Wikipedia talk:WikiProject Spoken Wikipedia

From Wikipedia, the free encyclopedia

Archives Archive 1 Archive 2 Archive 3 Archive 4 Archive 5 Archive 6

Contents 1 Currently recording? 2 Spoken article template link_to_recorded_version field 3 I screwed up--Corrected 4 Question about video 5 Spoken Article box 6 German Wikipedia project 7 Footnotes? 8 Quality assessment categories? 9 Quality assessment refinement 10 Main page featured articles 11 Request for brief feedback - The Four Stages of Cruelty 12 No Guidelines? 13 Jake Gyllenhaal 14 Placement of links to Spoken Version of Articles 15 Sung wikipedia? 16 A little change 17 Audio problems (question from MgM)

[edit] Currently recording?

Is there a list of articles currently being recorded? I see a list for post-recording editing, but not one so that we don't step over each other's toes and record something twice. --Brad Beattie (talk) 01:13, 17 January 2007 (UTC)

Since I think the post-recording list was actually serving both functions, and that seems a reasonable thing for it to do, I have changed the description to reflect that. -SCEhardT 03:12, 17 January 2007 (UTC)

Also, some articles have a template on the talk page indicating they are in progress (such as Talk:Nanotechnology) -SCEhardT 03:24, 17 January 2007 (UTC)

[edit] Spoken article template link_to_recorded_version field

People seem to be having some trouble knowing what to put here when they're uploading a recording, so there is now a brief guide. If anyone can see how to make improvements to it, please do so. Macropode 04:57, 26 January 2007 (UTC)

[edit] I screwed up--Corrected

I screwed up the formating when trying to write my pieces into the "being edited" table. I've looked over and can't figure out what I did wrong. I'm going to go hunting so I can figure out how to fix it, but if any body can correct it (without reverting or deleting my entries from the box...because I'll still have to face putting them back in) and tell me on my talk page what I did wrong, that would be super duper. Also, it would prevent me from making the same mistake next time I try to put a piece in the box. Ara Pelodi 20:41, 3 February 2007 (UTC)

I corrected it. It was as simple as putting the ending on the next space. Whew, wild ride of fear for the newbie!

Ara Pelodi 20:51, 3 February 2007 (UTC)

The wild ride of fear is the newbie's rite of passage in Wikipedia. I well remember my own. -- Macropode 06:15, 17 February 2007 (UTC)

[edit] Question about video

Is there any push to create video of wikipedia articles? For most cases you could just have the spoken version played over pictures describing the item. Remember 17:50, 7 February 2007 (UTC)

There would be some difficult technical issues to overcome, and in any case it's probably outside the scope of this project. -- Macropode 10:23, 17 February 2007 (UTC)

[edit] Spoken Article box

Why is the default location of the spoken article box in the external links section? The audio file is not an external link outside of Wikipedia. I suggest that this box should appear at the top of the article such that readers will see straight away that they have the alternative option of taking advantage of the recording. This is especially true of very long articles - having the spoken article box right at the end of the article means that readers can't see it easily and won't know that they can listen to that particular article unless they scroll right down, past the (maybe hundreds) of references and other external links. CupOBeans 16:30, 13 February 2007 (UTC)

There's some archived discussion relating to this. -- Macropode 06:05, 14 February 2007 (UTC)

[edit] German Wikipedia project

You guys might be interested in de:Wikipedia:Auszeichnungen, which is apparently an effort to vet quality FA Spoken Wikipedia articles. I wouldn't recommend actually developing a separate "featured" project here, but perhaps in the future having a quality Spoken article will become part of the featured article criteria.--Pharos 19:48, 15 February 2007 (UTC)

It would be an interesting exercise to examine how our sister spoken projects in other languages are doing things. Maybe someone with the necessary language skills would like to do this and draw some comparisons with what's happening here.

