How do we submit audio versions of PG texts?

I've been doing some volunteering for Librivox recently[1], and would like to know the best way to submit the end product (human-read versions of PG texts) to PG. [1] http://www.librivox.org/ -- Jon Ingram

On Sun, Dec 18, 2005 at 07:14:35AM +0000, Jon Ingram wrote:
I've been doing some volunteering for Librivox recently[1], and would like to know the best way to submit the end product (human-read versions of PG texts) to PG.
So far, it has been just me posting any form of audio eBook. This is a problem, because I have a serious backlog -- between new audio eBooks, fixing broken computer-generated audio eBooks, and contemporary copyrighted text works along with their fixes/updates, I have a pretty substantial collection of stuff waiting for my personal attention. Probably 50-75 new items, plus fixes/updates for another 30-50. Posting this type of content is a bit of a hassle compared to .txt and .htm files, because some sort of readme.txt or readme.htm needs to be written to accompany the files. For multi-part files, an index.htm would be a better alternative. Then, the ww's regular "makehead10k" program can "wrap" the readme.txt (or whatever) with our header/footer, and all the .mp3 files need to be renamed per our standard convention (such as, eBook # 17555 will have 17555-m/17555-m.mp3 , with variations for multi-part files). Jon, if you're willing to do the preparation by hand (perhaps following some of our earlier models, linked from http://gutenberg.org/audio), I could pre-allocate some eBook #s to use. If you could get me just one .zip with everything pre-configured, it would be very easy for me to upload. Oh, and a standard "posted" email message, too. At least a dozen or so items from literalsystems.org are also ready to be brought in, if anyone is inspired to do these. These audio eBooks are actually very popular, and the human readings often get complimented. (The computer-generated readings are not nearly so popular, but still surprisingly popular.) I just wish there were enough hours in the day for me to work through all of these pending items, or someone else interested enough to take the lead (hint, hint). -- Greg

On 12/18/05, Greg Newby <gbnewby@pglaf.org> wrote:
Jon, if you're willing to do the preparation by hand (perhaps following some of our earlier models, linked from http://gutenberg.org/audio), I could pre-allocate some eBook #s to use. If you could get me just one .zip with everything pre-configured, it would be very easy for me to upload. Oh, and a standard "posted" email message, too.
At least a dozen or so items from literalsystems.org are also ready to be brought in, if anyone is inspired to do these. These audio eBooks are actually very popular, and the human readings often get complimented. (The computer-generated readings are not nearly so popular, but still surprisingly popular.) I just wish there were enough hours in the day for me to work through all of these pending items, or someone else interested enough to take the lead (hint, hint).
Thanks for the reply. I'll look into the preparation you require. The ultimate aim is that all Librivox material would be uploaded to Project Gutenberg, in addition to the Internet Archive, which is the current destination. Ideally this would be done in a fairly automated way. Emailing the files is probably out, as they're quite large (128kbit MP3, so approximately 1 meg per minute of audio). Is there any mechanism for uploading them to a server in the same way that normal texts are uploaded to PG? -- Jon Ingram

While we're on the subject ... On 12/18/05, Jon Ingram <jon.ingram@gmail.com> wrote:
On 12/18/05, Greg Newby <gbnewby@pglaf.org> wrote:
Jon, if you're willing to do the preparation by hand (perhaps following some of our earlier models, linked from http://gutenberg.org/audio), I could pre-allocate some eBook #s to use. If you could get me just one .zip with everything pre-configured, it would be very easy for me to upload. Oh, and a standard "posted" email message, too.
Is there a link between items allocated new eBook numbers, and the existing texts they're being read from..? I've only uploaded audio for one book so far, but because it was added at the same time as the plaintext and HTML, everything is now sitting under the same eText number, with just a separate folder for the audio files (which are also linked nicely from the HTML.) But if, for example, one read "A Christmas Carol" and it was given a new number - would / could there also be a link from eText #46 ..? Cori

On 12/18/05, Cori <hiddengreen@gmail.com> wrote:
Is there a link between items allocated new eBook numbers, and the existing texts they're being read from..? I've only uploaded audio for one book so far, but because it was added at the same time as the plaintext and HTML, everything is now sitting under the same eText number, with just a separate folder for the audio files (which are also linked nicely from the HTML.)
But if, for example, one read "A Christmas Carol" and it was given a new number - would / could there also be a link from eText #46 ..?
It would make sense to use the number of the source text -- the audio version is after all just a new edition of the same text. -- Jon Ingram

