with them. It seems just a logical
> to display a bit rate for a MS Word file...

Bit rates for MP3s are a measure of the filesize/quality trade off.
The lower the bit-rate, the smaller the resulting file, and the lower the
quality.
Bit rate is independent of sample rate.
So, you can encode a 44.1k mp3 at 128kbps, or 256kbps or whatever.
The sample rate remains the same, but the fidelity is less.

They are not 'just digitised files', they are MP3s, which bear very little
relation to the original uncompressed CD data. The data MP3s contain is
more like a description of the sound, rather than an analog of it.
The description can be less exact (Less kbps) but still be recognisable.

> Confused (still)...
 

DaveC

Distinguished
May 8, 2004
204
0
18,680
Archived from groups: rec.audio.tech,rec.audio.misc,rec.audio.pro,uk.rec.audio,comp.sys.mac.misc (More info?)

On Sun, 23 May 2004 11:06:44 -0700, philicorda wrote
(in article <pan.2004.05.23.18.06.40.307716@nospnospamspaammntlworld.com>):

> Bit rate is independent of sample rate.
> So, you can encode a 44.1k mp3 at 128kbps, or 256kbps or whatever.
> The sample rate remains the same, but the fidelity is less.

So why not use the term "resample rate" rather than bit rate. The latter
implies streaming data, to me, at least. "Rate" implies a period of time over
which the "128K" -- or whatever -- takes place, when in fact it's just a
combination of the re-sample rate (ie, 22.05K) and the resultant file size.

Not just arguing terminology, here, but hoping that my premise is understood,
helping me better understand. (Understand?) :)

Thanks,
--
DaveC
me@privacy.net
This is an invalid return address
Please reply in the news group
 
G

Guest

Guest
Archived from groups: rec.audio.tech,rec.audio.misc,rec.audio.pro,uk.rec.audio,comp.sys.mac.misc (More info?)

DaveC wrote:
>
> On Sun, 23 May 2004 11:06:44 -0700, philicorda wrote
> (in article <pan.2004.05.23.18.06.40.307716@nospnospamspaammntlworld.com>):
>
> > Bit rate is independent of sample rate.
> > So, you can encode a 44.1k mp3 at 128kbps, or 256kbps or whatever.
> > The sample rate remains the same, but the fidelity is less.
>
> So why not use the term "resample rate" rather than bit rate. The latter
> implies streaming data, to me, at least. "Rate" implies a period of time over
> which the "128K" -- or whatever -- takes place, when in fact it's just a
> combination of the re-sample rate (ie, 22.05K) and the resultant file size.
>
> Not just arguing terminology, here, but hoping that my premise is understood,
> helping me better understand. (Understand?) :)


Actually, the terminology is important. I think your error is
using the term "resample". Those rates may be different due to
the sample size (in bits) that was chosen.

The data isn't resampled to stream, it's compressed into an MP3
format in order to send over the net.

The original recording is sampled at a fixed rate (samples per
second) which produces a fixed number of bits per second (i.e.,
based on the bits per sample). That means to play it back exactly
as it was recorded, you have to play it back at that specific
fixed number of bits per second in order to regenerate the sound
at the original sample rate. However, that is a LOT of data and
bits per second and not really feasible for streaming over the
internet (impossible if you have a dialup connection).

MP3 provides a way to compress the data by digitally rearranging
it and throwing away MANY bits of audio data in a way that the
sound is not impacted too much. The more you throw away, the less
data that you have to stream over a given time and hence the
ability to use a lower speed connection (e.g., like a 56kbps
dialup connection). However, more data thrown away means less
quality of the sound compared to the original. On a higher speed
connection you can use less compression and keep more of the
original data and hence a better quality sound.

- Jeff
 

DaveC

Distinguished
May 8, 2004
204
0
18,680
Archived from groups: rec.audio.tech,rec.audio.misc,rec.audio.pro,uk.rec.audio,comp.sys.mac.misc (More info?)