Spoken as part of the featured article criteria - now there's an interesting thought! -- Macropode 10:14, 17 February 2007 (UTC)

I notice most recent FAs already seem to have spoken versions. Presumably they are placed on priority in anticipation of the Main Page appearance? I would envision the FAC process giving a conditional approval, which is confirmed upon the recording of a quality spoken article based on the final version as passed (artticles often change drastically during the process). The Spoken Wikipedia project probably isn't mature enough for this yet, though. There would have to be considerable development of your proposed review system first, for example.--Pharos 03:44, 18 February 2007 (UTC)

Please excuse me, Pharos. What I should have said was that this is an interesting proposal in light of the Featured Article Director's historical willingness to take actions detrimental to this project without trying to solve the issues concerning him by engaging in any significant discussion here first. -- Macropode 01:45, 20 February 2007 (UTC)

[edit] Footnotes?

Are they left out, or how are they handled? What about other items like the headers and infoboxes? Шизомби 14:28, 16 February 2007 (UTC)

As a rule-of-thumb, read the core text, along with headings per the recording guidelines. If infoboxes provide information that isn't covered by the core text, and they won't add too much to the length of the recording, you could read them too. The "See also" and "External links" sections may also be useful to some listeners. Anything else would probably be tedious to read out, of little interest to the average listener (whoever that might be), and would make the recording unnecessarily long. -- Macropode 10:16, 17 February 2007 (UTC)

OK, thanks, although it doesn't indicate how to treat subheadings. Шизомби 21:39, 17 February 2007 (UTC)

Some people number subheadings, for example "section one point two". I just read them as they are, with pauses before and after to distinguish them from the rest of the text. It's up to you. -- Macropode 22:09, 17 February 2007 (UTC)

The recording guidelines will be modified to include information like this sometime soon. -- Macropode 00:37, 18 February 2007 (UTC)

[edit] Quality assessment categories?

Regarding the new article review section, is there any interest in having categories for spoken articles based on their rating? There are 3 evaluation criterea (technical quality, clarity, accuracy) and 3 ratings for each criterea (corresponding to 0, 1, 2). Thus there would be 7 categories corresponding to the possible sums of 0-6.

• If there is more than one review, the category is determined by the average of the reviews
• If a new version of the article is recorded, the reviews pertaining to the old version will be 'turned off' (they won't count toward the average)

I have an idea of how to write the template to automatically sort the articles (see User:SCEhardt/temp), but since I've run into a snag I thought I would check for interest here before I work on it more. Hope this makes sense :-) -SCEhardT 21:05, 19 February 2007 (UTC)

Perfectly, in both senses! I think this is an excellent idea, but one that would only really be simple enough in operation to be practical if it were automated. Please continue! -- Macropode 01:58, 20 February 2007 (UTC)

How will blank fields (i.e. un-rated evaluation criteria) be dealt with? -- Macropode 06:31, 21 February 2007 (UTC)

Sadly, this isn't going to work out. I can't figure out how to get multiple reviews to average in a manner that even comes close to being efficient or easy for other users to understand. -SCEhardT 11:20, 22 February 2007 (UTC)

[edit] Quality assessment refinement

I've got a review ready to post , but have run into a minor problem. The old boilerplate allowed the insertion of comments underneath each criteria heading briefly explaining why that particular criteria was given the rating that it was. Is there any way of setting up the new template to allow this functionality? The alternative would probably be to duplicate the criteria headings and put them and the comments underneath the template, which is a bit of a clunky solution. -- Macropode 06:27, 22 February 2007 (UTC)

If I understand what you're asking, the new template does allow this. Check out this, for example. -SCEhardT 11:24, 22 February 2007 (UTC)

I should have found that out for myself, shouldn't I. Thanks for dispelling a bit more of my ignorance. -- Macropode 12:41, 22 February 2007 (UTC)

I'm working on detailed definitions of the ratings. I need to know if it's acceptable to include things like sound clips or samples, and music in recordings. If so, they'd obviously have to be used appropriately and properly licensed. Opinions, please. -- Macropode 07:12, 28 February 2007 (UTC)

I think it's a great idea (actually it's an idea I've had myself), and so I've had the time to give it some thought. As written articles include pictures, so spoken articles should include sound clips. I would say that short clips (maybe under about 1 minute) would be placed at the relevant point in the text, while longer clips would fit best at the end of the article. Anything over about five minutes probably doesn't belong in an article (one can always give an excerpt instead). We should give "basic" licensing info for each sound clip on the recording, with a reference to where details can be found online; this "credits" section can go at the end of the recording. We should be especially careful about fair use clips, and explicitly state that they are copyrighted and only being used in an educational context.--Pharos 07:53, 28 February 2007 (UTC)

