Skip to main content
Topic: Why not to use WMA/mp3pro (Read 5075 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Why not to use WMA/mp3pro

Reply #1
Its not just those formats! RealMedia also has DRM.

This is my biggest headache with audio codecs at the moment - finding a good quality open streaming codec for 56k dialup use.

Ruairi
rc55.com - nothing going on

Why not to use WMA/mp3pro

Reply #2
The 'pro' part of the mp3pro algorithm - the SBR (Spectral Band Replication) will soon become open, as the part of ISO MPEG-4 audio standard extensions.

As I heard, the decoding part of SBR will be IPR free, but don't take my word for it, since I hard that from a source outside of MPEG commitee. It would be nice

Why not to use WMA/mp3pro

Reply #3
Peter made some great statements which hopefully get some mainstream audio users thinking correctly. His advise was head on. LAME, OGG, or MPC are the only reasonable formats. But I am going to focus on the WMA in Windows Media Player. He made a comment that the files have DRM compulsory. I don't think that is accurate. I have WMP 8.00.00.4477 in Windows XP Pro. There is a check box for "Protect Content." And if I recall accurately this box is unchecked by default. So no the files made are not DRM unless the user actively selects them to be.

rc55,
In response to your post. I don't think Real Audio has DRM any longer. I didn't see it as an option in the Real One Player trial I had a few months back. The files were very obvious unlike WMA. RMJ was unsecure and RMX was secure.

Bottom line is clarification that these formats do not force DRM, but allow the option. Not to be mistaken in my intent. I do not support either format. They are both "rogue" audio types in my opinion.

~ Darwyn

Why not to use WMA/mp3pro

Reply #4
I really like Peter's Job, but is this article really objective ?

First, there is mp3pro@128 kb/s (Steinberg MyMP3 3)
And I'm not sure that ogg vorbis beat wma and mp3pro at any bitrate (according to ff123 public test on 64 kb/s bitrate, ogg is not good at all on low BR). Maybe should Peter wait for RC4 before writing this...

Why not to use WMA/mp3pro

Reply #5
look. i am talking to masses. masses are dumb. they just eat any propaganda we write. if they can buy that 64kbps mp3pro is "cd quality", they will believe that 64kbps voirbs is "cd quality" as well. both mp3pro and wma are just a big pain in our techsupport's ass (details not worth mentioning here). vorbis quality isn't much worse (and will soon get better); vorbis is simply better because it's free and open. if you talk to masses, it's no use going into details (like "at this bitrate this is better but at that bitrate that is better") because people wouldn't understand you if you did so. at least this propaganda does less harm than M$'s or real's one.
did i say that DRM was compulsory ? where ? the problem is, people make DRM files without being aware of what they're doing, then watch all their music collection going poop after simple OS reinstall. again, this doc is for newbies, not for us geeks, 'cause we geeks know it all already.
yes, i'll possibly update it with info about realmedia someday, but not many people actually encode realmedia files so they aren't much trouble for us.
(note: edited about 128kbps mp3pro)

Why not to use WMA/mp3pro

Reply #6
Quote
Currently, Ogg Vorbis outperforms both WMA and mp3pro at any bitrate.


Source : http://205.188.228.81/showthread.php?threadid=81838


Is it true ? Not really.

Fighting propaganda by propaganda is not a good thing.
Many people don't believe me when I say that LAME is far better than Fraunhofer (Fastenc, Producer Pro...) : « you say that because you support open-source ; I just want quality, and Fraunhofer made with milions dollars very wonderful codecs...». Fatboy is helping me a lot to convice them.


Your article is good. But I'm not sure that lying is really a good thing. Better convince masses to forget low bitrate...

[sorry for my english]

Why not to use WMA/mp3pro

Reply #7
ff123's tests though good are outdated by almost 2 versions for Vorbis. They used the then just released RC2. We are at the tail end of RC3 nearing RC4 and 1.0. Lots of things have changed. RC3 and before have not been tuned for sub 80Kbps bitrates. And even then I found 64kbps Vorbis artifacts better than WMA or MP3-PRO to me. It had some problems, but nothing that a little tuning will not clean up for the most part.

When it comes to a music format quality is damned important. But so is portability. And I don't mean portable players. I mean jumping from on OS or platform to the next. That oblitterates WMA as a candidate. It severely reduces MP3-PRO's candidacy as well. That leaves us with MP3, MPC, Vorbis, LPC, APE, and a few others. Of those candidates MP3, MPC, and Vorbis best suit my quality/size ratio needs. And when it comes to supported formats OGG-Vorbis is popping out of the wood work. Under linux it is being integrated into some of the most popular window managers. Move on to BeOS and it has good support as well. MACs also play Vorbis. Windows again is flooded with support. BSD and Linux share almost everything. Hell I hear even QNX is dabbliing with Vorbis.