On Sun, 23 May 2004 19:06:36 -0700, Jeff Wiseman wrote
(in article <40B15825.E0F2B9B@earthlink.net>):

> Actually, the terminology is important. I think your error is
> using the term "resample". Those rates may be different due to
> the sample size (in bits) that was chosen.
>
> The data isn't resampled to stream, it's compressed into an MP3
> format in order to send over the net.

But this only applies to a *streamed* file, not a stationary
on-your-hard-disk file. Why does a file that's already on your drive have a
bit rate associated with it? A sample rate, yes. But it shouldn't have a bit
rate unless it's being streamed.

> The original recording is sampled at a fixed rate (samples per
> second) which produces a fixed number of bits per second (i.e.,
> based on the bits per sample). That means to play it back exactly
> as it was recorded, you have to play it back at that specific
> fixed number of bits per second in order to regenerate the sound
> at the original sample rate. However, that is a LOT of data and
> bits per second and not really feasible for streaming over the
> internet (impossible if you have a dialup connection).
>
> MP3 provides a way to compress the data by digitally rearranging
> it and throwing away MANY bits of audio data in a way that the
> sound is not impacted too much. The more you throw away, the less
> data that you have to stream over a given time and hence the
> ability to use a lower speed connection (e.g., like a 56kbps
> dialup connection). However, more data thrown away means less
> quality of the sound compared to the original. On a higher speed
> connection you can use less compression and keep more of the
> original data and hence a better quality sound.

I think I understand everything except why a file that is not streamed has a
bit rate spec. Digitized content should be described by the sample rate and,
in some cases, sample size. But only streamed content should be described by
a bit rate. Any other use of these terms is misleading.
--
DaveC
me@privacy.net
This is an invalid return address
Please reply in the news group
 
G

Guest

Guest
Archived from groups: rec.audio.tech,rec.audio.pro (More info?)

DaveC <me@privacy.net> wrote:
> I think I understand everything except why a file that is not streamed
> has a bit rate spec. Digitized content should be described by the
> sample rate and, in some cases, sample size. But only streamed content
> should be described by a bit rate. Any other use of these terms is
> misleading.

you're still streaming a file when you play it from a disk.

to playback uncompressed stereo 16bit/44.1kHz PCM, the disk (hard drive,
redbook CD, compactflash card, etc) needs a nominal sustained transfer
rate of 1411.2kbit/s (2*16*44.1).

likewise, when you play back a 128kbit/s MP3 from your local device, the
nominal sustained transfer rate from that device is 128kbit/s.

--
Aaron J. Grier | "Not your ordinary poofy goof." | agrier@poofygoof.com
"someday the industry will have throbbing frontal lobes and will be able
to write provably correct software. also, I want a pony." -- Zach Brown
 

DaveC

Distinguished
May 8, 2004
204
0
18,680
Archived from groups: rec.audio.tech,rec.audio.pro (More info?)

On Mon, 24 May 2004 00:08:54 -0700, Aaron J. Grier wrote
(in article <10b37s6f0i60597@corp.supernews.com>):

> you're still streaming a file when you play it from a disk.

True enough, apparently.

But misleading, nonetheless. "Bit rate" implies, traditionally, data
communications over a network or such. "Resolution" is a more-appropriate
term regarding files possessed on your local media.

To take a parallel example, you don't describe image files in "rate" terms.
They have been scanned at a set bits-per-inch resolution and are of a certain
file size. These two terms describe the image's "quality" but do not require
a description of how quickly (over time) the file will be displayed.

> to playback uncompressed stereo 16bit/44.1kHz PCM, the disk (hard drive,
> redbook CD, compactflash card, etc) needs a nominal sustained transfer
> rate of 1411.2kbit/s (2*16*44.1).

That's like telling the owner of a new car in the owner's manual that in
order to maintain a safe driving speed, he/she should depress the accelerator
pedal until the engine is operating at a sustained rate of 275.4 cfm (cubic
feet per minute). It is a technical description of the engine's mode of
operation that is totally inappropriate to discuss outside of engineering
circles. It leads to confusion in those to whom the term has no relevance.
More appropriately, an owner's manual typically discusses safe limitations in
terms that, generally, relate to the everyday use of the product ("at a safe
speed", "observe local speed laws", etc.). Only when the terminology is
necessary for a particular task (ie, towing) is such language appropriate
(weights, speeds, etc.).