Okay, so barring any opinions to the contrary, I'll work on the assumption that it's allowable, although I'm not inclined to encourage contributors to do it, at least at this stage. This project works fairly well, but has historically been lacking two important things: recording/uploading guidelines that are clear and understandable to non-technically inclined contributors, and an organised method of providing feedback to contributors and maintaining a quality standard for spoken articles. Having audio clips in spoken articles would complicate the implementation of solutions to both these issues, and would mean more ongoing work in the assessment process, due to the need to verify the copyright status of included audio. The horrible grey area of fair use only complicates things further.

There's certainly potential for future development in areas like this, but the small number of people we have working on the administrative/maintenance side of this project dictates that we should address the most pressing problems first, and I'd like to focus on implementing basic, workable solutions to the documentation and quality standard issues, at least in the short term. -- Macropode 07:22, 1 March 2007 (UTC)

Fair enough, that's a reasonable approach considering that the project is still at a relatively basic stage.--Pharos 14:50, 1 March 2007 (UTC)

Can anyone suggest a reasonable method of doing assessments on spoken articles that have been uploaded to Commons, other than just re-creating the template etc. over there? That's the only solution my currently sleep-deprived mind can come up with at the moment, and it's not a good one. -- Macropode 02:44, 5 March 2007 (UTC)

There's nothing stopping you from just putting the template on the local Wikipedia talk page.--Pharos 02:57, 5 March 2007 (UTC)

Thanks Pharos. I've already done that once but wasn't sure that it was the right thing to do. When you create the talk page, there's a little message asking you to verify that the (Wikipedia) parent page exists; technically it doesn't, as it's just some kind of link to the Commons page. I'll ignore the message and continue learning. -- Macropode 04:18, 5 March 2007 (UTC)

[edit] Main page featured articles

Should we be attaching spoken versions to featured articles in anticipation of them appearing on the main page, as currently recommended by the Article choice guidelines? Spoken articles like Image:Shadow.ogg, which was uploaded and linked to the article Shadow of the Colossus shortly before it went to the main page, would seem to bear out the concerns of Pharos. While this recording is fine as a novelty, it doesn't meet several of the spoken article recording guidelines, and isn't appropriate to be attached to a featured article that is being showcased on the main page as the best that Wikipedia has to offer. There have been instances in the past of main page articles with problematic spoken versions.

My own opinion is that this project should not be actively seeking publicity, but that it should earn its place in Wikipedia through the quality of the work of its contributors (which, of late, has been consistently high). This does not exclude, by any means, people who lack some of the skills necessary for the production of good spoken recordings from participating. Everybody has to start somewhere, and this project provides the framework to enable that, for those who are determined enough. There may be a case, however, for the application of a minimum quality standard to spoken versions of featured articles, to ensure that the spoken version matches the standard of the source text version before it's linked to it. The spoken article assessment system currently being implemented could be used for this.

This is an important issue that needs to be addressed, so I'd encourage anybody with an interest in this project (and you wouldn't be reading this if you didn't, would you?) to voice their opinion. -- Macropode 12:33, 22 February 2007 (UTC)

I think spoken articles should be held to a minimum standard to be placed in any article, not just featured articles. I agree that the spoken article review system would be good to use for this. Ideally, any article that has one or more scores of 'low' would be removed from the article until the issue is addressed. The scoring criteria would need to be written in a way that makes this reasonable. We should encourage the creator of any article receiving a 'low' rating to improve the article; this system should not act as a deterrent to anyone. At the same time, I think some system for removing poorly spoken articles should be in place to keep the output of this project up to standards. -SCEhardT 21:57, 22 February 2007 (UTC)

Image:Shadow.ogg... Wow, it's certainly hard to imagine a more embarrassing spoken article than that (our first audio WP:BJAODN?). While something of such a nature very obviously doesn't belong on any article, there are always going to be ambiguous cases of sincere efforts that fall short in various ways. We should have basic standards for all spoken articles, and more than that, it might be healthy to have a spoken FA standard that's just a step above (nothing extraordinary) what's considered minimal for the others, especially if all the stars on Wikipedia:Spoken articles are gonna have any meaning.--Pharos 08:22, 23 February 2007 (UTC)

