On 1 May 2000, Stainless Steel Rat wrote:
> * <[EMAIL PROTECTED]> on Mon, 01 May 2000
> | Actually, you will find that digital cellular systems virtually
> | universally use lossy compression for the voice. GSM for sure does, and
> | most others I have seen do.
>
> Actually, no, they don't. This comes straight from a guy who co-wrote the
> Sprint PCS software (he is now my co-sysmonster). The FCC does not allow
You'd better talk to him some more; the information you've given so far is
blatantly incorrect. Perhaps you've misunderstood him?
> compression or encryption; all civilian aerial broadcasts must be clear.
That is absolutely false. IEEE 802.11, aka WaveLAN, aka Orinoco is a
license free 2.4GHz wireless networking system. It is licensed and
approved and used (widely) in the United States. It includes either 40
bit or 128 bit encryption depending on the version you use.
Given that I am currently writing a driver for WaveLAN cards, I do
understand the standard somewhat. When I finish the driver, I will
understand the standard much more :)
http://www.qualcomm.com/cda/technology/display/0,1595,,00.html
Qualcomm makes phones in the US. Their phones are fully encrypted. And
fully FCC approved.
As a licensed radio amateur, I can assure you that I am fully allowed to
send compressed data. Amateur bands are explicitly prohibited from using
encryption, but that is not a limitation of all bands. The specific
limitation on the amateur band is that I have to use a publicly defined
encoding scheme.
http://www.qualcomm.com/cdmatechnologies/products/documents/CDMADataBook.pdf
Page 5 contains a description of the "variable rate vocoder". It's a
rather un-technical description, but it clearly explains the fact that
CDMA, a cellular standard used and approved in the US, uses lossy
compression like ATRAC.
> The lossy compression you describe is achieved by ratcheting back the
> sampling frequency rate to concentrate on the baritone and tennor ranges,
> ignoring the bass and alto ranges entirely.
The sampling rate has nothing to do with cutting off the low end of the
audio spectrum. The lossy compression I was referring to was that used in
GSM, the phone system used in pretty much everywhere but the US. GSM uses
a perceptual acoustic lossy compression system very similar to ATRAC. I
know this for a fact, having read through source code that did GSM
compression.
"ratcheting back"? Back from _what_? From infinite?
Your suggestion that sampling rate has anything to do with the low end
response of a system belies your lack of understanding. You could encode
a 0.01 Hz signal digitally. The analog circuitry of your system would be
unlikely to play it, but the sampling rate would have nothing to do with
bottom end response, just top end.
> It is the same thing CD-DA sampling does.
No, it's not.
Efficiency is _not_ compression. Is microfiche a "compressed" form of a
book? It's smaller, yes, but it is not compression. It's merely denser.
Please, before you continue to make assertions, do a bit of research and
cite some sources for your claims?
gopi.
-computer science major, radio ham, random electronics geek, etc.
-----------------------------------------------------------------
To stop getting this list send a message containing just the word
"unsubscribe" to [EMAIL PROTECTED]
