|
|
Posted by Randy Yates on 11/29/07 17:37
nospam@nospam.com (Don Pearce) writes:
> On Thu, 29 Nov 2007 11:50:28 -0500, Randy Yates <yates@ieee.org>
> wrote:
>
>>nospam@nospam.com (Don Pearce) writes:
>>
>>> On Thu, 29 Nov 2007 11:42:21 -0500, Randy Yates <yates@ieee.org>
>>> wrote:
>>>
>>>>nospam@nospam.com (Don Pearce) writes:
>>>>
>>>>> On Thu, 29 Nov 2007 13:09:22 +0000, Eeyore
>>>>> <rabbitsfriendsandrelations@hotmail.com> wrote:
>>>>>
>>>>>>
>>>>>>
>>>>>>Randy Yates wrote:
>>>>>>
>>>>>>> Eeyore <rabbitsfriendsandrelations@hotmail.com> writes:
>>>>>>> > [...]
>>>>>>> > ALL audio compression schemes rely on 'throwing away' information to get the
>>>>>>> > desired result.
>>>>>>>
>>>>>>> I suppose you meant to say "ALL lossy audio compression schemes ...".
>>>>>>
>>>>>>Fair enough.
>>>>>>
>>>>>>How much data compression can the non-lossy ones deliver ? I've never investigated. I imagine
>>>>>>it can't be that much.
>>>>>>
>>>>>>Graham
>>>>>
>>>>> I don't think the non-lossy ones are strictly codecs - just data
>>>>> compression and restoration systems.
>>>>
>>>>Lossless data compression is formally a type of "source coding," so
>>>>codec (meaning "coder/decoder") is a perfectly accurate term for the
>>>>process.
>>>>
>>>>A/D conversion is a type of quantization, which also falls under the
>>>>classification of source coding, so the application of codec is accurate
>>>>in this sense as well.
>>>
>>> Sure, I know all that; but that is kind of against the spirit of the
>>> word.
>>
>>How so?
>
> Because, as Graham has pointed out, under that terminology a Zip file
> would be a codec, and that isn't really what codecs are all about.
A zip file is a file. It isn't a codec any more than an mp3 file is a
codec.
However, the Lempel-Ziv algorithm is indeed a source coding algorithm
and in that sense can be called a codec.
--
% Randy Yates % "Maybe one day I'll feel her cold embrace,
%% Fuquay-Varina, NC % and kiss her interface,
%%% 919-577-9882 % til then, I'll leave her alone."
%%%% <yates@ieee.org> % 'Yours Truly, 2095', *Time*, ELO
http://www.digitalsignallabs.com
[Back to original message]
|