[edit] Request for brief feedback - The Four Stages of Cruelty

Hi. I've just started working on a recording of The Four Stages of Cruelty. Since this is my first recording, I ran off a quick test Image:4 Stages test.ogg. It would be useful to get a bit of feedback before I continue. Many thanks. I realise I may have spoken more quickly than the guidelines recommend, but have over-enunciated to compensate and wonder, in particular, whether the overall effect is still considered suitable. GDallimore (Talk) 00:34, 2 March 2007 (UTC)

The over-enunciation helps greatly for comprehension. Overall, I think it's pretty good, but don't ask me, I'm embarking on my first spoken article just now too. — Edward Z. Yang^(Talk) 01:10, 2 March 2007 (UTC)

I heard one or two minor breath-on-microphone "pops" near the end, so you might want to experiment with your microphone positioning a bit. Other than that, it sounded excellent. -- Macropode 05:56, 2 March 2007 (UTC)

I forgot to mention - it's encoded at a nominal bitrate of 84 kb/s, a wee bit high. Until I get a chance to work on the guidelines, look here for details on how to fix it if you're using Audacity, and a rough explanation of why it's important. -- Macropode 08:56, 2 March 2007 (UTC)

Thanks for the comments, guys. Macropode, I know what you mean about the breath pops - I'm more used to recording music and sound effects, so am still playing with the positioning. I think I'll have to make a screen.

More importantly, though, I wanted your advice on the bitrate. The stuff I do normally has to be uncompressed or CD quality (I do sound/music for my local theatre) so I don't really understand compression. My Ogg encoder gives the options of either using a "quality setting" or setting the max and min bitrates for VBR encoding. I could crank down the quality to 1, which should have the desired effect, but was wondering if you thought that setting a VBR range between, say 36kb/s and 52kb/s (or some other range) would give a better result than this blanket low quality. The test recording was done with a VBR range of 46 to 96, which explains the high nominal bitrate. How did you work that out, by the way? Was it as simple as dividing the file size by the time, or does Audacity tell you these things? GDallimore (Talk) 10:19, 2 March 2007 (UTC)

I've had good results from a home-made screen. If you want that particular sound you get from working close to the microphone, it's probably worth putting the effort into, otherwise just back off the mic a little bit or speak slightly across it.

Sorry, I shouldn't be referring to encoding as "compression". It creates confusion with the entirely separate process of increasing the average amplitude of the signal to make it sound louder.

You can't, unfortunately, get around the file size issue by manipulating the min and max bitrate settings. It still comes down to a question of average bitrate. To increase "quality", you need a higher average bitrate, which results in a bigger file. There's no way around it. The term "Quality" is something of a misnomer here, as it depends on what kind of signal you're dealing with. A higher "Quality" (average bitrate) setting will produce a file encoded with more "space" to allow a signal with a broader range of frequencies, like music or sound effects, to be accurately reproduced. Increasing the bitrate past a certain point will provide more leeway to accommodate a frequency range that your signal just doesn't have in it. That point, for reasonable voice in Ogg Vorbis, is something not very far above 48 kb/s. It's not "broadcast quality" by any means, but produces quite clean voice that doesn't take ages to download, and that minimises the need to break the recording up into separate pieces for large articles. In my limited experience, listeners generally don't like the extra complication in downloading multi-part recordings. I've had good results encoding at 48 kb/s. You should also consider that most people will be listening to these recordings through computer speakers, most of which range from being fairly average to downright awful in quality.

There's no arcane magic involved in determining the parameters an audio file has been encoded with. I use Linux with the KDE desktop (specifically, Kubuntu Linux with the Kubuntu default KDE modifications disabled, because I don't like them). Right-mouse-click on the audio file. From the menu, select "Properties". In the info box, select the "Meta info" tab. This shows some useful information about the file. I haven't checked, but I'd guess it's probably extracting most of that information from the file's header. There's some kudos waiting for the person who can chip in here and describe how this information can be accessed easily in Windows, without having to install extra software, as unfortunately I have no working knowledge of any recent version of Windows.

