Maintenance

Digitizing of the remaining Indonesian audio tapes is going to be done this semester, with the current TA working as a workstudy on our workstudy project. She's due to show up at 1 o'clock, so I prepared the audio station. Various devices had been moved and unplugged; that's sorted now, and the top cassette player is working (the bottom one is broken).

First, I went through the process of digitizing using the current SoundForge app on the station. Then I updated all of our online instructions (they date back a few years, and are based on CoolEdit instead of SoundForge). Then I logged in as a regular user, and discovered that SoundForge can't run (it defaults to a folder in the hcmc user space as its temp folder, and can't write to it). I spent a while tracking this down, and eventually modified a couple of registry settings, to make it point to c:\sf_temp instead. Then I tested the digitizing process. I put copies of the existing indonesian stuff (Chapters 10 through 12) from Squash in the work_in_progress folder.

The latest Ubuntu release, Gutsy Gibbon, is close to final, so I built a VM in VMWare with it, to test out stuff such as IMT running under WINE (it dies).

We've acquired a new ION USB turntable, which is now connected to the audio station. Here are some bits of info about it:

• The turntable shows up as "USB Audio CODEC" in the audio devices lists, so if you're recording from it, that's the device to choose.
• In Audacity and Sonar, you need to turn on Input Monitoring to hear the input from the turntable when armed or recording.
• In SoundForge, there seems to be no way to get Input Monitoring to work; the docs suggest there's a toolbar button for it, but it's not there. So if you're using SF, you'll need to record without hearing the input. The VU meters work fine, though.
• The recommended way to use this is to use Audacity to capture the tracks, then use SoundForge to clean them up; we have a vinyl record plugin for SoundForge.
• The manuals for the turntable are with all the other manuals for audio station kit, in the blue binder on the desk under the big reel-to-reel.
• You can capture 78 records by capturing at 33 or 45, then slowing the audio down in the wave editor.
• The turntable also has standard RCA outs, for connecting to a regular audio system. NOTE that the level for these outputs is controlled by a switch on the bottom of the unit; it can output at regular line level, or at "phono" level (used when connecting to the Phono input of an older stereo system). That switch is currently set to Line, and shouldn't be changed unless you know exactly what you're doing and why you're doing it.
• The RCA line-outs from the turntable are not connected to anything right now, but in the long run they should probably be routed to the patch bay in place of the smaller reel-to-reel recorder, which is no longer connected. This would enable input monitoring via audio (albeit with some latency), as well as giving the option of digitizing through the FW 410 unit instead of via USB (if, for instance, we needed a much higher sample rate). I'll rig this up when I have a few hours; it'll mean changing the documentation for the patch bay.
• The little record-centre used when playing singles with their middles punched out is underneath the deck.
• We found an ancient record-cleaning device from the 60s which still works, so that's now plugged in and sitting next to the turntable.

I've sent in a request to dispose of a bunch of old equipment:
4 AOpen computers - the last of the old lab machines
One of the old CRT televisions
Sony Remote Commander - from Lab B
Sanyo digital projector - non-functional
Sharp VCR - from Lab B. Works, but not well
sundry computer parts (DVD-ROM drive, keyboard etc.)

Also moved a bunch of spare monitors and other bits and pieces to get them out of the way.

Forgot to post this last week:
My motherboard is toast, and it took me nearly the whole week to deal with it.
Discovery had a quick response once it was in their shop, and the verdict is that it's a warranty-replacement motherboard. Their best guess is 3 weeks for me to get it back.
In order to get back to work I pulled the drives from my dead box and rebuilt another machine.

Update: Got an update on Oct. 29. Apparently the new board is at the borderr and I should have the machine back Wednesday-ish.

Meeting called by KL about web2.uvic.ca at 2.30; I'll be leaving from there directly, so we locked up the office.

Did the periodic clean-up of my desk, which was long overdue.

Every time I install VMWare (of any type - workstation, server, player) I need to edit some source code in order to get the network module functional. I've done it so often I figured I should document it.

1) Run the installer. It will try to build a network module (vmmon) and it will fail. The problem is that the Ubuntu kernels are a target that moves too fast for VMWare to keep on top of.
2) When it fails, head over to /usr/lib/vmware/modules/source/ (this path will be different if you are installing the server or player versions) and
3) Unpack the vmmon source code thusly: sudo tar xvf vmmon.tar
4) Go into the resulting vmmon-only/include directory and sudo gedit compat_kernel.h
5) Edit the compat_kernel.h file to include the text in bold (below).

/*
* compat_exit() provides an access to the exit() function. It must
* be named compat_exit(), as exit() (with different signature) is
* provided by x86-64, arm and other (but not by i386).
*/
#define __NR_compat_exit __NR_exit
#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,19)
static inline _syscall1(int, compat_exit, int, exit_code);
#endif

6) Just to be safe, go up the tree and backup the old vmmon.tar: sudo mv vmmon.tar vmmon.orig.tar
7) Repack the edited module: sudo tar cvf vmmon.tar vmmon-only
8) Run the config again: sudo vmware-config.pl

*** an effective alternative is the use the any-any patch from here: http://platan.vc.cvut.cz/ftp/pub/vmware/
Look for the file that starts with "vmware-any-any-update"

Finished installing and updating Ubuntu Feisty under VMWare, and installed Konqueror. Tested the Unicode/UTF-8 AJAX stuff on Mariage and Moses, and everything works well.

This is just to document the results of some testing we did on digitized VHS tapes, comparing AVI with QuickTime for quality. High-quality AVI output was nowhere near as good as QuickTime with Sorensen 3, using 44.1 audio, square pixels, and 640x480. The QT files are twice as big, but it's worth it. We'll see how good they look when we use them as the basis for a normal DVD.