LeftStrobeE
LeftStrobeE is English version of bilingual article «LeftStrobe».
LeftStrobe is old story about the Old Big Telescope and Very Noisy Data from that telescope.
And about Redshifts that were extracted from these data.
For the case if some colleague wants to find the background of this story, or to find related stories, here is the keyword: «Scanner spectra».
The name «Scanner spectra» may look strange, even irrelevant,
but Editor sees no need to replace this name to something more "scientific" in order to
make the story more believable.
The story is written from memory. There may be some non-correctness, mistakes, misinterpretations in the text.
Editor hopes, the personages of the story are alive.
If something needs to be corrected in the narration below,
then, please, communicate the Editor.
Spectra
Mukashi-mukashi, I sold one of my algorithms as software.
The algorithm extracts the redshifts from so-called "scanner spectra" measured at Very Big Telescope.
Some spectra were very noisy; visually they looked like an irregular fence; no spectral lines were seen there.
Each spectrum appears as a couple of files, so-called "strobes", and an additional file, "passport", with specification of date, time, coordinates, conditions, astronomer, etc.
The "spectra" had some strange mode of encoding of data, but, after serious efforts, I got the arrays of values from the files for both "strobes".
The "Right Strobe" referred to an object; we were supposed to measure its speed (how fast is it going away from the Earth).
The "Left Strobe" provided the similar data from the sky near the object, with very similar "noise" due to the scattering/absorption of light in the atmosphere.
The "true spectrum" appeared as some transform of these "strobes"; in the simplest case, as subtraction of data of the "Left Strobe" from those of the "Right Strobe".
The algorithm “knew” what it was looking for.
It succeeded to extract the redshifts from very noisy spectra, that were considered to be "useless".
The astronomers were impressed.
The organization agreed to pay for my algorithm; of course, after the serious testing.
I was forced to re-write the algorithm to the "standard", portable Fortran.
I was forced to write the interactive scripts to pick-up the spectra and to save the results in such a way, that any secretute could run the program.
En fin, the final exam. The Customer gives the spectrum.
My algorithm, instead of to print the value of the redshift, says:
- "Oops.. Something may be wrong.. may be some of these?.. " - and suggests several absurd values.
I also said: "Oops.."
The Customer said: "Hmm.."
The project was at the edge of the shameful failure.
Without any hope, I asked timidly: "May I see the spectrum?"
The Customer showed the curve; it was apparently upside-down, with negative values of intensities.
- "Ah!" - I cried, "The files are confused! The spectrum is at the Left Strobe!"
- "Yes", the Customer said, "what is wrong? There is no confusion. In the passport of the spectrum, it is clearly written «Left Strobe»".
- "One moment" - I said.
I added 4 lines to the code. The code checked the sign of the spectrum. If negative values were dominant, the code swapped the input arrays.
After a minute, I demonstrated the Redshift.
The astronomers compared it to the value they had and said: "Ohlalaa!"
Then we run the code for another spectrum, then the same for few more spectra, then the batch for some hundreds spectra; we compared the redshifts to the data, measured manually; we discussed the precision of the new values that they did not had before, and so on; that is another story.
Analogy
The text above is loaded due to the analogy with interpretation of propaganda treated by the Rule of Newspeak.
Once some message is identified as a propaganda, then, in the first approach, each non-trivial statement there is euristically replaced to the opposite; "good" happens to be bad, "bad" happens to be good, the "hero" happens to be a war crime, and "bad guy" happens to be a hero.
The only one needs to "swap the strobes".
Then, the apparently wrong propaganda, meaningless lies provide the some reasonable, useful information. Strongly propagandistic messages may contain useful information once their semantic polarity is revealed, recognized and inverted.
The more propagandistic is the initial message, the more meaningful can be the output of such a transform - in analogy with a scanner spectrum: the "more negative" (more absurd) is the spectrum, the stronger, more precise, more accurate can be the result after the swapping of the strobes.
Attribution
ChatGPT helped to improve the text above.
References
Keywords
«LeftStrobe», «LeftStrobeE», «Newspeak», «NewspeakDictionary», «Propaganda», «Rule of Newspeak»,
«Правило новояза», «Пропаганда», «Словарь новояза»,