Ricochet is the best place on the internet to discuss the issues of the day, either through commenting on posts or writing your own for our active and dynamic community in a fully moderated environment. In addition, the Ricochet Audio Network offers over 50 original podcasts with new episodes released every day.
Bumblebees and Frame Rate
For years it was posited that bumblebees could not fly — they simply could not displace enough air to make it possible. Nobody told the bees, of course, so they carried on anyway.
It turns out that the camera used to film bumblebees had a frame rate that captured half the bumblebee’s flaps, and nobody knew that the bee flapped twice as quickly as they thought it did — and so it flew.
I saw a tweet today where the frame rate of the camera shows a helicopter’s blades not moving at all. It is neat — check it out! Which reminded me of the bee, of course.
There is a deeper lesson here: we are limited by our instruments and very often don’t realize it. So instruments that do exactly what they are designed to do may nevertheless lead to erroneous conclusions. Hence, the bumblebee.
I thought it might be cool to hear from the engineers and scientists (or maybe even poets) on your stories of how instruments did their job, but people learned very incorrect lessons nevertheless. Please, share with the community!
Published in General
If you test light as a wave it behaves like a wave. If you test light as a particle it behaves like a particle. If you lay seven thermometers on a table and use a prism so light breaks into its colors and so that each color gets its own thermometer each color is a different temperature. Okay, you all knew all this and more but… what other delights is ‘light’ hiding that we have not even begun to dream of testing? What was that thing that “rang” off the coast of Africa a week or so ago that even scientists who specialize in unusual seismic events have never seen before? What does Mars know that Earth doesn’t?
So “Scientists” are SURE climate change will kill us all? Why? Because they know so much or their computer models are just that good? No and no. Just look at that amazing bee!
One of my favorite engineering stories is the investigation into the Apollo 13 accident. Thanks to NASA’s meticulous record-keeping and detailed engineering knowledge, they were able to figure out the complex sequence of events that led to the explosion of the oxygen tank, despite the fact that none of the actual hardware was available for inspection.
There were multiple small mishaps that contributed to the accident, but one of the causes was a temperature gauge on the oxygen tank. The gauge was designed to register temperatures up to 80 degrees Fahrenheit, because nobody expected the temperature inside the tank ever to go higher than that. But because of a voltage mismatch and a melted switch, the temperature inside the tank soared to over a thousand degrees during a pre-launch procedure, burning away insulation and leaving bare wires inside the tank. A spark from those bare wires is what ultimately caused the explosion.
The temperature gauge did not fail; it did exactly what it was designed to do, showing temperatures of up to 80 degrees and no higher. If your instrument is showing you a value that stays pinned at an extreme value, you really should ask yourself whether the real value is beyond what the instrument can show you.
We are also limited by our focused backgrounds to miss the “obvious” (a word used in a patent application) of the things around us.
Graduating with a Computer Science degree, and doing mostly microprocessor and Digital Signal Processing code for the first half of my career, I took a job with the title “Systems / Project” Engineer at a military radio company. A Systems Engineer (SE) is not expected to be an expert at anything, but to know enough to work with the various engineering disciplines such as Hardware, Software, Mechanical, Analysis, etc., that make up a Project. A Project Engineer (PE) oversees that the project has the resources to be completed on time and under budget.
I was assigned PE responsibility to update the military radio with a frequency synthesizer technique used commercially for about 10 years. With my SE hat on, I noticed that an extra input to the synthesizer would make it a Frequency Modulation (FM) radio transmitter building block with no special tuning needed. After reviewing the technical literature of other designs, there were none with this input. Like other patentable ideas, the company didn’t pursue this research project for budget reasons, but it was later proven to work.
Always ensure that your sampling rate is at least twice the rate of the phenomenon you’re trying to observe otherwise you get what’s known as “signal aliasing.”
Thus endeth the lesson.
Yep, the old Nyquist frequency. I learned about that years ago when I was learning about digital sampling keyboards … you used to get some weird aliasing noises with low-sampling-rate instruments, if you didn’t filter out anything above the Nyquist frequency. I suppose it’s not as much of a problem anymore, since I’m sure the Nyquist frequency of a modern sampler is way above the range of human hearing…
This is why audio on Compact Discs is sampled at 44,000 Hz even though humans can only hear up to about 20,000 Hz, correct?
(Telephone audio, by contrast, is sampled at only 8,000 Hz.)
