Gismu audio crowd source project (was: Lojban on Memrise)

77 views
Skip to first unread message

Jonathan Jones

unread,
Sep 6, 2012, 5:30:01 AM9/6/12
to loj...@googlegroups.com, lojban-b...@googlegroups.com
Each job will be essentially the same, with the only difference between them being the actual words in your chunk

With that in mind, these are the skills and software that are required to do this job:

You must be able to recognize a gismu from it's pronunciation. This means that you must be able to identify and type the word solely from the audio. If you know Lojban's pronunciation, this is not difficult. This is required because while Lindar recorded the audio /mostly/ in frequency order, some were skipped, and some needed to be re-recorded. If you need a list of the gismu in frequency order, you can find it here: http://www.lojban.org/tiki/Word+frequency+lists%3A+gismu

It is possible that there are more of both unrecorded and badly recorded words (badly recorded being words that are unidentifiable for any reason). If you encounter either in the course of your job, you must be able to contact Lindar, likely preferably via email, to inform him of these words so that he can submit new recordings. In the case of a word or three, preferably to yourself. In the case of a chunk's worth of words, preferably to me so I can add the recording to the chunk list for claimage.

You must possess, on the computer you will be using to work on, Audacity, available for free here: http://audacity.sourceforge.net/

You must possess, on the same computer, the LAME MP3 encoder. You can download it at this site: http://lame.sourceforge.net/, which provides the source code, or you can go here: http://lame.sourceforge.net/links.php#Binaries to download a binary, or for Windows users you can download the .dll directly from the instructions page that I will make if I get any volunteers.

You must know how to collect a group of files into an archive. It does not matter if you use .zip, .rar, .7z, .tar, .tar.gz, .tar.bz
 
, or any other format not mentioned.

You must know how to attach a file to an email.

You must be able to follow written instructions, which you can find here: http://tinyurl.com/gismu-instructions

Finally, you must be willing to sacrifice some of your free time to do the actual job. These last two are in my opinion no-brainers.

If you meet these criteria, please go to http://tinyurl.com/gacsp-signup, claim one or more chunks and type your name into the sign up field so everyone else knows you've claimed that chunk, and follow the download link to download your chunk in mp3 format.

There are 11 chunks total. Each chunk is ~6 minutes of audio, which based on my rather basic math should be ~100 gismu in each chunk.

When you have completed your chunk, place all the files into an archive and email them to me. Then, if you like, grab another chunk. If you don't like, go do something else.

I don't know how good of an incentive this is, but you should know that this audio, once it's been pulled apart into a usable format (the goal of this project), it will be the audio for the various learning programs currently in existence and probably for future ones as well. This means the current lessons on Anki and Memrise specifically, and more generally any other places where someone feels inclined to upload audio for how to pronounce the gismu.

--
mu'o mi'e .aionys.

.i.e'ucai ko cmima lo pilno be denpa bu .i doi.luk. mi patfu do zo'o
(Come to the Dot Side! Luke, I am your father. :D )

Lindar

unread,
Sep 6, 2012, 3:37:32 PM9/6/12
to lojban-b...@googlegroups.com
Once this becomes complete I will also tackle the ma'oste. Once that is complete, you can bet I'll do the experimentals and some of the broader vocabulary.

When do I get to hear the TTS engine based on my voice? I'm sure there's enough data out there to get a full sampling. =P

Jacob Errington

unread,
Sep 6, 2012, 6:24:15 PM9/6/12
to lojban-b...@googlegroups.com
On 6 September 2012 15:37, Lindar <lindar...@gmail.com> wrote:
Once this becomes complete I will also tackle the ma'oste. Once that is complete, you can bet I'll do the experimentals and some of the broader vocabulary.

When do I get to hear the TTS engine based on my voice? I'm sure there's enough data out there to get a full sampling. =P

Makes me want to record 'em all too ;)
 
.i mi'e la tsani mu'o 


--
You received this message because you are subscribed to the Google Groups "Lojban Beginners" group.
To view this discussion on the web visit https://groups.google.com/d/msg/lojban-beginners/-/rZY-4YlHgRoJ.
To post to this group, send email to lojban-b...@googlegroups.com.
To unsubscribe from this group, send email to lojban-beginne...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban-beginners?hl=en.


