Wednesday, March 17, 2021

A blog post with a low signal-to-noise ratio

 

If you are unfamiliar with a subject, then you need to do careful research before you post about it (or you will reveal yourself as a darn fool). There is a blog post by Dr. John Livingston at the Gem State Patriot News on March 13, 2021 titled Noise and stuff and Stinky Fish which begins as follows:

 

“Humans can hear sound in the frequency range of 20 – 20,000 Hz vibrations per minute. There are, of course, sounds above and below the levels that we can hear. Radio frequencies are between 20 kHz – 30 GHz — much higher vibrations.”

 

The units at the end of that first sentence are wrong. Our hearing range is 20 to 20,000 cycles per second. A Hertz (Hz) is a cycle per second, which you can find out in a few seconds by looking at Wikipedia. 20,000 cycles per minute is just 333 cycles per second – barely over halfway up a piano keyboard where the 44 th key of 88, E4, is 330 Hz. And the Wikipedia article on radio frequency lists a range from 3 Hz to 3 THz. Those frequencies begin below our hearing range. 3 THz is 30,000 GHz or a thousand times higher than the 30 GHz that Livingston gave as the top of the range.

 

Livingston’s first paragraph continues:

“Like listening to the radio, oftentimes adjacent frequencies can blur the frequencies that are being listened to—this is called noise. The difference between the desired signal and the surrounding noise can be measured as a “noise to signal ratio” When we listen to the mainstream media there is lots of “noise”. In fact, many times most of what we listen to and read is “noise”. Even in the sports media there is more noise than signal—Are the Seahawks going to trade Russell Wilson to the Cowboys—who knows, but until it happens it is just noise. When it happens it becomes real news.”

But noise can be at the very same frequency as the desired signal. And the ratio normally is described via a signal-to-noise ratio (the desired signal divided by the accompanying noise). A Google search for the phrase “signal to noise ratio” gives about 15,100,000 results, while “noise to signal ratio” gives just 77,700 results or a factor of 194 times lower. His getting the numerator and denominator flipped is a dead giveaway he really doesn’t understand communication. He might as well be comparing Covid-19 with cooties.

 

And then Livingston continues with a pro-Trump political rant. On September 11, 2020 I blogged about Editing tips for speechwriters and other writers and used a couple of Livingston’s posts as examples for spelling errors. His current post has the abbreviation DJR where he clearly meant DJT (for Donald J. Trump).  

 

No comments: