Advogato blog for mlsm

10 Apr 2002

Wed, 10 Apr 2002 13:43:39 GMT

Stevey: We're sorry about the lack of documentation - lack of resources mean that documentation has taken a bit of a back seat. We will have documentation of the fileformat for 1.0 though.

If you don't mind looking at code - take a look at vorbis/lib/info.c: _vorbis_unpack_comment - while it isn't documentation, it should explain the format of the comment header (header number 2 - which is packet number 2) pretty clearly. The ogg layer is fully documented. We're happy to help out if you ask on the list or find us in #vorbis on OPN. Please do so.

For an example of parsing a number of important things from vorbis files without using libvorbis (it DOES use a significantly cut down version of libogg, but you could reimplement the bits of that needed seperately if you wanted to), you could look at the vorbis support in gnapster (it's possibly a bit out of date, I haven't looked at it since writing it, but it should be largely correct)

12 May 2000

Fri, 12 May 2000 08:07:41 GMT

Today is a good day. Why? Because it's friday. What this actually means, of course, is that I'll spend all of the next two days catching up on assignments for comp sci and physics that I haven't been doing, and on studying for elec eng stuff.

I also need to work on waking up - this has been a major weakness of mine recently, to the point where I've managed to make it to 2 lectures in the last three days. Next week I'll do better, promise.

Spent some time with the pain of windows again, updating the vorbis example (en|de)coder to work with windows, and misc. bug fixing. We've got a good variety of players too, now - freeamp, xmms, winamp, and at least 4 others.

Also got mid-semester test results back today - did pretty well on physics (thermodynamics), and passed the elec eng one by a reasonable amount (which was a relief).

Now I go to pick up some yummy takeaway chicken (mmm... foooood.....), and then in an hour or so I get to meet a Nobel prize winner (something I've never done before).

24 Apr 2000

Mon, 24 Apr 2000 14:45:47 GMT

Today I got nothing done. This was a good feeling a few days ago, at the start of this break, but it's beginning to get boring. Might have to do something productive tomorrow.

Sat around and watched The Phantom Menace again (on video - great birthday gift!). It's still pretty good.

I've also been reading through some books on signal processing/compression - notably "Vector quantization and signal compression" by Gersho and Gray. This is good stuff - I'm actually seeing relationships between the stuff I do at uni and the stuff I'm interested in for the first time. I might actually get motivated to do some real work on it as a result, rather than just scraping through with zero effort, as I seem to have done for the last 15 years of my life. That's a good thing.

Next, I need to find a good book on basic audio coding theory. Might go searching for one on wednesday, unless I get a good suggestion beforehand.

As predicted, windows programming has successfully got me ready to run screaming - those APIs look like they were designed by rabid monkeys. I think I'll leave it for a bit, and play around with some real programming. Much more fun.

21 Apr 2000

Fri, 21 Apr 2000 15:30:46 GMT

So, today I found myself at the beginning of a much-needed 10 day mid-semester break. And what did I do with it? Sat around doing nothing - damn, that feels good.

Then I got bored, so I went and rewrote the primary interface to vorbisfile. Oh - and sped up decode by 10% by changing a single line (now THAT'S optimisation). Vorbisfile is now much more flexible, so doing interesting things like streaming becomes relatively easy.

I think tomorrow I'll sit down and figure out how to do network stuff, and implement streaming in the winamp and xmms vorbis plugins. However, I'll probably run screaming from the windows one. This could be interesting...

I've also been spending some time getting into the real technical material concerning audio coding. This is lots of fun, though it's really hard work, compared to the generally trivial stuff I seem to do at university. Two big books sitting in front of me on VQ techniques, up next on my reading list.

Some notes on advogato: Found this place pretty recently. Some really smart people, and damn high s/n ratio compared to most of the net. Pretty nice - but it's somewhat hard to find your way around, a lot of things are only linked to from obscure sub-pages. I agree with what a lot of people have been saying - the range of the certifications is overly limited, and their scope is unclear.

Keeping a diary is a good idea. I think I might try and do this regularly - writing it all down helps organise stuff in my mind, and encourages actually thinking instead of just going and wasting time doing stuff that turns out to be entirely trivial