?

Log in

No account? Create an account

Hitlerto! - The online computery journal thingy of a turtle

Aug. 16th, 2008

01:30 am - Hitlerto!

Previous Entry Share Next Entry

This sample of badly OCR'd text is hilarious:

LOSE TIME IN SUBWAY. LOSE TIME IN SUBWAY. Investitures in England. From London Truth. Tile King has made an important change inI the regulations for investitures. Hitlerto the order tns been: "o insignia are worn by gentlemen ,vho are to be invested." In future, hob ever, the recipients of honors are to wear any order-=, decorations, or medals which they may have received. Inves: are now State ceremonials. Queen Victoria observed semiState if the function was at 'V'indsor Castle, But at O.borne there was a considerable reduction in the ceremonial, with fewer members of the Court in attendance. It is understood that next yeal' the periodical investitures el the Bath, the St. Michael and St. George, the Indian Orders, and the military decorations wlll be held by the King in the Throne Room at St. James's Palace. Queen Victoria usually investitures to the higher el ases o the several brders -- l e., Knights Grand Crosses and Knights ! Commandel's. ii The King,. how. e.er;' :h a i-d!yl Coilns?nted' to[: :th? om; :.pl Ons ;;Of. . :[6de .rs;!( wh: are ? tn .tro du.lc. :: !nto: ; hi :!Majesty,s f p$gence in:-:;:: S =vJt? ,':

Comments:

[User Picture]
From:thecanuckguy
Date:August 16th, 2008 03:36 pm (UTC)
(Link)
This creates more questions than it answers. How did you come across that? Why is the search term supposedly "Hitlerto" - who besides you would search for that word? How did it come to be? What was the original text? Why is badly OCR'd text on the web?
(Reply) (Thread)
[User Picture]
From:kinkyturtle
Date:August 16th, 2008 03:53 pm (UTC)
(Link)
The pun just popped into my head, and I decided to search for it and see if it occurred anywhere.
(Reply) (Parent) (Thread)
[User Picture]
From:kinkyturtle
Date:August 16th, 2008 03:59 pm (UTC)
(Link)
And if you follow the link I provided, you can find the original text in a PDF of an uncleaned-up scan of the original document.

The thing that made me laugh the most is how indecipherably mangled the OCRed text gets at the end. That, I figure from examining the PDF, is because the bottom of the paper was not lying flat against the scanner glass, making the scanner see the paper as grey instead of white, and when the image was converted to monochrome (every pixel either black or white) in the OCRing process, that grey turned into a lot of spots and blobby shapes surrounding the letters, which the computer desperately tried to interpret as text.

OCR is an amusingly fragile process. :}
(Reply) (Parent) (Thread)
[User Picture]
From:wbwolf
Date:August 16th, 2008 06:51 pm (UTC)
(Link)
Add to the fact that the original next is from 1905 and taken from what appears to be microfiche, it's surprising that the OCR turned out as (marginally) coherent as it did.
(Reply) (Parent) (Thread)
[User Picture]
From:patcat
Date:August 16th, 2008 05:11 pm (UTC)
(Link)
'V'indsor Castle, indeed!
(Reply) (Thread)
[User Picture]
From:deckardcanine
Date:August 16th, 2008 11:41 pm (UTC)
(Link)
That is how it would be pronounced Hitler-to, isn't it?
(Reply) (Parent) (Thread)