Jon Ingram wrote:
On 12/18/05, Cori <hiddengreen@gmail.com> wrote:
Is there a link between items allocated new eBook numbers, and the existing texts they're being read from..? I've only uploaded audio for one book so far, but because it was added at the same time as the plaintext and HTML, everything is now sitting under the same eText number, with just a separate folder for the audio files (which are also linked nicely from the HTML.)
But if, for example, one read "A Christmas Carol" and it was given a new number - would / could there also be a link from eText #46 ..?
It would make sense to use the number of the source text -- the audio version is after all just a new edition of the same text.
-- Jon Ingram _______________________________________________ gutvol-d mailing list gutvol-d@lists.pglaf.org http://lists.pglaf.org/listinfo.cgi/gutvol-d
If the audio file is read from our text, it should go into the etext number it was read from. If it derived from someone else's edition, then it would get a new number. At least, that's how I understand it. Josh

On Sun, 18 Dec 2005, Joshua Hutchinson wrote:
On 12/18/05, Cori <hiddengreen@gmail.com> wrote:
But if, for example, one read "A Christmas Carol" and it was given a new number - would / could there also be a link from eText #46 ..?
It would make sense to use the number of the source text -- the audio version is after all just a new edition of the same text.
-- Jon Ingram
If the audio file is read from our text, it should go into the etext number it was read from. If it derived from someone else's edition, then it would get a new number.
At least, that's how I understand it.
Josh
Yes, that sounds like a good idea. If there is any question, I'd say that posting the audio book under a new PG number would be the best choice. It's worth keeping in mind that, particularly for well known books, there can be multiple forms out there. A few examples: We have more than one edition of Darwin's "Origin of Species". We have a "young Folk's edition" of "Black Beauty" as well as "regular" edition. I understand that much of our P.G. Woodhouse is from American editions which sometimes vary greatly from the original British editions. So I would not be too hasty to put audiobooks under the same number unless we are certain that they prepared from that same text. We can make "See also" links in the bibrec pages if needed. Andrew

On Sun, Dec 18, 2005 at 11:09:08AM -0800, Andrew Sly wrote:
On Sun, 18 Dec 2005, Joshua Hutchinson wrote:
On 12/18/05, Cori <hiddengreen@gmail.com> wrote:
But if, for example, one read "A Christmas Carol" and it was given a new number - would / could there also be a link from eText #46 ..?
It would make sense to use the number of the source text -- the audio version is after all just a new edition of the same text.
-- Jon Ingram
If the audio file is read from our text, it should go into the etext number it was read from. If it derived from someone else's edition, then it would get a new number.
At least, that's how I understand it.
Josh
Yes, that sounds like a good idea. If there is any question, I'd say that posting the audio book under a new PG number would be the best choice. It's worth keeping in mind that, particularly for well known books, there can be multiple forms out there. A few examples: We have more than one edition of Darwin's "Origin of Species". We have a "young Folk's edition" of "Black Beauty" as well as "regular" edition. I understand that much of our P.G. Woodhouse is from American editions which sometimes vary greatly from the original British editions.
So I would not be too hasty to put audiobooks under the same number unless we are certain that they prepared from that same text.
I'm with Josh and Andrew. _If_ the audio is made from the same edition as ours, it's just a new format of that number. In some previous cases, the producer of the computer-read file claimed copyright, and that's why Greg gave them new numbers, because we couldn't post both copyrighted and PD content under the same number, even though it was made from the same source text. We also said we'd replace them when the technology improved. If we don't know that the audio is from the same edition, then it should get a new number. jim

On 12/18/05, Jim Tinsley <jtinsley@pobox.com> wrote:
I'm with Josh and Andrew. _If_ the audio is made from the same edition as ours, it's just a new format of that number. In some previous cases, the producer of the computer-read file claimed copyright, and that's why Greg gave them new numbers, because we couldn't post both copyrighted and PD content under the same number, even though it was made from the same source text. We also said we'd replace them when the technology improved.
If we don't know that the audio is from the same edition, then it should get a new number.
Sounds very sensible. In the case of all the Librivox projects I've been involved with, they've definitely been based on specific PG texts -- see http://librivox.org/forum/viewtopic.php?t=376 for an example. Librivox also explicity releases all the material it records back into the public domain, rather than retaining any rights over it. -- Jon Ingram

