10. I would say this is a very important paper, but don't kill yourself doing it. Let the software guide much of the translation if possible.
If you don't do it, I'm going to have to learn German the fast way.
Jump to content
Posted 29 July 2013 - 06:56 AM
I agree this paper could turn out to be very crucial. I am having troubles getting the OCR to read it in properly and this is a very good OCR program (ABBYY). I would be happy to pay you for a translation. No immediate rush, although I am bursting at the seams to be able to read this paper in it's entirety I also know we all have lives to lead. Like geo said, don't kill yourself doing it but if you could do this I could see it being VERY helpful.
Posted 29 July 2013 - 06:50 PM
I had some thoughts regarding Hund’s paper. I think I may be able to help with the OCR process. I can rip each page of the paper into high resolution .pngs and edit them manually to carefully remove artifacts then we could proceed with OCR. Also I'll extract the diagrams and other illustrations from the paper. It would be nice if someone else could do the formulas to break up the work load. Then we could supply Gestalt with the OCR text, and cleaned up original pages for translation. And then finally piece all the work together, (diagrams, formulas, text etc.) in a new document with LaTeX, MS Word / Open Office or something similar. Literally, this project will require a team effort. I think if we break it up into sections we could get this done with minimal work on everyone’s part, except maybe Gestalt's as translating will be task in and of its self.
Also I would be willing to chip in some money to compensate Gestalt for taking the time to translate, if you’re still willing to help out.
Let me know what you guys think,
Posted 29 July 2013 - 06:56 PM
Oh, I had another thought to help increase the quality of this project. We could contact the website that Router got the papers from and see if they could do a higher resolution scan and then give them our finished work in exchange. They might be willing to go for it. Doesn't hurt to ask at least. But we would need Gestalt to write the email as I think they wouldn't speak Engish.
Posted 29 July 2013 - 07:31 PM
I don't mind chewing on the equations in LaTeX. Actually, we could just TeX up the whole document if that would make it easy to distribute the work load and give us some more experience fiddling with the pretty print equation output. Someone would do the equations, another the image extractions, and another the text translation, and so on.This way we just insert the translations in later and tweak the format to what ever floats your boat (paper, article, dual column, ...).
MSWord Bad. LaTeX good.
Posted 29 July 2013 - 07:55 PM
Haha, good call on using LaTeX Geo. I'll get the images extracted and clean up the document for OCR. Then the OCR system can focus on text and not worry about equations, diagrams and lots of weird black dots that somehow appear from bad scanning.Then maybe Router will have better luck with converting the scanned text with ABBYY? And finally Gestalt, when he gets some free time, should have an easier time in translating the document. I'll admit that I haven't made a document with LaTeX before but it should be a fun learning experience, plus it will look great when we are all done. On that note, I would like to mail this paper off to Eric when we finish with it, as I would love to hear his comments on it. And would be cool to post his thoughts on the new forum.
Posted 29 July 2013 - 08:46 PM
I did a bit of translating yesterday and in the process did two equations. My OCR has scanned in some pages ok but others it renders so poorly that it might as well have not rendered them. It detects the text fine though just doesn't render it =/ I've only used Tex for equations but I'd be keen to much around with Tex for getting the paper translated. Here is the site I got it from:
What I could do is break it down so that I extract sections of the pages that are stuffing up at a time and just put it all together because the OCR isn't too bad it just needs some tweaking. I'm not sure how much a higher resolution will help, I mean for sure it'll make some difference but I don't think it's vital we get a higher resolution copy for this project to succeed.
Posted 30 July 2013 - 01:58 AM
This is great guys this will be a massive help.
When you guys do up the document do it sequentially, with images and the block texts in between for ease of flow and translation.
The txt doesn't necessarily have to be OCR since just block images will work as well although OCR would help that is IF it can do the special german characters. (the dots above certain vowels and the "es zet" ß.)
If it can't do those just images of the text will suffice. I have to reword entire sentences anyways.
I also just remembered I have a relative that lives a couple streets from me an older retired guy, german, who was an electrical engineer for Siemens.
Apparently a smart guy who received several patents, and probably learned EE in an era closer to Steinmetz than what is taught in schools now so i'm crossing my fingers he is interested and knows this stuff well.
I have to do some genealogy stuff with him, and there is a good chance he can help me with the translation!!
I got all the translation covered.