Jonathan Jones

unread,
Sep 6, 2012, 6:28:59 PM9/6/12
to lojban-b...@googlegroups.com
On Thu, Sep 6, 2012 at 4:24 PM, Jacob Errington <nict...@gmail.com> wrote:
On 6 September 2012 15:37, Lindar <lindar...@gmail.com> wrote:
Once this becomes complete I will also tackle the ma'oste. Once that is complete, you can bet I'll do the experimentals and some of the broader vocabulary.

When do I get to hear the TTS engine based on my voice? I'm sure there's enough data out there to get a full sampling. =P

Makes me want to record 'em all too ;)

I'd be happier if it made you want to help with processing the recordings we have already, instead.
 

.i mi'e la tsani mu'o 

--

Jonathan Jones

unread,
Sep 6, 2012, 6:44:08 PM9/6/12
to loj...@googlegroups.com, lojban-b...@googlegroups.com
If anyone who is willing to help is not doing so because they think they don't enough Lojban for the task, this page of the CLL is /all/ the spefically Lojban knowledge that is required:

http://dag.github.com/cll/2/2/:

....

Lojban has six recognized vowels: “a”, “e”, “i”, “o”, “u” and “y”. The first five are roughly pronounced as “a” as in “father”, “e” as in “let”, “i” as in “machine”, “o” as in “dome” and “u” as in “flute”. “y” is pronounced as the sound called “schwa”, that is, as the unstressed “a” as in “about” or “around”.

Twelve consonants in Lojban are pronounced more or less as their counterparts are in English: “b”, “d”, “f”, “k”, “l”, “m”, “n”, “p”, “r”, “t”, “v” and “z”. The letter “c”, on the other hand is pronounced as the “sh” in “hush”, while “j” is its voiced counterpart, the sound of the “s” in “pleasure”. “g” is always pronounced as it is in “gift”, never as in “giant”. “s” is as in “sell”, never as in “rose”. The sound of “x” is not found in English in normal words. It is found as “ch” in Scottish “loch”, as “j” in Spanish “junta”, and as “ch” in German “Bach”; it also appears in the English interjection “yecchh!”. It gets easier to say as you practice it. The letter “r” can be trilled, but doesn’t have to be.

The Lojban diphthongs “ai”, “ei”, “oi”, and “au” are pronounced much as in the English words “sigh”, “say”, “boy”, and “how”. Other Lojban diphthongs begin with an “i” pronounced like English “y” (for example, “io” is pronounced “yo”) or else with a “u” pronounced like English “w” (for example, “ua” is pronounced “wa”).

....

All Lojban words are pronounced as they are spelled: there are no silent letters.

la .lindar.

unread,
Sep 7, 2012, 1:22:59 AM9/7/12
to loj...@googlegroups.com, lojban-b...@googlegroups.com
As incentive for participation, I will award participants one of the following:

1. A wicked-awesome commissioned drawing with Lojban stuffs on it.
2. A hand-written thank-you note in Lojban.
3. A picture of me somewhere in Hollywood holding up a sign that says, in Lojban, whatever you would like.
4. A musical vignette, in Lojban, about how awesome you are.

Jonathan Jones

unread,
Sep 8, 2012, 4:17:18 AM9/8/12
to lojban-b...@googlegroups.com
I'd like to take this moment to thank Pierre for claiming one of the chunks, and also to wonder aloud why it is that v4hn, who previously stated that he desired to assist, has not as yet done so.

Again, thank you Pierre. At the least you have my sincere gratitude.

--
You received this message because you are subscribed to the Google Groups "Lojban Beginners" group.
To view this discussion on the web visit https://groups.google.com/d/msg/lojban-beginners/-/t5gGxW9sDyMJ.

To post to this group, send email to lojban-b...@googlegroups.com.
To unsubscribe from this group, send email to lojban-beginne...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban-beginners?hl=en.

Karin Dawson

unread,
Sep 8, 2012, 9:57:03 AM9/8/12
to lojban-b...@googlegroups.com
On 06/09/2012 5:30 AM, Jonathan Jones wrote:

You must possess, on the same computer, the LAME MP3 encoder. You can download it at this site: http://lame.sourceforge.net/, which provides the source code, or you can go here: http://lame.sourceforge.net/links.php#Binaries to download a binary, or for Windows users you can download the .dll directly from the instructions page that I will make if I get any volunteers.

I need the lame_enc.dll file. The above did not help me.

If you give me that, i will try to do a chunk.

