Hello. Maybe this has been answered before.
I have a CD with some songs I downloaded years ago, and they were burned from Fraunhofer 128k CBR mp3 (yeah, I was stupid back then). Anyway I ripped and encoded that CD today and was surprised when the tracks, once encoded by Dibrom´s Lame -aps, had average bitrates of 130-140k?
This is completely irrelevant, but couldn´t those tracks be of any higher quality than 128 CBR? I mean, that was the limit bitrate they were encoded, right. How can they seem...better?

I'm not sure I understand your inquiry... Are you wondering why the --aps encoded files are better, even though the bitrate is roughly the same?

Well, not all bits are created equal, of course! Lame --aps makes smarter decisions about how to compress the audio. Just check out --death-preset ultimate to see how you can make Lame produce a 256kbps file that sounds like garbage

Fraunhuffy... its hard to beat it with its 20kHz or higher cutoff @ 128k. the new 3.94a9 sounds pretty good tho, at least on the joint-stereo tip, so far. it seems a little "cooked" at 128cbr still, but has 18kHz headroom so cant complain...

pre-echo on Farunhuffy still isnt touched tho imo

I would think that the increased noise/distortion/artifacts from the FhG 128kbps CBR file would make it harder to transcode, hence the larger file size?

for reference Fhg mp3... i would consider Fhg 1.2 (or radiums edit), Krystal Studio's rip of fhg 4.2 (included in nemo codec pack), the Fhg "cool" filter/encoder in cooleditpro2, or audioactive 2.0.4j (best ripper imo). set to MS-Stereo or Stereo, HQ (of course)

any advanced fhg or vinblows version is subpar in quality to those versions - imo B)

I'd nix that FhG 1.2 recommendation, it's got bugs in it (mangles out-of-phase signals in joint-stereo).

And your argument earlier, that FhG's 20KHz cutoff @ 128kbps is a strong point... I have to disagree; it's better to use those precious bits on frequencies that most people can actually hear.

As for your pre-echo comment, no MP3 encoder is near Lame --aps when it comes to pre-echo handling. I guess you're talking about 128kbps CBR, in which case you probably have more experience testing codecs than I do, so I'll take your word for it.

And I saw you snuck in some comment about "MS-Stereo or Stereo"... I suppose you'll need simple stereo to overcome that stupid, stupid bug in FhG 1.2 that's still got people afraid of joint-stereo, but other than that, Stereo is a bad thing; I'm sure you knew that, but I wanted to clarify for other readers who might not be as informed.

And your favorite ripper... well, I guess everyone is entitled to their own opinion. :x

When encoding with LAME --aps, the encoder can't know what is music and what are the artifacts introduced by the previous decoder/encoder combination. So it'll encode both the music and the artifacts, which increases the bitrate.

A higher bitrate doesn't automatically mean better quality BTW.

I think, the threadstarter did evil.

He burned a 128 kbit/s CDDA.

Then he ripped this CD and encoded AGAIN to aps now.

with typical result, quite low bitrate (compared to averaged aps), 130 - 140, but the sound will be worse then the 128-original or the CD................



Again:

Re- or Trans- Coding is EVIL !
from lossy to lossy format, like mp3 to mp3 in this example.