Posted 31 July 2013 - 10:30 AM
I came across Bakoma, which is evidently the best WYSIWYG Latex editor. It compiles/interprets the code on the fly as one types, with the final result displayed in one pane, the code that one is working on in another pane, and error messages in a third pane. That and doing online searches has allowed me to make minimal use of manuals and quickly get a third of the way through that page of equations. I'm using the program on the 30-day evaluation plan. (screenshot attached)
Bakoma is very smart in correcting ones syntax errors, so the desired final (visual) result is easy to obtain.
BUT, there's a "problem" (not really): I must eventually understand and correct all the errors in the error pane before handing the code over to Geo, because that code must be fed into a compiler that might not be as forgiving as Bakoma. Which means (gulp) reading the manuals..
Still, it's an easy entry into the language for a beginner.
bakoma.jpg 121.33KB 0 downloads
Posted 31 July 2013 - 04:44 PM
Looks great. I will just need the *.tex output file generated from that process. You're right, the compiler/interpreter is not forgiving, but that's life.
Hmm. If there is way to to put the entire document together with a WYSIWYG editor, it would me much easier for editing. Additional formatting could be applied to the output later (or multiple with a make file).
I'll play around with lyx: http://www.lyx.org/
There's windows binaries too: http://www.lyx.org/Download
Good call Robert, I'm hooked.
Posted 01 August 2013 - 12:25 AM
At your mention of it, I tried Lyx but it didn't seem to have any advantages over Bakoma, and it lacked its WYSIWYG.
In Bakoma I then found that nearly all the errors (apart from a missing or excess brace) were due to my not having grasped the different "math modes" (inline or display) and the tags that are used for them. But even that was simply my wanting to make the page look pretty, which is redundant to producing the raw code that Geo requested.
Conclusion: Bakoma makes the user very efficient once one know the basics of Latex.
ADDENDUM: One thing to note is that the unlicensed version of Bakoma prints any hard copy out with a banner saying that it is an unregistered copy, which is alright for proof-reading and draft copies but not for final publication.
Posted 03 August 2013 - 09:19 PM
I did some searching, and I'm a little surprised that the subject matter of the paper has received almost zero attention anywhere in nearly a century, unless perhaps it is commonly called something else entirely (but my basic EE knowledge has no reference point). That is, a search on the terms bisymbolic AND electrical (on startpage.com, since I don't use google) gives exactly one result: garagehacker.com (i.e., the current discussion). Some of Hund's books are online, so I had a look through their contents and indexes for bisymbolic (no luck) but I did spot a footnote that referred the reader to the paper (in the German publication). I looked through the diagrams in the hope of finding clearer versions of some of the small and indistinct diagrams in the paper (no luck) since I reasoned that the author might re-use some of his diagrams in other works.
Lastly, I'm waiting to hear back from UC Berkeley to my enquiry as to whether Hund was ever on the faculty there and, if so, whether there is an archive of his papers. This is because, in the paper, Hund's name is followed by "Berkeley, California", which might mean the university.
Posted 03 August 2013 - 10:37 PM
I own a three of Hund’s books (soon to be four now), all in English, and can confirm that he did live in America and worked at the University of California in Berkley.
May I ask which book contained the foot note? As I would like to read that section to see if there is anything important to be found.
Interestingly, some of Hund’s diagrams, from the last pages of his Bi-Symbolic paper, appear in a reply to another paper: A discussion on "Oscillations in Antennas & Inductance Coils" by John M Miller. The diagrams appear in the 3rd page of this discussion.
It may be more fruitful to ask the German magazine for a higher quality scan of Hund's paper, as I don't think UC Berkley will have it.
Finally, it's interesting to note that in Hund’s later books, in English at least, he dropped the symbolic method and went back to calculus. Specifically his 1952 two volume set Short-Wave Radiation Phenomena; Vol1 and Vol2.
Posted 04 August 2013 - 01:02 AM
My god Robert!
Thank you! The preceding page (27) was most enlightening on negative resistance and Hund openly states that a negative resistance can act as a source of energy! This blows my mind right now. As I would have thought that "nonsense" such as this couldn't and wouldn’t be found in a text book. This is one of the many reasons I like Hund.
Hund's High-Fequency Measurements has two editions, each with slight variations on the negative resistance subsection.
1st Ed, 1933
2nd Ed, 1951
Posted 04 August 2013 - 01:31 AM
To anyone who cares,
After I finish my English final and get caught up at work, I'm going to build some negative resistance tube oscillators. While I couldn't find any dynatron tubes  I'm going to attempt to use another type in its place and see what happens. I'll post details once I start building.
0 members, 0 guests, 0 anonymous users