Karin

Álvaro Vallejo

unread,
Sep 8, 2012, 4:07:36 PM9/8/12
to lojban-b...@googlegroups.com, loj...@googlegroups.com
Hi,

This thread seems interesting, but I think it is incomplete. What is the background here? It just starts in the middle of I don't know what: "Each job will be essentially the same, with the only difference between them being the actual words in your chunk ..."

I thank in advance any clarification, maybe I can help with this.

BR,

Álvaro

Jonathan Jones

unread,
Sep 8, 2012, 4:11:17 PM9/8/12
to lojban-b...@googlegroups.com

Lindar

unread,
Sep 8, 2012, 9:51:11 PM9/8/12
to lojban-b...@googlegroups.com
Looks like nine are unclaimed. Keep in mind that not only do you get audio in your study decks, but I will compensate each participant with some kind of soft good or mailed letter.

Thank you to Pierre, livefree, and aionys for helping out. Let me know if you'd like a music vignette, written thank you, or silly picture in Hollywood, in Lojban.

la .lindar.

unread,
Sep 9, 2012, 5:03:05 PM9/9/12
to loj...@googlegroups.com, lojban-b...@googlegroups.com
Please post here or e-mail me personally if there are errors or you need repeats. I'm going to re-record a mess of them later on (not today it seems) in one go. Also, if you're participating and want to grab one of those participation gifts, please let me know. The gifts are a hand-written thank-you note in Lojban (North America only, otherwise you'll get a scan of it e-mailed to you), a music vignette with Lojban lyrics, or a picture of me in Hollywood holding up a sign in Lojban saying whatever you'd like.

la .lindar.

unread,
Sep 10, 2012, 4:11:18 AM9/10/12
to loj...@googlegroups.com, lojban-b...@googlegroups.com
Craaaaaaaaaap... apparently I skipped a big bit? I was reading from the list of words in the gismu deck on Memrise, so maybe that deck isn't complete? Ugh... Oh well... Did you get both files when you started the whole thing? I have them both saved to my Google Drive, so please make sure you've got both of them. If that's still not the whole list then please try to get me something by Wednesday when I'll have a day off and the apartment to myself.

For those of you that are helping, please let me know what gift you want and fill in the request portion if applicable.

la .lindar.

unread,
Sep 10, 2012, 4:11:39 AM9/10/12
to loj...@googlegroups.com, lojban-b...@googlegroups.com

Jonathan Jones

unread,
Sep 10, 2012, 5:23:24 AM9/10/12
to lojban-b...@googlegroups.com
On Mon, Sep 10, 2012 at 2:11 AM, la .lindar. <lindar...@gmail.com> wrote:
Craaaaaaaaaap... apparently I skipped a big bit? I was reading from the list of words in the gismu deck on Memrise, so maybe that deck isn't complete? Ugh... Oh well... Did you get both files when you started the whole thing? I have them both saved to my Google Drive, so please make sure you've got both of them. If that's still not the whole list then please try to get me something by Wednesday when I'll have a day off and the apartment to myself.

For those of you that are helping, please let me know what gift you want and fill in the request portion if applicable.

I'm absolutely certain that, minus the portion I had already done myself, the 11 chunks were created from the entirety of the audio you made- both the original recording and the second one I asked you to make. I don't know /which/ deck on Memrise you used, but if it wasn't one that had 1343 words, then no. Considering which words /I/ personally noticed you skipped, I would assume that the most likely reason for the skipped words is that when you printed them out, the print-out did not print them all, possibly because on Memrise, they list 100 words per page, in a format that will not fit on one 8.5x11 sheet of paper. There's a scrollbar on that list, so I'd guess it only printed the words that appear when the scrollbar is all the way at the top of it's track.

I'm guessing, of course.

I don't know how long it'll take to finish the work we have right now, and I'll need that work done before I can determine which words were skipped. There's two days between now and Wednesday, so if the work is complete by tuesday afternoon my time, I'll be able to get you the list on time. Otherwise, you'll need to make other arrangements, find another good day. I wouldn't make any definite plans right now.

In order to prevent- or at least minimize the chance of- any words being skipped this time around, I'm going to ask you to do some things:

<snip> I was going to put this in the thread, but on reflection, I think the rest of this message would be better off just sent to you personally, which I am about to do.
 

Jonathan Jones

unread,
Sep 10, 2012, 6:46:57 AM9/10/12
to lojban-b...@googlegroups.com
Everyone that is working on this, please check the sign up page at your convenience, as I've added a few comments I'd like you to read.

