[LAD] A little quiz about audio measurements...

Veronica Merryfield veronica.merryfield at tesco.net
Sat May 29 01:59:18 UTC 2010


On 2010-05-28, at 3:26 PM, fons at kokkinizita.net wrote:

> 49 Hz is impossible. It would mean that a clock using this
> frequency would be slow by more than a minute per hour. 
> 
> Long-term accuracy is extremely high, precisely because clocks
> depend on it. And since almost all of Europe is interconnected,
> adjusting the frequency must be a _very_ slow process. So errors
> of 2 percent really can't be tolerated.

Assume it at your peril. The usual spec is to arrange that over a given 24 hour period, the total cycle count will result in a mains clock being within 1 second. Often, the power generators can get much closer and I believe these days they use atomic clocks as the reference 24 hour time period.

The frequency will drop upon sudden demand. This is often coincident with the start of adverts in a national sporting event or high ratings soap, end of the news broadcast and so on. It takes a while to add more power and as a consequence the rotating machines will slow. Also, things like storm damage removing a significant section from the grid will result in an increase in frequency. 

Vrnc


More information about the Linux-audio-dev mailing list