Anyone have a URL to point to the changelog for 2.4? I checked out www.vorbis.com but didn't see anything.

Thanks!

This is about as close as you'll get:

With the release of vorbis 1.1.0, there was an additional 'quality' mode made available which allowed the user to specify the required quality by entering an approximate target bitrate in place of the usual 'q' setting. The standard oggenc from the SVN has still not been updated to reflect this, so oggenc2.4 makes that facility available. I've added a line in the help to explain how to use it.

OK?

John,

I'm a little confused by this. What does your new option actually do that the current oggenc from SVN (and your fork, which looks like it retains all the relevant code) does not?

I assume (from a quick glance at the --help message) that this is the -V option? I can't see what this does that the -b option doesn't do.

-b (by itself), in oggenc, sets an approximate target bitrate, in VBR mode (i.e. it's roughly equivalent to specifying the corresponding -q value). This seems to be the same as what -V does.

-b ALSO sets the target managed bitrate mode, but that only matters if managed bitrate is turned on (by either using --managed, or by specifying a min and/or max bitrate value).

Seems like -V is redundant?

Mike

Let's see if I've understood this correctly.

1. The new -V uses new features in libvorbis
2. The old VBR -b is largely similar to the new -V in function.

If the above is true, why not just make a new -b that uses new features in libvorbis? If it's successful, and if it results in cleaner code, I guess it won't be bad.

Not quite.
1. The new -V uses new features in libvorbis
2. The old -b uses precisely the same new features in libvorbis (but also does extra things if you turn on other options).

As far as I can tell, the -V option is unneccesary.

Ah! However... The documentation (the --help message, not the main documentation) is a bit misleading on what -b does. I'll go and fix this in mainline oggenc. John's then welcome to copy over my documentation fix, and possibly then remove his -V option :-)

Mike

Ok. Thanks for pointing out (indirectly) that the documentation for the -b option was confusing. If people are writing new options to do exactly the same things, there's something wrong :-)

I've committed some updates to the --help text and the manpage to xiph svn.

Mike

Thank you. While you're at it, the man page and the help output still show 0.0, not -1.0, as the lowest quality setting.

Thanks Mike for pointing out that I managed to add duplicate code adjacent to the original without even noticing it!!!!!! You just know that your brain must have been seriously out to lunch when that happens, don't you?

Anyway, to anyone that used the '-V' option, your resulting encodes will be identical when compared with having used the '-b' option without '--managed'.

Since I can't really reverse the version numbering, I've made the appropriate changes, adding the additional help info that Mike added, and uploaded oggenc2.41 to Rarewares in all flavours.

Sorry for any confusion here and thanks Mike for straightening this out.

Well, they mostly say -1 is the lowest quality setting, but there were a couple of other places where it implied otherwise. Anyway, fixed now, thanks for pointing that out.

Mike

p.s. Though I sometimes read things here, I often don't - so if you find other problems in oggenc, please feel free to email me about them.