Also, we need someone or ones to claim the remaining 3 chunks. The sooner we get these finished, the sooner we can figure out what's missing and get those recorded and put up for working on.

Jonathan Jones

unread,
Sep 10, 2012, 4:32:05 PM9/10/12
to loj...@googlegroups.com, lojban-b...@googlegroups.com
Chunk 9 is officially finished. It contained 49 gismu. This brings the total of completed gismu to 507, and the total completed chunks of 11 to 8.

Thank you again Salad, for the three chunks you've brought to completion.

The remaining three chunks have been claimed and are in progress. The people working on these are Pierre, whose latest report says he has finished 38 gismu and is thus nearly finished, livefree, who had just begun work on chunk7, and iesk, who is working on chunk8 and is having some computer problems.

Thanks again to all of you who have helped so far, and to anyone who wants to help and hasn't had a chance yet: Once we finish these chunks, we will be able to determine which gismu has not recorded audio for yet, and so we'll have a bunch more chunks put onto the sign up sheet a few days or so after the first 11 are all finished, so don't be discouraged.

Jonathan Jones

unread,
Sep 10, 2012, 11:40:48 PM9/10/12
to loj...@googlegroups.com, lojban-b...@googlegroups.com
Chunks 1 and 7 are officially finished. They contained 40 and 50 gismu respectively. This brings the total of completed gismu to 578, and the total completed chunks to 10.

Thanks go to Pierre and livefree for these two chunks. We are now just waiting for  iesk, who is working on chunk8 and is having some computer problems.

Jonathan Jones

unread,
Sep 11, 2012, 4:07:28 PM9/11/12
to loj...@googlegroups.com, lojban-b...@googlegroups.com
Chunk 8 is officially finished. This brings the total completed gismu to 623. All 11 chunks have been completed, so I will now figure out which gismu still need to be recorded and will send the list to Lindar, who will then record the new chunks, and I will let you all know when those are available.

Thanks to iesk for getting this done.

la .lindar.

unread,
Sep 13, 2012, 3:07:26 AM9/13/12
to loj...@googlegroups.com, lojban-b...@googlegroups.com
I'm starting tomorrow, Thursday, after work.

iesk

unread,
Sep 9, 2012, 4:21:00 PM9/9/12
to loj...@googlegroups.com, lojban-b...@googlegroups.com
@ Lindar:

I am now working on chunk 8. One odd thing so far: As heard by me, you definitely say {pano panlo pano}. Which is interesting insofar as, if you didn’t mis-pronounce it two times out of three, my (German-native) ear recognises your (suppressed American-English-background, I guess) realisation of /l/ as … well, nothing.

By the way, I find your pronunciation surprisingly excellent so far! (The surprise is not because it is you who did the reading – I’m not familiar with your voice or usual accent – but because I hear no recognisable accent at all. It’s great!) Thanks for the audio.

iesk

iesk

unread,
Sep 11, 2012, 9:01:50 AM9/11/12
to loj...@googlegroups.com, lojban-b...@googlegroups.com
done

la .lindar.

unread,
Sep 16, 2012, 12:45:28 AM9/16/12
to loj...@googlegroups.com, lojban-b...@googlegroups.com
https://docs.google.com/open?id=0B2twDfvSFg4nN2tyak1DT0FJWms

Chunk 12
(Sorry!!)

Next chunk coming soon.

la .lindar.

unread,
Sep 16, 2012, 1:19:55 AM9/16/12
to loj...@googlegroups.com, lojban-b...@googlegroups.com
https://docs.google.com/open?id=0B2twDfvSFg4nSFJDbUtFUmQyMlk

Chunk 13
(Kids, don't drink and voice act.)

la .lindar.

unread,
Sep 16, 2012, 3:34:31 AM9/16/12
to loj...@googlegroups.com, lojban-b...@googlegroups.com
Chunk 14: https://docs.google.com/open?id=0B2twDfvSFg4nN2hKZzR5TGx4bG8


Oh man... I'm so sorry. Seriously.
I hope it's fun and not annoying.

More to come tomorrow.

coyo

unread,
Sep 21, 2012, 11:36:27 AM9/21/12
to loj...@googlegroups.com, lojban-b...@googlegroups.com
i am willing and soon able to assist with this.

i will sign up later today, if i remember to do so.