Meanwhile, GDallimore, you might like to re-encode your test file at a lower "quality" setting and use the "upload a new version of this file" link in Image:4 Stages test.ogg so we can listen to the difference and check the bitrate. If the cold hasn't grabbed you too hard by now, that is. :) -- Macropode 10:18, 3 March 2007 (UTC)

That approach doesn't seem to work in Windows. At least not for me as I haven't downloaded any OGG codecs. As for file bitrate, rather than fiddling with the VBR settings, I just cranked the quality setting down to the minimum. Doesn't sound too bad and is about half the size of the previous version - size of file / time gives a 44kb/s bitrate.

Cold has now killed my voice, so I'll be waiting a week or so before recording any more of the article. :( GDallimore (Talk) 19:02, 3 March 2007 (UTC)

'Scuse me for taking so long to get back. The later version's showing nominal 48 kb/s, average 45 kb/s, sounds good, and the file size, as you can see, is just over half of the original. Whip that cold, man! -- Macropode 03:04, 8 March 2007 (UTC)

No problem and thanks for all the useful feedback and support! GDallimore (Talk) 14:06, 9 March 2007 (UTC)

[edit] No Guidelines?

Hey, can you direct me to some recording guidelines? - I have a few questions.
Let's say I wanted to record a good article that has a nice big introduction, an info box, a table of contents, and then all the sections.
Should I read the contents of the infobox, and if so when, and if so should I introduce it by saying "Here is a brief summary of factual information about [Article Name]?"
With the contents, should I read out the contents perhaps after I've read out the introduction, for example:

1. Read introduction
2. Introduce and read Infobox
3. Read out Contents
4. Begin reading article.

How does one deal with things like mathematical formulas, images, etc? There may be cases where the article refers to a picture - there should probably be a guideline that a reader mustn't describe a picture - unless it is absolutely necessary. What if the article's text actually refers to a picture, as such: "Below is a picture of a rainbow showing the spectrum of colours" - after that should the reader just say "Picture" or "Editorial: Picture".

If there are no guidelines, then you're going to have a lot of inconsistancy, and mistakes.

I am more than happy to begin writing a set of guidelines -- please first tell me if there isn't one already.

Rfwoolf 00:45, 4 March 2007 (UTC)

Here's an example of where a reader might want to slightly paraphrase (from Afrikaans):
Similarly original qu and x are spelt kw and ks respectively. For example ekwatoriaal instead of "equatoriaal" and ekskuus instead of "excuus".
Might be read as:
Similarly, original QU and X are spelt KW and KS respsectively. For example the word "Ekwatoriaal" spelt E-K-W-A-T-O-R-I-A-A-L (that is with an "KW" for the "qwa" sound) instead of being spelt E-Q-U-A-T-O-R-I-A-A-L (that is with a "QU" for the "qwa" sound), and, "ekskuus" spelt E-K-S-K-U-U-S (that is with "KS" for the "X" sound) instead of "excuus" spelt E-X-C-U-U-S (that is with a "XC" for the "X" sound). Rfwoolf 01:15, 4 March 2007 (UTC)

Just thinking again about the infobox, for articles that have lengthy info boxes, it might be an idea to have them read last in the article -- and that the reader should decide this when he records the article. Rfwoolf 01:26, 4 March 2007 (UTC)

• I've created Reading Guidlines

[edit] Jake Gyllenhaal

I've just finished my first recording, Image:JakeGyllenhaal.ogg. Would someone (preferably English) mind listening to it and letting me know what my accent is? Dev920 (Have a nice day!) 15:06, 4 March 2007 (UTC)

[edit] Placement of links to Spoken Version of Articles

I am curious to know the reasoning behind the placement of the audio links? Why are they at the bottom of the article alongside the external links section? I think this topic is something that should definitely be discussed and perhaps changed.

To me this seems like an illogical spot to place the link for this as one has to scroll through the entire article to find out if there is an audio version available or not. It would seem like it would make more sense to place the link at the very top of the article above the introduction, thus letting people know prior to reading the text version that an audio version does exist. I also think it would be significantly easier for visually impaired people to locate the audio version.

