Which MP3 encoder?

  • Thread starter Thread starter regebro
  • Start date Start date
regebro

regebro

Insane Genious!
I thought I should upload my things to mp3.com, but they refuse anything else than 128kbps as ecoding speed, and the encoder (or rather the front-end, I think) I have used so far in it's freeware version does not allow this.

I have recognized that several encoders sound like crap. I think LameEnc seems to do alright, but maybe theres one that is even better? And which front end should I use?

I might add that I could not care less about ripping CD's or playing and orginizing mp3s at this moment. I won't do that, at least not from the studio PC. I just want to convert WAVs to MP3, and I normalize the WAVS and everything beforehand.
 
Oh Yeah.

Slackmaster's right...
SCMPX is the best freebie out there..
Love it's capabilities.
DJ
 
Hmmm. Got some weird effects where sounds seems to move around in the panning and just behave weirdly.
I'll do a "shootout" with LameEnc and the current stuff I use later in the week. (And I might include more shtuff if people ask me too).

Thanks for the tip.
 
MP3 shootout heat results!

I tested five different encoders, and both 128 and 56kbps mode (except Xing, where the free demo only allows 128kbp).
I used five different songs from a reference CD that came with a HiFi magazine I bought. I had an excellently produced rock-song, some new-age-ethnic stuff and some classical. I excpected the classical stuff with its exceptionally clear recordings to make a big impact, but in fact it was the rock song that had the clearest differencies.

The following encoders were tested:
- Fraunhofer (via Yunasoft, AudioActive and MusicMatch)
- Blade (via CDCOPY)
- Lame (also via CDCOPY)
- Xing (via AudioCatalyst)
- SCMPX (via SCMPX)

Of these, the Blade Encoder sucked big. It sounded like crap in 128k, and the 56k wasn't even listenable. There was a lot of the typical low frequency modulation you get with crappy mp3 encodings.

I wasn't able to the the Xing Audiocatalyst in 56K, but I tested the others, and here I found the really big differencies. With LameEnc the biggest problem was the heavy loss of high-end you'd expect from an encoding speed that is rated as 22.05khz instead of 44.1khz. Also some modulation was discernible in reverbations on some of the material. The SCMPX encoder also had the loss of high end, but has severe modulation problems over all the materials, and even some distortion. The happy surprise here was definitely Fraunhofer, who not only has no discernible modulation problems, but also a very small loss of high-end! The 56k encoding af Fraunhofer is ALMOST as good as the 128k decoder.

In the 56k contest Fraunhofer crushed the other encoders like flies! Too bad the Xing did a walk-over... :)
If somebody has a license for AudioCatalyst I would be happy to send over some huge Wav-files with program material to test it in 56K too!

When it comes to 128k, which is the most interesting, differences between the encoders were very small, (except for Blade that sucked bigtime) and they all sounded excellent. There were small differencies, but it was hard to say which was the best.

I will therefore continue with a "final" maybe this weekend, where I try to get more problematic material. If you have found material that is difficult to encode, please tell me!
 
You must log in or register to reply here.
Back
Top