Jonathan Jones

unread,
Sep 25, 2012, 7:11:38 PM9/25/12
to lojban-b...@googlegroups.com, loj...@googlegroups.com
Chunks13, 15 and 21 are officially finished. Thanks go to livefree, .iesk., and .triliyn.

We still have 9 chunks that are unclaimed, so anyone that's willing to help, get off your butt!

Jonathan Jones

unread,
Sep 28, 2012, 2:49:54 PM9/28/12
to lojban-b...@googlegroups.com, loj...@googlegroups.com
Chunks 12, 14, 17, and 19 are officially finished. Thanks go to Daniel Lyons for completing these chunks.

Chunk 18 is finished as well, thanks to .pier.

We now have 1,019 of the 1352 gismu audio finished, and only 6 more chunks to go until we're finished with this project. Keep up the good work!

Lindar

unread,
Sep 28, 2012, 8:43:19 PM9/28/12
to lojban-b...@googlegroups.com
Only two have claimed gifts.
I can't remember who requested the pictures, but they're done.
A written thanks has also been requested, which will go out over the weekend and will have lots of neat jbo stuff.

Daniel Lyons

unread,
Sep 28, 2012, 4:44:28 PM9/28/12
to loj...@googlegroups.com, lojban-b...@googlegroups.com


On Friday, September 28, 2012 12:49:57 PM UTC-6, aionys wrote:
Chunks 12, 14, 17, and 19 are officially finished. Thanks go to Daniel Lyons for completing these chunks.

I hope there's somewhere to upload Lindar's hard lemonade remarks, even though they aren't germane to this project. It was rather cool to hear some casual Lojban.

Jonathan Jones

unread,
Nov 27, 2012, 7:03:06 PM11/27/12
to lojban-b...@googlegroups.com
So, I realize it's been awhile since my last update. My motherboard fried, and I had to wait until I could get a replacement before I could verify the work that's come in.

So, I can now officially announce that Chunk 20 is complete. Thanks go to vahn for the effort, and I'm sorry it took so long - he sent me his work on 7 Oct.

I also received Chunk 22 from Jesko. but it still needs work since he didn't do four of the words, so I'll announce that once I've finished his job for him.

At this point we only need work done on chunks 16. 23, 24, and 25, and all of these are unclaimed except for 23, which .pier. is working on.

We're nearly done people! Everyone say "Yay!"

Jonathan Jones

unread,
Nov 27, 2012, 7:20:31 PM11/27/12
to lojban-b...@googlegroups.com
Chunk 22 is now officially finished. Apparently there was a bit of overlap, since {rinsa} was in that chunk and at least one other. (I already had the word, so I compared it with the one from this chunk and decided which to keep.)

Thanks go to .iesk., who now can have three prizes if you so chooses.

doi.lindar. I'll need to have you do another recording session when we get these chunks done. It doesn't look like it'll be more than about 10-12 words, so it won't be nearly as long and arduous as the first two. :) Also, because it's only going to be a few words, I'll do the editing myself. Of course, none of that is going to happen until we get these last few chunks done, because I need to know which words we have before I can give you the final script, so this is just a heads-up.

Jonathan Jones

unread,
Nov 27, 2012, 8:01:44 PM11/27/12
to loj...@googlegroups.com, lojban-b...@googlegroups.com
Minor update: I went onto the sign-up page to claim one of the chunks myself, and it turns out that now the only chunk that hasn't been claimed is Chunk25.

So, who wants to be the really awesome person that claims that final chunk?

For anyone that doesn't remember what we're doing here, the first message in this thread tells you what software, skills, etc. you need, and provides links to all the required stuff, including the instructions, which I'll repost here:

On Thu, Sep 6, 2012 at 3:30 AM, Jonathan Jones <eye...@gmail.com> wrote:
Each job will be essentially the same, with the only difference between them being the actual words in your chunk.


With that in mind, these are the skills and software that are required to do this job:

You must be able to recognize a gismu from it's pronunciation. This means that you must be able to identify and type the word solely from the audio. If you know Lojban's pronunciation, this is not difficult. This is required because while Lindar recorded the audio /mostly/ in frequency order, some were skipped, and some needed to be re-recorded. If you need a list of the gismu in frequency order, you can find it here: http://www.lojban.org/tiki/Word+frequency+lists%3A+gismu

It is possible that there are more of both unrecorded and badly recorded words (badly recorded being words that are unidentifiable for any reason). If you encounter either in the course of your job, you must be able to contact Lindar, likely preferably via email, to inform him of these words so that he can submit new recordings. In the case of a word or three, preferably to yourself. In the case of a chunk's worth of words, preferably to me so I can add the recording to the chunk list for claimage.

