is oggenc the best ogg encoder in turms of quality and not speed

Well, I'm not an Ogg Vorbis user (yet), but I'd say there's no other encoder B)

If you go here and download both:
"Command line encoder - 2002-07-27 - ICL6 compile - 189Kb - 1.0 - by John33"
and
"Command line encoder - 2002-08-20 - GCC3.2 compile for testing - 159Kb - 1.0 - by John33"

You can try them both and see which runs quicker on your PC. They both have the same quality and as far as I'm aware are the newest builds of oggenc and thus are both the best quality available.

IMHO all the Vorbis encoder produce exactly the same quality audio files. I have used the command line encoder for the RC3 release, plus someone did a ICL compile of the RC3 and mentioned the files were identical (someone can clear this up if it is not right). Right now I use john33 excellent program OggDropXPd, easy and fast!

If for some reason you do want to use oggenc then you can use Oggshell as a front end. You can get it here.

If you wish to be able to encode directly from lossless encoded files, take a look at oggenc2. It's exactly the same encoder as oggenc/oggdropXPd but with the additional input option. It's available at Mirror 1, below, as are all the others.

Will you be adding the extra input options to oggdropXP?

Hopefully, yes. It will take rather more time and effort, though!

If you are interested in highest quality you should have a look at Garf's "oggencgt2". It is IMO better than the latest encoder at high bitrates (~160 and 350 kbit/s).

Link plz !

Garf's 'website': http://sjeng.sourceforge.net/ftp/vorbis/

But Garf's version is not based on Vorbis 1.0, right?

Correct. Garf's version is based in RC2, IIRC.

ErikS:


I can't comment on the 350kbps mode as I have never used it seriously. But I surely think that Vorbis 1.0 @ -q 5 provides better quality than Garf's 160kbps mode. I had a few samples that provoked really horrible HF artifacts in Garf's mode (one of which I remember posting here) - Vorbis 1.0 performs a lot better on these, even at -q 4 IIRC.

CU

Dominic

I can verify that Garf's GT2 modes did indeed perform quite a bit better on many samples than 1.0 at similar bitrates. Of course, I'm sure there are samples where 1.0 does better now because of other bugs being fixed (like what you mention), but for example, on many pre-echo inducing clips, Garf's modes were significantly better.

I was actually a bit surprised when I first tried 1.0 because it seemed like a general step backwards in quality to me. Not trying to start any flames here or anything, but I think a lot of this just has to do with the fact that Garf's Tuned modes where just that, highly tuned to provide the best quality within the limits of what Vorbis already provided. A lot of that was not carried over to 1.0 because of changes in code that made it not quite a simple task, according to what Garf has told me. So Vorbis 1.0 may be more stable overall in regards to quality than what the GT2 modes did, but it doesn't seem to produce as high quality as GT2 most of the time, at least on most of the difficult samples I've tried both on.

This really just kind of underlines the issue that 1.0 needs to have it's higher bitrate settings retuned in the same manner as before.

Well, as I said, I only tested very few samples (on which the difference was huge, though). I guess I was over-generalising a bit.

we all would love it if the higher bitrate modes could be improved