When it comes to newbies and non-techies Vorbis wins as well. WMA, MP3, MP3-PRO are the main  formats recognised by them. But Vorbis is quickly joining their ranks. Of these formats which would you rather they use. WMA which was purchased and is now being developed by a company with a track record and penchant  for lieing and writing inferior software. MP3 and MP3-PRO which were developed by many commercial audio experts acrost the board. But due to its old technology and kludged improvements still has fundamental drawbacks and flaws. Vorbis which is being written and improved daily by audio experts from many fields and even audio afficianados here as well. It is a bit at a disadvantage because they seek to avoid pattents. But this could also be one of it's greatest strengths. It's VBR methods almost always spend just the right amount of bitrate for the signal to provide for excelent quality. And it can be easily accessed by the -q modes. For them Vorbis should be the choice. When it comes to MP3 Lame is one off the best. But all the different switches confuse the lay person quite a bit. And everybody has got it in their head that 128Kbps is superb quality. With Vorbis we have a second chance to mold them. Bitrates are almost thrown out the window. I recomend -q 5 to all of em.


In the end vorbis is not the be all end all audio codec. Actually I have yet to see that codec. But it is definatly up there with the best of em.

Why not to use WMA/mp3pro

Reply #8
Quote
Originally posted by Neo Neko
ff123's tests though good are outdated by almost 2 versions for Vorbis. They used the then just released RC2. We are at the tail end of RC3 nearing RC4 and 1.0. Lots of things have changed. RC3 and before have not been tuned for sub 80Kbps bitrates. And even then I found 64kbps Vorbis artifacts better than WMA or MP3-PRO to me. It had some problems, but nothing that a little tuning will not clean up for the most part.


The listening tests referred to in this thread appear on the r3mix forum at:

http://66.96.216.160/cgi-bin/YaBB.pl?board...&num=1017521986

and

http://66.96.216.160/cgi-bin/YaBB.pl?board...&num=1017905544

Panos and I used RC3 for ogg at -q 0

ff123

Why not to use WMA/mp3pro

Reply #9
Overall I believe mp3pro is a bit better at 64kbps than Vorbis. Even Monty himself has said this.
Also CoolEdit 2 Pro's MP3Pro VBR is certainly an impromevent to Mp3pro cbr.
Right now as we speak, Monty is coding techniques which are supposed to improve low bitrates alot. We will see when it hits the CVS.
Juha Laaksonheimo

Why not to use WMA/mp3pro

Reply #10
Vorbis Model:

I think the Vorbis physcoacoustics algorithm can be tweaked quite well to work around low-bitstream audio coding, because it open-source there isn't anything that says you "cannot improve the physcoacoustics algorithm", I don't like to think of using VBR as source of better "quality" per say, I like to think of it as using what is best to get the most effcient results out of physoacoustics algorithm, (i.e allocating more bits for transients where there is sharp attack as apposed to using less bits for stationary signals or silence) maybe to prevent pre-echo. I also think the algorithm takes the best aspects in it's physcoacoustic design out of all the industry "standards" while not infringing upon any patents of any sort. In the future there will be newer apects of the codec worth paying attention to in physcoacoutic design (i.e two-switched iterated filterbank, (Discrete Wavelets) and anything else that will further benefit the codec.

Somethings I would like to see in Vorbis maybe in the future pre-beta releases as well is the ability to experiment a little more with the codec if your into the technical aspects or a developer (command line switches) and having the ability to choose your own free-bitstream with the -b switch once it has been optimized rather than using the -q switch for all of your basic needs. I think the quality coeffcient switch would benifit front-end GUI's more especially if you are the rather simplicit user or you want to encode on the fly.

Terminology:

As for "CD Quality" well I think that when your talking in terms of Digital Audio Compression that term is not appropriate, instead "CD Transparency" would sound better.
Transperancy various upon your average user (i.e one person may strongly prefer to use lossy coupling and not hear any diffrence while the other will use lossless coupling, because they can distinguish a diffrence or do not prefer lossy data physcologically), as for me I will use either lossy or lossless coupling it doesn't matter to me, I find subjective diffrences between the two to be least noticeable at times. With lossy coupling well that's the point of audio compression isn't it?


MP3PRO Model:

For the MPEG-1 Layer 3 Pro-Codec (MP3PRO), I think what Fraunhoffer IIS came up with to improve the physcoacoutics algorithm wasn't a bad idea in that of Spectral Band Replication for low-bitstream audio coding, however I think that the only drawback is that the codec (MP3) is more than 15 years old and the LAME project has since well experimented with the codec in a rather interesting fashion as well as implementing new feature to improve upon it's performance. Something like Spectral Band Replication would prove to be more efficient for the AAC physcoacoustics algorithm, rather than MP3PRO which just seemed more like a mass marketing scheme, but turned out to work quite well.
budding I.T professional

Why not to use WMA/mp3pro

Reply #11
Re: compulsory "Protected" WMA content on WMP8.

Peter,

I stand corrected. I just finished a new install on Windows. Looking into the default setting I find that it infact is set with a checkbox on Protection. The unwitting user with think this is a better feature. (Since when in the world has "protection" been a bad thing?) Thereby compelled my Microsoft to making horsesh!t files. The other kicker is the default bitrate slider bar at 64 labelled, CD quality.

~ Darwyn


 
SimplePortal 1.0.0 RC1 © 2008-2019