pentatonic Senior Member United States Joined 7247 days ago 221 posts - 245 votes
| Message 57 of 237 29 March 2005 at 9:57pm | IP Logged |
Update on digitizing the FSI courses
As an experiment on what it would take to digitize an FSI course I digitized one lesson of my Barron's German Level II course. The pages are just copies of the original FSI version but smaller. The audio is in tape format although probably higher quality than your typical course from the Foreign Service Institute.
Audio
Digitizing is painless because little supervision is needed; you just start recording and leave.
I didn't really have a good audio sample because the Barron's tapes have obviously been processed some. The hiss is very minimal and the volume levels fairly even. I experimented with processing the sound even though what I did didn't make that much of a difference in quality.
My conclusion regarding quality processing is that for the most part only the hiss needs to be removed. This can be done with free tools. The sound files need to be split to separate the drills. It's not too much work if you have the right tools.
I also experimented with Malcolm's recordings and they don't really need to be split. However, I believe he said they are from the Defense Language Institute and I don't think they are in the same format as any FSI course. He did the best he could given his circumstances and the rules of his library, but his recordings had a lot of hiss and benefitted a great deal from processing.
Text
OCR is tedious work and takes more time than working with the audio. I'm sure it would have been easier if I had a sheet feeder. Scanning requires supervision and the constant flipping of pages and cropping of images to be saved.
It's much easier if you are willing to tear your book apart. Pages held in three-ring binders would be ideal. I took the binding off my book because I have a paper trimmer to straighten the edges and a comb binder to put it all back together. The end result is actually pretty nice as it lies flat.
The type in my Barron's book is blurred from photocopying and I would guess that is representative of the quality of any FSI course. This makes the OCR less accurate and results in lots of editing during the spell-check phase. There are a lot of underlined words and this gives the OCR software problems.
Also, the end result is often not formatted correctly and editing is needed. Luckily the layout of the pages is fairly simple so it's really easier to just create a simple html template with the right columns and such for the page and to cut and paste into that.
The amount of work required for this project depends on the level of quality we are willing to accept. A high-quality finished product will take lots of work but I think we will be doing a great service to language learners and it will be well worth the effort.
Edited by Malcolm on 30 March 2005 at 1:39am
1 person has voted this message useful
|
Eric Senior Member Australia Joined 7228 days ago 102 posts - 105 votes Speaks: English* Studies: Spanish, French
| Message 58 of 237 30 March 2005 at 12:46am | IP Logged |
Side note - Any chance you guys will do the Spanish one after you've done with Mandarin?
1 person has voted this message useful
|
delectric Diglot Senior Member China Joined 7181 days ago 608 posts - 733 votes Speaks: English*, Mandarin Studies: German
| Message 59 of 237 30 March 2005 at 5:45am | IP Logged |
The website and the address of the guy selling this course online is:
Cris Stephens
The Language Warehouse
459 Bryant St,
Ormond Beach, FL 32174 U.S.A
www.tlwh.com
You can get the phone number of the company from this web page (my computer won't open it at the momment).
I have a feeling the courses on this website are from the Defense Language Institute as they offer a "headstart course" which I'm sure is from the DLI?
Previously you phoned up the company and bought their tapes over the phone. They offer 60 tapes in their French basic course and 74 in their Spanish basic course.
To keep the cost of the courses down they've digitized the text, which they give to you for free.
Now the website says they are digitizing these courses so they will not be selling their tapes anymore.
By the way, are there any plans to digitize the Mandarin Defense Language Institutes courses? From what I hear they are supposed to be excellent.
Why is it that the best learning material is always so scarce to find? You would think that having a more educated society, especially in today's world, would only benefit various governments.
I'm living in China right now. Who knows perhaps there is somewhere I can get these FSI course over here for a very cheap price.
Edited by Malcolm on 30 March 2005 at 2:51pm
1 person has voted this message useful
|
Malcolm Triglot Retired Moderator Senior Member Korea, South Joined 7315 days ago 500 posts - 515 votes 5 sounds Speaks: English*, Spanish, Korean Studies: Mandarin, Japanese, Latin
| Message 60 of 237 30 March 2005 at 3:12pm | IP Logged |
Eric: I will probably be digitizing the rest of the Mandarin course. I don't have access to the Spanish tapes, so the conversion of this course will happen independently of the Mandarin course. I believe ElComadreja said he owns all the tapes for this course and will convert them when they get returned to him. Also, many of us have the Barron's version of the first two levels, so it's very likely that this will be the next course to be converted.
Delectric: The course you're talking about is called "Standard Chinese: A Modular Approach". We've discussed it at length in this and previous threads and I've already converted one of the modules.
Everyone else: If anyone here has converted a non-copyrighted FSI course for personal use and would like to share it with the rest of us, please send me a private message or reply to this thread. I can arrange to share these courses with the other members through an FTP.
1 person has voted this message useful
|
guillaume Pentaglot Groupie France Joined 7181 days ago 59 posts - 57 votes Speaks: French*, English, German, Spanish, Japanese Studies: Mandarin
| Message 61 of 237 30 March 2005 at 4:51pm | IP Logged |
Malcolm: you said before that you wanted to start with Korean, do you
still plan on doing it ?
Edited by guillaume on 30 March 2005 at 4:51pm
1 person has voted this message useful
|
Malcolm Triglot Retired Moderator Senior Member Korea, South Joined 7315 days ago 500 posts - 515 votes 5 sounds Speaks: English*, Spanish, Korean Studies: Mandarin, Japanese, Latin
| Message 62 of 237 31 March 2005 at 12:15am | IP Logged |
Guillaume: I still plan to digitize the full FSI Korean course. This will happen gradually over the next few months. After that, I'll do whichever course is in demand.
1 person has voted this message useful
|
Alex Newbie Germany Joined 7188 days ago 29 posts - 31 votes
| Message 63 of 237 31 March 2005 at 10:32am | IP Logged |
I' currently digitizing levels 16-30 of the FSI Basic Spanish Course. I only have the tapes sold by Barron's though, so it probably wouldn't be legal to share these digitized versions. I'm trying to figure out the best way to convert the .wav-file to SMALL mp3s. Has anyone here experimented with different codecs and different settings? LAME (version 3.96) with --alt-preset voice gives good results but the the mp3s are still about 20MB for 1 hour of audio. Using --preset 40 -mm will yield smaller mp3 files but then you can clearly here a difference between the .wav and the mp3.
1 person has voted this message useful
|
pentatonic Senior Member United States Joined 7247 days ago 221 posts - 245 votes
| Message 64 of 237 31 March 2005 at 11:12am | IP Logged |
Alex wrote:
Has anyone here experimented with different codecs and different settings? LAME (version 3.96) |
|
|
I've had the smallest results with Fraunhofer's mp3Pro codec using the voice preset. With LAME I use the --voice switch with good results. I'm using version 3.93.
1 person has voted this message useful
|