Talk:mp3PRO

Latest comment: 9 years ago by Bp0 in topic Class B, importance Low

reverted a change by "MrOrlie"

edit

2011/Nov/08: Whoever MrOrlie is, he was wrong to remove the Linux section. The link I added to my Linux/mp3pro page is not spam, it genuinely shows how to deal with mp3pro files under Linux. And since there are no adds there, the classification as "spam" is unwarranted.

possible research

edit

It has struck me that it might be interesting information if someone would research a bit more of how it actually works (not anything patented obviously.) It has occured to me for quite some time that MP3Pro may very well just use the rather simplistic method of simply dropping half of the samples. Normally digital audio is sampled twice (eg 44100Hz audio is actualy 22050Hz sound sampled twice) for a number of reasons. However, if one dropped half of the samples (or even just merged two into one) and had the decompressor automatically double them, the sound might not sound so bad at half of the bitrate (while it would not be equal to true MP3 at the normal bitrate, it would surely exceed normal MP3 at the same bitrate when in lower bitrates.) This is just a theory, but if it were true, I'm not certain this method actually could even be patented, and there's no reason that more CODECs couldn't implement their own version of MP3Pro... (Of course it may be complete bunk, but that's why I wonder if anyone has ever researched the matter.) This theory might also explain why MP3Pro files play just fine in a normal MP3 CODEC except at half the sample rate...

Such a thing would make a great addition to the MP3Pro page. —Preceding unsigned comment added by 68.209.196.233 (talk) 09:00, 26 October 2007 (UTC)Reply


First, digital audio is not "sampled twice", it is sampled at a frequency which is at least twice as high as the highest frequency of the signal.
Second, merely dropping samples and then duplicating them is a very crude form of resampling, without any anti-aliasing. Higher quality could be obtained by using proper resampling algorithms that include low-pass filters before performing the encoding. However, if the signal were to be resampled to 22050 Hz, as suggested here, the bandwith would be severely limited and it would greatly impact audibility. Anything less than 32 kHz is only used in ultra-low bitrate, speech-only coding.
Third, many encoders already include customizable low pass filters.
Fourth, and most important, resampling and Spectral Band Replication are not the same. 69.165.144.30 (talk) 23:51, 19 December 2009 (UTC)Reply

additions

edit

added the truthful reasoning behind the lack of a linux player. bite my ass, rca. --andrew 06:15, 6 July 2006 (UTC)Reply

Native Linux players

edit

AFAIK at least with mpg123 (and probably other mp3 command-line players) one can manually specify target samplerate. Not a big problem, really. 217.76.116.171 00:59, 11 August 2007 (UTC)Reply

December 2014 rewrite, history of a dead codec

edit

I rewrote this article to reflect that it is unmaintained and largely considered obsolete. It is now just a short history of its development and existence. I can't tell when mp3PRO was officially dropped but there are references of it being dead as early as 2005.

Most of the products and software on the mp3pro page are dead links, but two interesting ones:

SAM Broadcaster
Still proudly supports mp3PRO in 2014!
JetAudio Changelog
Removed mp3PRO in version 8.1.1.

--Bp0 (talk) 03:12, 27 December 2014 (UTC)Reply

Class B, importance Low

edit

Reasonably complete description of the existance of the thing. Its not very important, but it is interesting. I think the page should exist if only to inform people that mp3pro is not a successor or improvement to mp3, but only a now unmaintained and disused variant of it. --Bp0 (talk) 03:48, 27 December 2014 (UTC)Reply