You must possess, on the computer you will be using to work on, Audacity, available for free here: http://audacity.sourceforge.net/

You must possess, on the same computer, the LAME MP3 encoder. You can download it at this site: http://lame.sourceforge.net/, which provides the source code, or you can go here: http://lame.sourceforge.net/links.php#Binaries to download a binary, or for Windows users you can download the .dll directly from the sign-up page.

You must know how to collect a group of files into an archive. It does not matter if you use .zip, .rar, .7z, .tar, .tar.gz, .tar.bz, or any other format not mentioned.


You must know how to attach a file to an email.

You must be able to follow written instructions, which you can find here: http://tinyurl.com/gismu-instructions

Finally, you must be willing to sacrifice some of your free time to do the actual job. These last two are in my opinion no-brainers.

If you meet these criteria, please go to http://tinyurl.com/gacsp-signup, claim one or more chunks and type your name into the sign up field so everyone else knows you've claimed that chunk, and follow the download link to download your chunk in mp3 format.

When you have completed your chunk, place all the files into an archive and email them to me. Then, if you like, grab another chunk. If you don't like, go do something else.

I don't know how good of an incentive this is, but you should know that this audio, once it's been pulled apart into a usable format (the goal of this project), it will be the audio for the various learning programs currently in existence and probably for future ones as well. This means the current lessons on Anki and Memrise specifically, and more generally any other places where someone feels inclined to upload audio for how to pronounce the gismu.
Message has been deleted

Jonathan Jones

unread,
Dec 7, 2012, 1:06:28 AM12/7/12
to lojban-b...@googlegroups.com
Email me directly with the specific problem you're having. I may be able to help you fix it.

On Sat, Dec 1, 2012 at 1:37 PM, jongausib <so.co...@gmail.com> wrote:
So I claimed chunk 25, but I've got problems converting the gismu audio files to mp3. I installed audacity and the lame.dll plug in, but it won't work.
So either I fix the problem somehow or maybe someone else would like to take over chunk 25?

.uu go'i .i mu'omi'e jongausib


Den torsdagen den 6:e september 2012 kl. 11:30:04 UTC+2 skrev aionys:
Each job will be essentially the same, with the only difference between them being the actual words in your chunk

With that in mind, these are the skills and software that are required to do this job:

You must be able to recognize a gismu from it's pronunciation. This means that you must be able to identify and type the word solely from the audio. If you know Lojban's pronunciation, this is not difficult. This is required because while Lindar recorded the audio /mostly/ in frequency order, some were skipped, and some needed to be re-recorded. If you need a list of the gismu in frequency order, you can find it here: http://www.lojban.org/tiki/Word+frequency+lists%3A+gismu

It is possible that there are more of both unrecorded and badly recorded words (badly recorded being words that are unidentifiable for any reason). If you encounter either in the course of your job, you must be able to contact Lindar, likely preferably via email, to inform him of these words so that he can submit new recordings. In the case of a word or three, preferably to yourself. In the case of a chunk's worth of words, preferably to me so I can add the recording to the chunk list for claimage.

You must possess, on the computer you will be using to work on, Audacity, available for free here: http://audacity.sourceforge.net/

You must possess, on the same computer, the LAME MP3 encoder. You can download it at this site: http://lame.sourceforge.net/, which provides the source code, or you can go here: http://lame.sourceforge.net/links.php#Binaries to download a binary, or for Windows users you can download the .dll directly from the instructions page that I will make if I get any volunteers.

You must know how to collect a group of files into an archive. It does not matter if you use .zip, .rar, .7z, .tar, .tar.gz, .tar.bz
 
, or any other format not mentioned.

You must know how to attach a file to an email.

You must be able to follow written instructions, which you can find here: http://tinyurl.com/gismu-instructions

Finally, you must be willing to sacrifice some of your free time to do the actual job. These last two are in my opinion no-brainers.

If you meet these criteria, please go to http://tinyurl.com/gacsp-signup, claim one or more chunks and type your name into the sign up field so everyone else knows you've claimed that chunk, and follow the download link to download your chunk in mp3 format.