If you increase the frame rate while recording a Chuck Schumer speech, how high do you have to go to before there is 50% chance of having one frame where he is not in the process of lying?
If you increase the frame rate while recording a Bill Clinton speech, how high do you have to go to before there is 50% chance of having one frame where he is not thinking of banging the twenty-something blonde in the first row?
Correct on both counts. Originally, for a realizable analog low pass filter in the frequency band (~300Hz to 3 kHz in telephones), you need 8,000 Hz rather than the Nyquist frequency of 6,000 Hz. Here ends a very simplified lesson, now it’s time to get back to iWe’s post (no insult implied!):
Hmm… once upon a time when I was a support tech I had to investigate an order that failed at the Summit. A ‘Summit’ is a device for measuring horizontal distances quickly and repeatably, don’t know why they call it that. In this case it was measuring the width of traces and the distance between them to ensure our copper plating process was doing it’s job right.
The dates didn’t line up. It had gone through the copper plater two weeks before. Why is it only getting here now? (As one might imagine measuring your work comes pretty quick after doing the work. If it doesn’t you can make a lot of scrap before you realize it.)
As it turns out the operator had run the darn thing upside down. Most products only go through that machine once, measuring the copper layer. This one had an extra two steps measuring the stainless steel etching job from the bottom. This guy had gotten the order, figured it was at the other step, loaded it upside down, and ran it with the copper measuring program. Wanna know why? Here’s what they named the workflow steps:
Care to guess which is which?
My son is a photographer, and we were talking about perception the other day. He pointed out that the visual cortex in the human brain takes up almost a third of the cerebral cortex. Vision is the most complex of all our senses, and that’s saying a lot when we realize how complex the other senses are.
In a case like that there is no reliable way to guess. (Been there, done that to myself, even. Well, it isn’t at all my preferred naming system, but who knows what I might do in a pinch.) So you have to figure it out. Then when you come back the next time, figure it out again because it’s too easy to forget.
Being both an oboe player, surgeon and amateur astronomy buff, the concept of being limited by my instruments is second nature. We all are.
Most people don’t know it.
Think of an analog: we are limited by our inputs. So, if one only reads the New York Times..
Speaking of bees flying or not, my father had said that back in the forties he mathematically proved that bumble bees couldn’t fly based on size and weight and so forth, but concluded: Of course, they can fly, so that just tells you not to believe everything you read. I asked him how he did it. He said: I used a few false preconceptions [or was it starting measurement error]; it didn’t matter, I was just doing it for fun, to see if I could do it.
Apparently people liked the work, too.
This reminds me of my own oboe-playing days.
Despite being, overall, a fairly musical person, I made a terrible oboist. Some of this could have been due to instrument failure (teachers told me my oboe really was particularly difficult to keep in adjustment and make reeds for), but some of it was due to my pinkies being so much shorter than my other fingers you might think they’d been amputated at the top knuckle if you didn’t look closely. That, plus joint hypermobility, meant I could only reach all an oboe’s myriad pinkie keys while bending my other fingers backwards, a fault no teacher ever managed to correct, and which may have been anatomically impossible for me to correct.
Nothing wrong with an oboe having pinkie keys. Definitely something wrong with you if you try to play the oboe without really being able to reach them.
The instrument I really wanted to play was viola, but my parents’ rule was, no young string players in the house, on the assumption that listening to such a young ‘un practice would simply be too painful. I definitely got a fitting revenge on my parents’ prohibition with my horrible oboing.
Vision is also arguably controlled hallucination, occasionally corrected by input from the outside world. Or at least so say theories of the Bayesian brain (here’s one person’s informal summary). To these theories’ credit, they seem to do one helluva good job explaining why we experience optical illusions the way we do.
The bumblebee story dates to the late 1930s or early 1940s. Apparently some aeronautical engineers from Grumman bored while at a social event. (I understand alcohol was involved. When engineers get bored (especially with alcohol involved) they look for off-the-wall technical problems to solve,
(Back in the early 1980s when my late wife Janet was an assembler building science experiment packages at Johnson Space Center, the engineers she worked with got into this kind of argument at an office party. They argued over whether the Super Guppy could take off if you filled it with ping-pong balls. Calculators and slide rules came out, and consensus was soon reached. It would be to heavy.)
