I'm going to conduct a multi-format listening test very soon, and I need only one encoder per format. I'm deciding between oggdropXPd 1.9.0 vs. oggenc2.85 (Either libVorbis v1.2.3 aoTuVb5.7), and I wanted on what your thoughts are, since I don't have much experience with Ogg Vorbis. All the encoders are coming from Rarewares.org.
I'm deciding between oggdropXPd 1.9.0 vs. oggenc2.85
That doesn't matter.
Hi ezlez. When og with what software? What bitrate?
I vote for libvorbis 1.2.3
Because it is newer and for me it sounds better than aotuvb5.7 at ~150 kbps.
Which bitrate you have choosen for test?
libvorbis 1.2.3 uses aoTuV b2 tweaks. So w.r.t. audio quality it is much older than aoTuV b5.7.
As long as you use aoTuV b5.7 it's fine. So yh oggdropXPd 1.9.0 using aoTuV b5.7
as lvqcl said, oggdropXPd vs. oggenc doesn't matter, that's simply an issue of frontend.
what does matter is which encoder. and it sounds like aoTuV b5.7 is the newer encoder (and more recently optimized, and thus likely but not guaranteed to be better)
Just my 2 cents - as already said, the question relates to frontends and not to encoding library. Since the results should be identical, it should be a matter of personal choice and have zero influence over the test result.
I vote for libvorbis 1.2.3
Because it is newer and for me it sounds better than aotuvb5.7 at ~150 kbps.
Which bitrate you have choosen for test?
Have you ever ABX'd this? Either one should usually be completely transparent at such a high bitrate. AoTuv Beta 5.7 is widely considered to be mostly transparent (except on a few killer samples) at -q3 or -q4, which correspond to nominal bitrates of 112 and 128 kbps respectively. Even -q2 (nominal bitrate 96 kbps) seldom produces severe artifacts that are noticable in casual listening. I can ABX LAME at ~128 kbps on a lot more samples than I can ABX Vorbis at ~96 kbps.
If someone wants to do some adventurous testing. Try out FFvorbis which is the vorbis encoder internal to ffmpeg. It generally produces horrific results[1]. It is suddenly interesting because a lot of people will be using ffmpeg to transcode into webm and some people seem to have a hard time believing that the ffmpeg encoder sounds worse than libvorbis or aoTuV. Some independent results might go a long way towards saving the world from billions of files encoded with that encoder.
[1] No blind testing is really required, for example: http://myrandomnode.dyndns.org:8080/~gmaxwell/fffail/new01/ (http://myrandomnode.dyndns.org:8080/~gmaxwell/fffail/new01/) and http://myrandomnode.dyndns.org:8080/~gmaxwell/fffail/new02/ (http://myrandomnode.dyndns.org:8080/~gmaxwell/fffail/new02/)
the ffmpeg guys lost my respect a while back due to how they treated the creator of mediacoder, Love some of the results of their products, but at least their groups mouth peices are a bunch of arseholes, dosnt surprise me at all that ffmpeg's encoder sucks, their theora encoder was HORRIBLE(if you can even get it to actually work.....)