For those of you who understand German...

... just take a look at this: http://computer.t-online.de/TOnl/Comp/soft...audiofiles.html

It's a T-Online report based on a so-called test at TecChannel. You simply won't believe it, I just don't know what to say. :puke:

And while you're at it, to shock yourself even more, take a look at this forum post at TecChannel.

OK I'll post a short conclusion for those of you who don't understand any German.

The core statements:
[list]

"... between the programs": That in other words means that Opticom producer pro (= MP3Enc) is just as good as Xing.

Oh hell. I haven't read the complete test on tecChannel yet, but that short conclusion on t-online.de basically is enough...

:jawdrop:

CU
Dominic

two conclusions

a) those guys are dorks

b) they don't have the ears to hear artifacts like 95% of the MP3 consumers.

I don't think they have a huge reader base. So it's not even worth complaining. Besides, I don't see an EMail address to send complaints to

another quick addendum:

Just post a link to the r3mix and hydrogenaudio pages/forums in the TEC channel forums. ;-) So maybe you can attract some Newbies.

some of their information is also out of date:

audioactive production studio pro now retails for 150$.. they say its still the old 370$ (:-o).. and i really dont know why they bothered testing both pro and light versions of audioactive.. they're both the same, except that the lite version is limited to a 20 file batch and 128kbits max bitrate, and obviously costs less. the encoding engine however is respectively the same as the pro version

a few other points:

the test they did is at 128kbits/sec. according to them, with some of the music "stressing" the algorithms to the max :

- xing removes information above 17kHz all the time at this bitrate as a tradeoff of its speed
- fhg does a "better job" based on a frequency analysis (it appears they used a spectrograph in cooledit for determine that)
- fhg does a better job encoding more "complex" music

Ooooh, they have quite a large reader base! We're talking about the webzine equivalent to IDG's PC-Welt (you have PC World over in the USA)! As long as the information they give is correct, they can be as superficial as they like, but this is completely wrong.

With cheeseheads like that writing reviews about MP3 software, I'm actually glad they didn't mention LAME, because they'd just grab the DLL, set it to high quality (@128 kbits/sec), get absolutely horrible sounding MP3s and, (because according to them it is a bad idea to cutoff frequencies above a certain point) they'd critisize the fact that the frequencies above 16 kHz are "lost".

@outscape: The information is bound to be outdated, as this review was issued in late 1999. I'll suggest to them to take it off their site completely...

CU

Dominic

good idea, t-online should do it the same way

It really is a pity to continue to see this kind of misinformed reporting by sites/publications with a large reader base. When you are working towards trying to educate people and provide real objective information about testing and quality, these kinds of things only continue to hamper the effort. Unfortunately though, most people don't seem to be too interested in really getting to the truth of the matter, instead resorting to short cuts or quick (and flawed) observations like those you see made from frequency analysis graphs and the like.

What can you do though really? I guess some people will just never quite get it. All of this kind of reminds me of the Washington Post article where they tested Vorbis and mp3PRO and a few other codecs. In that case (as it appears is the case here as well) the people had no idea how to perform any sort of objective testing and really had no real knowledge about the codecs they were testing themselves but came to some seriously off base conclusions.

This is exactly the problem. How shall we be able to create a high quality audio community if some "underinformed" testers, for not being to harsh , "steal" away the newbies and make them use Xing.

Improve Xing?

I doubt that any LAME developer would put some effort in a commercial encoder I don't wish it either.

But it would be the only solution: To improve the quality while maintaining the speed. That would be nearly impossible i think and if it would be, why don't we have lame versions that run at 20x realtime on a P2 200 MHz .

I know, but it'd probably be easier than trying to educate the masses to use Lame.

i don't think there's even a way to do that. imho you'll be better off writing the psy-model from scratch instead of improving all this terribly outdated encoding routine

in my view, xingenc is all about being a commercial product: fast and easy to use, it receives very positive reviews from utter morons, some of the biggest web and music merchants use xing to make downloadable music available to millions.. taking all this overwhelming publicity into considertion, why shouldn't the average user use a program like audiocatalyst?? quite honestly i'm not surprised why there are so many mp3s out there encoded with this piece of shit

i think realnetworks are quite satisfied with the attention it gets in its current state. thus why they don't bother improving it

It's a sad fact of life that Xing will always be a very popular encoder, simply because it's in everybodys face. Most people use Audiocatalyst and Xing because they don't know better. The program tells them 128kbps is CD quality, and they believe it is. After all, they're the "experts".
Maybe once the next generation format comes to the fore (like Vorbis), the amount of dire-quality rips floating about will lessen by a significant amount. However, I do not see this happening for a while yet, and I fear people will still be encoding in Xing for many, many years to come.