The engineers modeled the bumblebee as if it were a fixed-wing aircraft. And sure enough, it cannot fly as a fixed-wing aircraft.
Now if it is modeled as a type of helicopter, a bumblebee flies just fine. Of course, if you model a helicopter as a fixed-wing aircraft it cannot fly, either.
The bumblebee cannot fly is more a cautionary tale about using models properly (climate science folks, are you paying attention?) than it is about sampling rates. Not that sampling rates are unimportant. I have a sampling rate story illustrating this.
In my first year working on the Shuttle program, back in 1979 an experienced engineer showed me a plot of approach data for a Shuttle landing. An SR-71 flying flat out (Mach 3 or so) was used to simulate part of a Shuttle approach. (The part between the Texas border and the Florida west coast as I recall.) The approach velocity was negative, even as the range decreased. The range was skewed, too. The Blackbird was a lot farther than it actually was. The sampling rate was such the system was picking up a previous signal rather than the time-homogeneous signal. (You can see the effect in the movies of a propeller speeding up until its starts going backwards.).
The purpose of the exercise was to knock it through my engineering-college educated skull to run all test results through an “is this reasonable?” filter, and not blindly accept data,
I can’t resist quoting ET Jaynes here — bolding added:
Bringing this back to @iWe’s point about instruments that are in perfect working order, but nonetheless give flawed results due to user error, “the apparatus” gathering the data here is really the whole instrument-user complex. An instrument that’s exceptionally accurate if you use it just right — but almost no-one can — would usually be less useful than one somewhat less-accurate but more easy to use.
I heard some ingenious experimental physicist (maybe Milikan?) was quite brilliant at devising experiments, but too klutzy to carry them out himself and get anything like meaningful results. Dunno if that’s just urban legend, but there do seem to be people like that — clever, but still mired in apparatus error at the user end.
Here’s a review of a high end power line conditioner for audiophiles and studios. I know the guy that makes them. He can hear the difference that capacitors with the same spec made from different materials make; he worked that out in the prototypes. The alloy the contacts in the receptacle in the wall can make a difference.
He claims the higher harmonics above our auditory range make a difference in the sound. Maybe they do. We have photoreceptors in our eyes that don’t project to the visual cortex. Some are involved in circadian rhythm regulation. Others do other things. Individuals with specific focal brain lesions that render them blind respond to the emotions associated with facial expressions that they cannot see on a person in front of them whom they cannot see.
This makes sense. A friend of mine who is a doctor once told me that the most important thing he learned during his training was when NOT to believe the numbers.🙂
but were kiling teh be bc of colomy colaps bc globle warning!!!11
Also getting research dollars are scientists studying the effect of neonicotinoids on bees, for exa,ple.
Or by our ability to interpret input. Whichever applies. Maybe both. Maybe they overlap.
I remember a time somewhen around high school when my father claimed there was no music in a P.O.D. song. Maybe this one.
I could hear the music just fine–the wonderful musical arrangement of that sound an electric guitar can make that is, in itself, so terrible and non-musical. (I find it very therapeutic hearing that sort of music, not to mention spiritual–a symbol of something Boethius was talking about.)
Why couldn’t my father hear it? Was there something wrong with his ears, or with his brain? And why don’t I like opera? Something wrong with my ears, or with my brain? Not unlikely.
It’s long been evident that my ear-brain system is not very good at understanding words that aren’t enunciated properly. I hear gibberish on the radio when others hear lyrics.
I told them myself.
They got worried about it and had trouble flying for a while. They begged me never to mention it again. I agreed after extracting certain concessions. Now they never sting me, and I have an army of bees at my command.
Have we ever mentioned how much we like your posts and how we would never be a threat to you in any way whatsoever?
The human hearing system is a device that detects the pitch of a tone spoken or sung by a person over the phone even when the fundamental–the frequency that normally determines the perceived tone–is missing. The phone filters out most of pitches of a male voice.
The sensory system’s “incorrectness” corrects for the severe harmonic error introduced by the phone line.
The hearer subconsciously infers the missing low note from the overtones that are passed through the system.
That’s what I’ve been led to believe, anyway.
True. Here’s an example of how it’s used in audio mixing — and pipe organs:
“Have you listened to music sometime between 1920 and 2018? If so, you may be entitled to compensation for Frequency Fraud practiced by several music delivery systems. Call 888-555-1212 to see if you are due to get a payout.”