It also seems odd to me that these are placed under the external links as I don’t believe they really are as the info page for the file is on Wikipedia itself, and the actual audio files I believe are hosted on Wikimedia along with images et cetera. An image is usually plunked in the relevant section of the article, not at the very bottom. It makes sense to me to put links to other projects such as Wikiquotes as this is linking to further reading or other related media. The audio version is of the actual article itself and not something that is related that should be at the bottom. Links to versions of the same article in other languages are given more prominence on the article pages, and most of the time they are in different stages of completion and are dramatically different then the English version. I think defiantly placing these links at the top of the article is definitely the way to do it.

I am very new to editing on Wikipedia, I hope I have posted this in the right place for discussion.

Leeuwekoning 10:02, 7 March 2007 (UTC)

Hi. If you look carefully (it's not very prominent) there's a little icon at the top right corner of a spoken article so you can immediately see on opening any article whether it is spoken or not. I'm not sure of the reasoning for putting the main template in the external links section, though - just joined this project myself and learning my way around, but then there are always new things to learn, I'm finding!

Oh, and a quick word of advice on editing Wikipedia: it's easiest to start a new conversation topic on a talk page by cliking the "+" tab at the top of the page. That adds your new discussion to the end of the page, which is the preferred format. GDallimore (Talk) 11:53, 7 March 2007 (UTC)

See archived discussion -SCEhardT 14:24, 7 March 2007 (UTC)

[edit] Sung wikipedia?

I've an interest in traditional (out of copyright) songs, and have recently been looking over a few articles such as Miller of Dee, Greensleeves and Heart of Oak. All great songs, but little to say about them, so the articles are a bit stubby and probably not worth recording. However, I was wondering if there was room in this project to record sung versions of these old songs to help readers who may not know the tune? It's easy enough to make instrumental midi or even wav files of the tune, but you never get the same feel as from actually hearing someone singing.

I'd be interested in getting involved on such a project on a couple of levels. I'm not a great singer, but can hold a tune, so could probably record some songs myself. Also, I'm in a position to produce backing tracks for others to add their singing to karaoke style.

What do people think? Would it be possible for Wikipedia to release their first music album? :0 GDallimore (Talk) 14:18, 9 March 2007 (UTC)

There are already some efforts going on in this vein. User:Makemi, for example, has done a number of recordings of old songs (she helped me recently with Tippecanoe and Tyler too). You could ask her about recording advice. There are also several instrumental musicians who've contributed to Wikipedia. See also Wikipedia:Instruments and Wikipedia:Requested recordings--Pharos 19:09, 9 March 2007 (UTC)

[edit] A little change

I gave our main Wikiproject page a yellow background and a speaker image in an attempt to make it more appealing to potential participants who visit the page. Let me know what you think! We can always get rid of it if everyone hates it. =P  Panser Born  (talk) 17:57, 13 March 2007 (UTC)

The speaker image is nice, but is the yellow background really necessary? — Edward Z. Yang^(Talk) 21:39, 13 March 2007 (UTC)

Not particularly, no! I just thought it helped the page stand out from the multitude of other grey pages. Thanks for the feedback anyway, hopefully some other people will have their say too.  Panser Born  (talk) 07:34, 14 March 2007 (UTC)

Looks good to me. Nice work cleaning up the "in progress" table, too. -- Macropode 22:46, 14 March 2007 (UTC)

Yep - a clear and attractive new image. Nice work! Hassocks5489 13:28, 20 March 2007 (UTC)

[edit] Audio problems (question from MgM)

I can't help MgM with this speech file-related query, so I have copied the question across from my talk page: (Hassocks5489 08:52, 21 March 2007 (UTC))

"I've recorded two audio pronunciations. One I did in the standard Windows sound recorder before converting it to .ogg with an online utility. The other was recorded and exported from Audacity. Do you have any idea why my VLC Media Player refuses to play the Audacity-created file? Did I do something wrong when I exported it?"

There are too many possibilities to speculate about here. You might like to upload the Audacity recording, or a short test recording made in Audacity with the same settings so we can have a look at it. When uploading, call it something like Spoken_test_recording.ogg so we can re-use the image page for others. -- Macropode 09:12, 22 March 2007 (UTC)