In reply to:

---

I don't think there have been any real conclusions about the sound stage issue, which is what interested me in the first place :) I guess the only real conclusion is that low bitrate MP3 files definitely destroy the sound stage, that higher bitrate files may or may not destroy it, and that what we really need is WAV playback to prove for sure how good the empeg hardware is.

---

Sounds about right.. It's not really fair to blame the Empeg for mp3's shortcomings..
-mark

...proud to have owned one of the first Mark I units

In reply to:

---

Thanks for the links, I'll study them more closely and will certainly have a go with LAME and check for myself. (I passed testing Frauenhofer as they wanted some US$250 even for trying it out ). It's funny that, although the second link you quote is really negative on AudioCatalyst, the first link recommends it as it is supposedly LAME based:

---

Audiograbber and audiocatalyst are technically different products...
-mark

...proud to have owned one of the first Mark I units

In reply to:

---

Has anyone actually tried LAME? What's people's opinion's about it? I would like to include a little section in the FAQ about what PEOPLE (and not biased reviwers) think about the different encoders...

---

Well, I followed that link posted here not too long ago. http://www.r3mix.net/

It gave some pretty staggering results. It's makes a rather strong claim that Xing AudioCatalyst chops off the high frequencies above 15khz intentionally (and may be the reason for it being so damn fast). That really concerned me, although I've never personally noticed it when listening to a song. So, I downloaded LAME. The url above indicated that LAME was the only encoder that didn't mangle the song badly.

I encoded about 10 songs at the highest quality VBR setting of both encoders (normally, I encode my songs in Xing AudioCatalyst 2.1 at the higest VBR setting).

I did several of my own unprofessional and probably inconclusive tests, but they at least helped me formulate my own opinion. The first test was simply to play the song from beginning to end using the WAV first, then the LAME mp3 second and then the Xing mp3 last (for all 10 songs). The second test was to find a spot in the song that had a lot of high frequencies, then open up three players each with a different form of the song (wav, lame mp3 and xing mp3). The section of the song (2 or 3 seconds) was played in sequence using all three players in the same order as above.

Now these tests aren't in any way highly scientific. I just did them to try and decide if it was worth re-encoding my collection all over again. The nice graphical display at r3mix.net is probably true but I wanted to know how it sounded.

The biggest difference I noticed was that LAME takes much longer to encode the song (5 minutes for the song as compared to Xing's 45 seconds).

Through all the songs and tests, I could not tell the difference between Xing and LAME encodings at the highest VBR setting. At times I thought there might have been a difference between the mp3's themselves, but they both sounded identical to the original WAV file so the effect was probably pyschological.

I think the strongest selling point I can offer to LAME is that it is FREE, and it is a high quality encoder. It is probably a higher quality encoder than AudioCatalyst 2.1 but I couldn't tell the difference myself. For anyone out there who hasn't bought a good encoder yet, a good one is available for free and I would see no reason to purchase one. Another program that makes a great companion to LAME is the EAC (ExactAudioCopy) program (http://www.exactaudiocopy.de/), which is also free (I believe the author wants a postcard for registration, but the software doesn't appear to be crippled in anyway). These two products together make almost the same thing as Xing AudioCatalyst (without the speed).

Xing AudioCatalyst 2.1 has the benefits of speed and the look and ease of a little more professionalism. I didn't notice any of the quality sufferings that r3mix.net claimed existed, but I might just be deaf! I'm quite happy with it myself. If you've already been using Xing AudioCatalyst 2.1, I don't think there is a need for you to re-encode all of your songs again.

On the other hand, my tests were only with the highest VBR setting. r3mix.net indicated that even the highest VBR setting in Xing suffers the high-freq flaw, but I sure as hell didn't notice.

These are just my opinions of what I heard. I am not arguing any technical differences between the two, and I have not compared the CBR settings of the encoders (so for all of you out there that use a constant bitrate, you may have to do your own tests). Also, I only did testing with 2.1 of Xing. Prior versions will likely produce different results.

Hope this is helpful. Consider this just a review as others will have their own opinions.

Kureg

Note: VBR is currently under heavy development. Right now it can
often result in too much compression. I would recommend using VBR
with a minimum bitrate of 112kbs. This will let LAME increase
the bitrate for difficult-to-encode frames, but prevent LAME from
being too aggressive for simple frames.

In reply to:

---

This means that I have to specify a "bitrate floor" of 112kbps in the command line. Which, although understandable, means that I'm not getting the full bang-for-the-buck out of the VBR format. In other words, why go to the trouble of doing VBR if you can't have it compress the simple frames below 112? The result is larger file sizes, and the whole point of VBR was to get smaller file sizes.

---

What I've noticed on a few files is that the minimum bitrate seems to be a preferred minimum bitrate, and that there is a "force" option elsewhere. There are parts of the song that compress more than the minimum should limit, but hovers very close to that mark. This is very distinguishable in the fade-out portion of a song (at least for me ).

Kureg