There are 11 chunks total. Each chunk is ~6 minutes of audio, which based on my rather basic math should be ~100 gismu in each chunk.

When you have completed your chunk, place all the files into an archive and email them to me. Then, if you like, grab another chunk. If you don't like, go do something else.

I don't know how good of an incentive this is, but you should know that this audio, once it's been pulled apart into a usable format (the goal of this project), it will be the audio for the various learning programs currently in existence and probably for future ones as well. This means the current lessons on Anki and Memrise specifically, and more generally any other places where someone feels inclined to upload audio for how to pronounce the gismu.

--
mu'o mi'e .aionys.

.i.e'ucai ko cmima lo pilno be denpa bu .i doi.luk. mi patfu do zo'o
(Come to the Dot Side! Luke, I am your father. :D )

--
You received this message because you are subscribed to the Google Groups "Lojban Beginners" group.
To view this discussion on the web visit https://groups.google.com/d/msg/lojban-beginners/-/ibMrNcM0TcoJ.

To post to this group, send email to lojban-b...@googlegroups.com.
To unsubscribe from this group, send email to lojban-beginne...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban-beginners?hl=en.

Jonathan Jones

unread,
Dec 23, 2012, 8:58:17 PM12/23/12
to lojban-b...@googlegroups.com
Chunk 25 is now officially finished, for 69 additional words, giving us a running total of 1186 of 1342 total gismu with human audio.

Thanks go to .jonguasib. for the effort. Talk to .lindar. about your prize, if you want one.

We have three chunks left to finish, all of which are in progress as I write this.

Jonathan Jones

unread,
Dec 31, 2012, 12:54:29 PM12/31/12
to lojban-b...@googlegroups.com
Chunk 23 is now officially finished, bringing the total of gismu with audio to 1236. Assuming the last two chunks have exactly 50 words with no overlap, that means that when these are finished, we will have 6 words that still haven't been recorded by Lindar. I should have my chunk done later today or tomorrow, which means we're just waiting on .andares. to finish his chunk so we can find out what those last few words are, get Lindar to record them, and then I'll finish them up myself.

Thanks go to .pier. for the chunk, and thanks to everyone for the work so far.

doi.andares., if you are having difficulties completing your chunk, please let us know so someone can take over and get this done as quickly as possible. You can send me what you have so far and I'll upload what's left of the chunk to the sign-up page for someone to finish, if you want.

Jonathan Jones

unread,
Dec 31, 2012, 4:09:45 PM12/31/12
to lojban-b...@googlegroups.com
Alrighty. I've just finshed Chunk16, adding 50 more completed gismu audio for a running total of 1286.

Now we're just waiting on Chunk24. If I don't hear from .andares. in the next day or so, I'm going to assume he's abandoned his chunk. As far as I'm concerned right now, though, if anyone else wants to work on it, be my guest.

mcc

unread,
Jan 1, 2013, 10:00:37 PM1/1/13
to lojban-b...@googlegroups.com
I'd like to do the job.

在 2013年1月1日星期二UTC+8上午5时09分45秒,aionys写道:

Jonathan Jones

unread,
Jan 1, 2013, 10:17:25 PM1/1/13
to lojban-b...@googlegroups.com
On Tue, Jan 1, 2013 at 8:00 PM, mcc <me.x...@gmail.com> wrote:
I'd like to do the job.

You're welcome to it. Do you need me to repost where the stuff is so you know what to do?

I assume you know English well, yes? I ask because of "在 2013年1月1日星期二U...",which looks like it's Japanese to me, and all the instructions are in English.
 
--
You received this message because you are subscribed to the Google Groups "Lojban Beginners" group.
To view this discussion on the web visit https://groups.google.com/d/msg/lojban-beginners/-/spw8He7F-9wJ.

To post to this group, send email to lojban-b...@googlegroups.com.
To unsubscribe from this group, send email to lojban-beginne...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban-beginners?hl=en.

Jonathan Jones

unread,
Jan 1, 2013, 10:19:28 PM1/1/13
to lojban-b...@googlegroups.com
Oh. Wow. I already got the work from you. Nice. I'll check it to make sure is passes muster in a jiff.

Assuming they're good, we'll only be missing the three that Lindar, um, forgot to record.

Jonathan Jones

unread,
Jan 1, 2013, 10:36:05 PM1/1/13
to lojban-b...@googlegroups.com
Well, except for the fact that you named the file for {surla} "sutra.mp3" for some reason, it's perfect. Thanks for being so quick.