> likewise, when you play back a 128kbit/s MP3 from your local device, the
> nominal sustained transfer rate from that device is 128kbit/s.

Understood. But an inappropriate terminology for general description to the
public.
--
DaveC
me@privacy.net
This is an invalid return address
Please reply in the news group
 
G

Guest

Guest
Archived from groups: rec.audio.tech,rec.audio.misc,rec.audio.pro,uk.rec.audio,comp.sys.mac.misc (More info?)

On Sun, 23 May 2004 21:23:18 -0700, DaveC wrote:


> I think I understand everything except why a file that is not streamed has a
> bit rate spec. Digitized content should be described by the sample rate and,
> in some cases, sample size. But only streamed content should be described by
> a bit rate. Any other use of these terms is misleading.

Just giving the sample rate would not tell you anything about the quality
of the encoding. The same sample rate can be mp3'd at a number of
different bit rates.

There is no difference to an MP3 decoder if the file is streamed from the
net or streamed from a hard drive.
The messy stuff about getting the data to the right place at the right
time is dealt with in another layer.
 
G

Guest

Guest
Archived from groups: rec.audio.tech,rec.audio.pro (More info?)

On Mon, 24 May 2004 06:23:38 -0700, DaveC wrote:

> On Mon, 24 May 2004 00:08:54 -0700, Aaron J. Grier wrote
> (in article <10b37s6f0i60597@corp.supernews.com>):
>
>> you're still streaming a file when you play it from a disk.
>
> True enough, apparently.
>
> But misleading, nonetheless. "Bit rate" implies, traditionally, data
> communications over a network or such. "Resolution" is a more-appropriate
> term regarding files possessed on your local media.

I guess 'resolution' would be clearer to some people... But to say an MP3
has a resolution of 128kbps would make it more confusing in the long run,
as the meaning is lost in the abstraction.

Sometimes it's not a good idea to try and make a parallel of a computing
term in what appears to be clearer language.

The 'Campaign for Clear English' came across this in the UK when they
tried to find a simpler term for 'modem'.
There really isn't one, as there is no 'real life' alternative explanation
that can be used.

>
> To take a parallel example, you don't describe image files in "rate"
> terms. They have been scanned at a set bits-per-inch resolution and are
> of a certain file size. These two terms describe the image's "quality"
> but do not require a description of how quickly (over time) the file
> will be displayed.

That's because static image files are not shown over a defined period
of time.

If it's a film (AVI,MPG etc) then the kbps or Mbps is used. Ie, 784 Kbps,
2.5Mbps.

There is no rule for how fast a static jpeg image whould be decoded, or
a zip file extracted, but if you are decoding an MP3 or AVI, then you had
better do it at the defined bit rate, or it would be
unwatchable/unlistenable!

>
>> to playback uncompressed stereo 16bit/44.1kHz PCM, the disk (hard drive,
>> redbook CD, compactflash card, etc) needs a nominal sustained transfer
>> rate of 1411.2kbit/s (2*16*44.1).
>
> That's like telling the owner of a new car in the owner's manual that in
> order to maintain a safe driving speed, he/she should depress the accelerator
> pedal until the engine is operating at a sustained rate of 275.4 cfm (cubic
> feet per minute). It is a technical description of the engine's mode of
> operation that is totally inappropriate to discuss outside of engineering
> circles. It leads to confusion in those to whom the term has no relevance.
> More appropriately, an owner's manual typically discusses safe limitations in
> terms that, generally, relate to the everyday use of the product ("at a safe
> speed", "observe local speed laws", etc.). Only when the terminology is
> necessary for a particular task (ie, towing) is such language appropriate
> (weights, speeds, etc.).

Again, you can create a simpler alternative, but the useful meaning is
lost.

>
>> likewise, when you play back a 128kbit/s MP3 from your local device, the
>> nominal sustained transfer rate from that device is 128kbit/s.
>
> Understood. But an inappropriate terminology for general description to the
> public.

What would you suggest?
 
Status
Not open for further replies.