On Sun, Dec 18, 2005 at 10:33:22PM +0000, Jon Ingram wrote:
On 12/18/05, Jim Tinsley <jtinsley@pobox.com> wrote:
I'm with Josh and Andrew. _If_ the audio is made from the same edition as ours, it's just a new format of that number. In some previous cases, the producer of the computer-read file claimed copyright, and that's why Greg gave them new numbers, because we couldn't post both copyrighted and PD content under the same number, even though it was made from the same source text. We also said we'd replace them when the technology improved.
If we don't know that the audio is from the same edition, then it should get a new number.
Sounds very sensible. In the case of all the Librivox projects I've been involved with, they've definitely been based on specific PG texts -- see http://librivox.org/forum/viewtopic.php?t=376 for an example.
Librivox also explicity releases all the material it records back into the public domain, rather than retaining any rights over it.
I didn't know this -- I thought they kept a copyright with a creative commons-style license. Thanks for setting me straight. Mixing public domain with copyrighted items in the same eBook # is something I'd (still) like to avoid. So: Yes, we can just put these public domain audio performances in with the other formats. That makes the job slightly more difficult: for pre-10K items, it would be best to go ahead and update them for post-10K. (Not 100% required, but it would be nice -- otherwise, it's leaving a mess for someone to clean up in the future. Remember that the etext?? directories do *not* have subdirectories, making the post-10K structure much more suitable for multi-file audio eBooks.) For post-10K items it's a little simpler...just adding in the files. -- Greg

At 05:43 PM 12/18/2005 -0800, you wrote:
So: Yes, we can just put these public domain audio performances in with the other formats. That makes the job slightly more difficult: for pre-10K items, it would be best to go ahead and update them for post-10K. (Not 100% required, but it would be nice -- otherwise, it's leaving a mess for someone to clean up in the future. Remember that the etext?? directories do *not* have subdirectories, making the post-10K structure much more suitable for multi-file audio eBooks.)
Hello. Why not just add an extra zip file with the individual mp3 files? Not neat, but it would work. Just take the base etext name and add -mp3.zip to the end. That only adds one extra file. The problem of course is that you're stuck downloading everything. Is there any reason why there can't be subdirectories in the etext dirs? That way you could have one big zip file with mp3 files and one subdirectory with them individually available for download. Yes, it's still a mess but would get us by until all the ebooks are reposted.

On Mon, Dec 19, 2005 at 07:37:49AM -0800, Tony Baechler wrote:
At 05:43 PM 12/18/2005 -0800, you wrote:
So: Yes, we can just put these public domain audio performances in with the other formats. That makes the job slightly more difficult: for pre-10K items, it would be best to go ahead and update them for post-10K. (Not 100% required, but it would be nice -- otherwise, it's leaving a mess for someone to clean up in the future. Remember that the etext?? directories do *not* have subdirectories, making the post-10K structure much more suitable for multi-file audio eBooks.)
Hello. Why not just add an extra zip file with the individual mp3 files? Not neat, but it would work. Just take the base etext name and add -mp3.zip to the end. That only adds one extra file. The problem of course is that you're stuck downloading everything. Is there any reason why there can't be subdirectories in the etext dirs? That way you could have one big zip file with mp3 files and one subdirectory with them individually available for download. Yes, it's still a mess but would get us by until all the ebooks are reposted.
These things are possible, of course, but would be inconsistent with our other holdings. I'd rather be consistent, for many reasons. -- Greg

On Sun, Dec 18, 2005 at 11:08:51AM +0000, Jon Ingram wrote:
On 12/18/05, Greg Newby <gbnewby@pglaf.org> wrote:
Jon, if you're willing to do the preparation by hand (perhaps following some of our earlier models, linked from http://gutenberg.org/audio), I could pre-allocate some eBook #s to use. If you could get me just one .zip with everything pre-configured, it would be very easy for me to upload. Oh, and a standard "posted" email message, too.
At least a dozen or so items from literalsystems.org are also ready to be brought in, if anyone is inspired to do these. These audio eBooks are actually very popular, and the human readings often get complimented. (The computer-generated readings are not nearly so popular, but still surprisingly popular.) I just wish there were enough hours in the day for me to work through all of these pending items, or someone else interested enough to take the lead (hint, hint).
Thanks for the reply. I'll look into the preparation you require. The ultimate aim is that all Librivox material would be uploaded to Project Gutenberg, in addition to the Internet Archive, which is the current destination. Ideally this would be done in a fairly automated way. Emailing the files is probably out, as they're quite large (128kbit MP3, so approximately 1 meg per minute of audio). Is there any mechanism for uploading them to a server in the same way that normal texts are uploaded to PG?
Yes -- there is a non-anoymous ftp server on pglaf.org. Or, you can just put 'em where I can find 'em on ibiblio.org. If we can figure out how to make this "fit" with our existing processes, there is also http://upload.pglaf.org . So far, as I described earlier, it's not been a great match... -- Greg
participants (7)
-
Andrew Sly
-
Cori
-
Greg Newby
-
Jim Tinsley
-
Jon Ingram
-
Joshua Hutchinson
-
Tony Baechler