And as I'm sure you're aware from Lindar's comments at the end of the chunk, if you like you can geta "prize" from him for the work you did.

mcc

unread,
Jan 2, 2013, 2:33:57 AM1/2/13
to lojban-b...@googlegroups.com
I heard it again, yes. it should be {surla}. thanks for your information. it's my mistake.

and it is my pleasure to participate. and thanks Lindar for providing the learning material,
that is the best "prize".

在 2013年1月2日星期三UTC+8上午11时36分05秒,aionys写道:

Jonathan Jones

unread,
Jan 2, 2013, 2:42:36 AM1/2/13
to lojban-b...@googlegroups.com
So, this project is officially finished. All the gismu have been recorded, analyzed, a separated, so we can use them to update the Anki decks and put them into the Memrise lessons.

We have a bash script to automagiaclly create places audio from the gismu as well, and it works perfectly (thanks wobsite!), but unfortunately the audio we have for "lo <SE>" is pretty crap and the resulting files sound, well, odd, to be kind.

I might be able to do some creative editing like I did to get the last three missing gismu (I used samples for other words with the needed sounds and pasted them together. :) ), but I haven't really looked into it as of right now.

I'll make a report once I've uploaded the files to the tiki on where to go to get them in you want them for anything.

Jonathan Jones

unread,
Jan 2, 2013, 2:56:57 AM1/2/13
to lojban-b...@googlegroups.com
All the audio for the gismu has been uploaded. It is linked at http://www.lojban.org/tiki/Multimedia in the "Other Recordings" section, and you can download it directly here: http://www.lojban.org/tiki/tiki-download_wiki_attachment.php?attId=931&download=y.

So, who wants to tackle the cmavo now? ;)

v4hn

unread,
Jan 2, 2013, 9:00:12 AM1/2/13
to lojban-b...@googlegroups.com
On Wed, Jan 02, 2013 at 12:56:57AM -0700, Jonathan Jones wrote:
> All the audio for the gismu has been uploaded. It is linked at
> http://www.lojban.org/tiki/Multimedia in the "Other Recordings" section,
> and you can download it directly here:
> http://www.lojban.org/tiki/tiki-download_wiki_attachment.php?attId=931&download=y

Nice! Thanks to everyone!

You mentioned Anki.
When/How will this be added to the anki decks?


v4hn

Jonathan Jones

unread,
Jan 2, 2013, 12:51:05 PM1/2/13
to lojban-b...@googlegroups.com
There was some discussion about that at the start of this project, but I don't remember the details. All I know for certain is that Robin originally created the Amki decks, and he expressed ideas for what to do with them once we got this audio.

I'll see if I can find that discussion. I'm pretty sure it was in the original topic.

Jonathan Jones

unread,
Jan 2, 2013, 1:08:48 PM1/2/13
to lojban-b...@googlegroups.com

Jonathan Jones

unread,
Jan 14, 2013, 1:30:22 PM1/14/13
to lojban-b...@googlegroups.com
Just a quick announcement:

The "simplified gismu" lesson on Memrise (http://www.memrise.com/course/37344/simplified-gismu/) now has Lindar's audio for all of the gismu.

Based on a thread in the forums for "gismu (frequency order)" (http://www.memrise.com/course/17295/gismu-frequency-order-3/thread/1285944/), the "simplified gismu" database may be imported into the "gismu (frequency order)" lesson, which means that the two lessons will share everything except the actual definitions.

For those that don't know, the "gismu (frequency order)" lesson uses the definitions as they are in the gismu list, while the "simplified gismu" lesson uses definitions edited to be more simple and concise. Other than this, and the available content (i.e. audio and mems), the two lessons are exactly the same, even to the point of having the words appear in the same order.

As an example of the difference, this is how the definition for {klama} appears in the two lessons:

gismu (frequency order):

"to come/go": x1 comes/goes to destination x2 from origin x3 via route x4 using means/vehicle x5; x1 goes to x2 from x3 via x4 using x5

simplified gismu:

x1 goes to x2 from x3 via x4 using means x5


If my guess is correct, the long-term goal is to have both lessons share everything /except/ the definition, so that all the mems, audio, etc. appear in both, and so that any progress made in one will be reflected in the other as well.
Reply all
Reply to author
Forward
0 new messages