Hi there!
I don't know very much about these things, but I would like you to analyze these two versions
of the same MP3 file, and tell me which one is better and the differences. I would appreciate it.
The thing is just I have the sensation that the Fraunhofer version sounds more grounded, but I don't know why. And the Lame one is like yeah it must be better, but somehow I feel that it hasn't got the richness of the another one. I know these things may have personal views and etc etc but I am not sure by myself. I need some references.
Please tell me your opinions.
The links here:

Moderation: Links removed.

Greetings

UPDATE: An image of frequency analysis with CoolEdit Pro 2.1
http://img115.imageshack.us/img115/9748/fhgvslamebw9.jpg

Well, anyway for the links, I think I have my problem solved. Fraunhofer one had the best high frequency chart and sounded better for me. Thank you all.

What you need is a copy of the original recording in either .wav or a lossless format like flac or tak to compare them to and to carry out an abx test yourself to decide which is more like the original if you can even distinguish them at all...other than that, you just have to decide for yourself which sounds better to you as none of us have your ears...

sorry i can't be of any more help but what you are asking is purely subjective and you are the only person who can decide which sounds better...

There is a difference between attempting to test your ability to distinguish a difference, which is what an ABX test is for and attempting to test your preference for one over the other, which is what an ABC/HR test is for.
An ABX test can not tell you which one is "more like" the original, unless one is transparent to you while the other is not.

I noticed in the frequency analysis that the LAME version seems to have a lack of higher frequencies. Maybe it makes it worse...what do you think? I've heard that most of the MP3 have a lack in frequencies over 16 Khz; anyway, actually that means that an MP3 which has got all the frequencies until 20 Khz sounds better compared to one which has a lack, not?

@Vitilsky2: I suggest you do some reading on the forum before posting. You will find tons of information on the topics you are mentioning. Also, prepare to get a storm of "ABX" replies, so you better read up on that as well.

Not.
When you only have a limited number of bits to play with, it is often better to not spend those bits encoding high frequencies which not only are you unlikely to hear, but also cost more bits than lower frequencies.

a) Why must Lame be better? Even if Lame 3.96 (which is a bit outdated BTW) is better than the FhG version you used (whatever the meaning of 'better') this doesn't mean Lame 3.96 is better with every sample. Moreover FhG encoders are considered to be excellent no matter whether Lame may be 'better' or not. But when talking about Lame I suggest you try version 3.98 -V2. This should especially tackle your high frequency concern.
b) Looking at the context of your post chance is that your judgement is influenced by your high frequency considerations. To make sure there is a quality difference for you just do a blind listening test (use for instance the foobar ABX tool) and find out whether you can really differentiate the two encoders' results.
c) Other than that there's nothing wrong obeying to personal feeling. At a bitrate of ~192 kbps mp3 quality usually is perfect anyway no matter whether you use Lame or FhG. But again if you consider using Lame my advice is to give current version 3.98 a try, and do use -V2, -V1, or -V0 if you worry about high frequency content.

Why do people who know nothing about audio always insist on using spectrum analyzers to determine subjective quality of particular file? I have even seen some technical journalists do this it's vaguely annoying.





I would recommend you do a simple listening test to determine what frequencies you can hear up to. It consists of a simple sine tone sweep through the frequency spectrum. Do a search on these forums as it has been done before.

It looks like the Lame version was transcoded from another lower bitrate mp3, beause Lame sets the lowpass way higher than 16k at 192kbs. So I'd say the FHG encoding is possibly better quality.

What about LAME 3.97, 3.98.2 or 3.90.3 APS ?

Did you transcode these from the same source? i.e. take a lossless file and transcode it once with lame and once with fhg? unless you did you can't be sure theres not something else going on such as a transcode. Grab yourself a lossless file and try to abx that.