If you were to plot a graph of speaker fatalities in the lab, and you chose the appropriate scale, the curve described would be the letter `l'. Note the lowercase.
In a totally unrelated event to my own speaker incident, we've just fried one of the speakers from one of the 5.1 systems that make up our sonic array. What follows is a tale of tragic irony and the evil of Windows.
Etienne has been doing some calibration measurements of the signal emitted by each speaker in response to a sound source spatialized by OpenAL. To do this, he places a tie-clip microphone right at the speaker, plays the sound, and records the microphone signal. Since he is interested in the latency of OpenAL, the microphone signal is read by ASIO drivers, which provide very low-latency sound I/O. For sensible reasons, he does this at a low volume.
He was working on this yesterday, when it was discovered that the speaker cable driving that channel was lousy. He bought a new cable this morning, and connected it up this morning after successfully rebooting Windows (foreshadowing here). He moved to other tasks but left the speaker and microphone in position (italics for suspense).
A few minutes ago, Paul restarted the machine that was connected to the fateful speaker and microphone. He then logged in remotely from out in the lab.
Windows has two handy features:
1. When it starts up, it forgets a bunch of your sound settings like whether the microphone is on.
2. Regardless of your volume settings, it plays the triumphant "Windows is Here!" song at full volume.
By the time we reacted to the sound and entered the chamber, the feedback had reduced the speaker in question (fortunately the only one that was connected at the time) to a bad smell.
After the fact, we realized that it was the faulty cable that saved the speaker this morning and postponed its chilling fate. That's the irony part.
Posted: June 24, 2003 07:07 PM