Bad quality rips one the internet are something that no-one will get away from, that is unless someone comes up with some amazing lossless wavepacker that gives smaller filesizes than MP3 does today, but somehow I doubt that.

Lets face it, if you want a job doing right, do it yourself. I don't know about anyone else, but the amount of poor quality rips about has led me to buy more CD's and encode them myself. Maybe there are some benefits to Xing after all? ... maybe not.

After obtaining a copy of the Xing software development kit (SDK), it might be possible to replace the xing DLLs with the Lame engine by creating a replacement with exactly the same interface and some useful presets for encoding.

Now that would be a *real* improvement

>>>'it might be possible to replace the xing DLLs with the Lame engine by creating a replacement with exactly the same interface and some useful presets for encoding.

Now that would be a *real* improvement'<<<

hehe.. yes it would

i assume you're talking about audiocatalyst.. i guess it could work out but how are you gonna get all the people or newbies to use this "modified" version of audiocatalyst? we're gonna have to go pirate if thats the case because i really doubt realnetworks will agree to switch from xing to lame. its actually kinda sad for them to insist on using xing in their software

also, if we were to do that, then we basically end up with a replica of audiograbber.. or something that resembles that.. audiocatalyst 2.1 is essetially a really old version of audiograbber with the newer xing encoder.. if we substitute lame instead of the xing engine we end up with the equivalent of audiograbber with lame_enc.dll

if thats the case then i prefer to buy a more recent (less buggy)version of audiograbber and just use it with the latest lame_enc.dll

... or have i misunderstood?

Naaaah, leave that LAME DLL, it sucks
Doesn't have all the faetures, and even if it had, then there would be no CD ripper that would be able to use them. The command-line encoder is the only option!
Hell, I think we really should start some huge campaign "Use EAC + LAME", but how?

CU
Dominic

Or maybe we should just give up on trying to get the idiots to make good encodes. They probably wouldn't make proper rips first of all anyway. It's really not that these people are lacking in information on how to make high quality MP3s - they just don't want to do too much work for it (like setting all those "fiddly" options in EAC). It's not a matter of being uneducated, it's a matter of ignorance.

I disagree. Those people don't have the knowledge! They think there's nothing more to making an MP3 than to insert a CD in the CD-ROM drive, read it, and that's all. I know lots of people who don't even know that an MP3 is encoded, not ripped. And you could throw the worst sounding MP3s at them, they'd say "it's supposed to sound like that".
Oh yeah, and I have also had conversations like this many times: << 'What prog do you use for CD ripping/encoding?' - 'MusicMatch 6.' - 'Oh no! The MP3s sound really horrible with that!' - 'Ha, they might on your PC, they don't on mine!' >>.

But as you said, there still are many people who are just ignorant. Some of my friends (whose nerves I really have been getting on lately, trying to convince them to make _proper_ MP3s) do know about the dangers of MP3 encoding, but they excuse themselves "I don't hear what's wrong with 128kbps MP3, so why shouldn't I use it".

This is where I have got by trying to convince everybody to use proper programs (after many fierce, neverending flamewars, after having many people laugh at me and saying "rubbish, ringing noises, you must be imagining things, MP3 only takes out noises that humans can't hear, how should it introduce other noises?"): ONE out of probably 6 or 7 friends has listened to me, but he needs the knowledge as he's coding a Half-Life mod and wants to implement the sounds as MP3. He rips maybe 2 CDs a year. :confused: And the others who regularly rip CDs continue to use either FreeRIP MP3 (the most laughable ripper in the world, go check it out at http://www.mgshareware.com - I mean it seriously, just take a look at this software, you could nearly call it criminal) or Audiograbber + Radium ACM codec.

Anyway, we'll get there somehow, some day we'll have at least 10% of all MP3 making people convinced

CU

Dominic

PS: Umm, about that Half-Life mod... as I have the task to promote it: http://www.deltaeagle.net

Well, I agree that if they wish to learn, they should be provided with enough information to make good encodes. However, if they bring up arguments like "it's supposed to sound like that", don't you feel they're being ignorant already?