Wikipedia talk:Featured sound criteria

From Wikipedia, the free encyclopedia

Contents

[edit] Not clear

The criteria on the main page completely fail to explain what makes a sound in this case. Is a spoken article version considered a "sound"? How about a pronunciation? Or audio recordings of an event? -Amark moo! 21:26, 28 January 2007 (UTC)

While I'm not sure I agree with this, it is now clear, at least. -Amark moo! 22:24, 28 January 2007 (UTC)

In response to your crossed out comment, spoken articles should not be candidates, nor should pronunciations. In my opinion, they wouldn't have any significance, really. As for an audio recording of an event, if it falls under the "significant" criteria, then sure. Those that wouldn't may include a wonderful (but uneventful) commentary of the woman's Olympic team table tennis event in 2008. Unless it has (or will have) historical significance, then it lacks importance and therefore should not be a FS. └Jared┘┌talk┐  22:45, 28 January 2007 (UTC)
  • The "historical, social, or current significance" criterion is still unclear to me. Exactly what kind of significance do you mean? --KFP (talk | contribs) 23:14, 28 January 2007 (UTC)
  • Perhaps what I am suggesting is redundant, but here's what I think: a great sound of a walrus making whatever noise they make--without the "significance" clause in effect--could potentially get into FS. Now, some people may think this is OK, but myself, I believe that the sound, as nice as it is, could not possibly be "1 in a million." In other words, I could go to my local zoo with a high quality recording system and prod a walrus until it makes a noise, and upload it to Wikipedia. The sound would be similar to the first, but still not significant. See what I mean. On the other hand, a famous speech, or a special performance, or a sound that is unique in some way should qualify. FSs are just like FPs because you wouldn't want something that's so-so. You want something that could not easily be recreated, or something that has some significance in history/society. But others say that this is already a "built-in feature" of wikipedia in that things here should have some significance, but I think it wouldn't hurt to accentuate this fact. └Jared┘┌talk┐  23:24, 28 January 2007 (UTC)
  • I think it would be far better to require the sound to be encyclopedic, without this very fuzzy "significance" requirement. Mak (talk) 23:32, 28 January 2007 (UTC)
  • It seems that that may be far better. Because using the word "significance" means something totally different to multiple people, while encyclopedic is more defined. But encyclopedic has significance built-in, sort of. └Jared┘┌talk┐  23:41, 28 January 2007 (UTC)
  • Well, anyone could also go to their local zoo and take a picture of the walrus instead. That would certainly not guarantee the photo a place on FP today. The key thing to remember here is, evolving standards. In the future, an FS of a walrus may well be a sophisticated radio piece explaining their whole system of communication.--Pharos 01:47, 29 January 2007 (UTC)
  • Keeping with the ways of FPC, I think the criteria should emphasize high quality and encyclopedic value. This means that rare or historical subjects can be of lower quality, and more common subjects should be exceptional, i.e., a scratchy recording of a famous speech and a clear recording of a walrus could both be made featured. These basic criteria should be the starting points, and as the ball gets rolling more specific requirements may be added. But trying to enforce strict requirements this early in the game isn't going to attract possible candidates. Start small. --Tewy 01:52, 29 January 2007 (UTC)
  • I totally agree - if someone does a) happen to have some high-quality recording equipment they don't mind taking to a zoo, b) manage to get the zoo's permission to take it in, c) manage to get a walrus to make some characteristic noises without poking it and then d) decide to upload their high-quality, encyclopaedic recording here under a free license, then it would probably make a very good Featured Sound - after all, the particular walrus doing the talking doesn't need to be "notable" for a recording to be highly encyclopaedic. I doubt we'll find ourselves in a situation where we have a vast excess of walrus-noise nominations anytime soon, though. I expect that, if we make the criteria relatively flexible to start with, then we'll encourage plenty of nominations and can see where consensus takes us in terms of tightening up the requirements (and potentially delisting any early noms which no longer seem to make the grade) as we go along. --YFB ¿ 02:57, 29 January 2007 (UTC)

[edit] Description

I think that for right now the description requirement is a bit too harsh. I think it's good for the image discription page to describe what the piece was performed from (so that we can be sure that the performance can be freely licensed), if music is in question, but it's quite likely that someone could perform from a not-yet-digitized public domain score, and I don't think we should necessarily expect the uploader to digitize it. Also, I'm confused by "sufficient to explain the contents of the file for users who cannot play or hear it." I don't think I could give the entire meaning of a music file in prose, nor do I think an uploader should. Perhaps "A descriptive caption" should be the requirement for now, with more specifics when we run across them? Mak (talk) 00:46, 30 January 2007 (UTC)

I agree, although it would be very useful to have digitised music scores available if they exist. Perhaps that could be suggested but not stipulated as a requirement? I think "a descriptive caption" is sufficient for now - going back to the walrus example, it'd be difficult to describe the contents in any more useful detail than "typical vocalisations of the male walrus in confrontation with rivals" or whatever... perhaps some attempt could be made to characterise the sound (e.g. "a deep, breathy grunting noise") but obviously that wouldn't be possible for a piece of music. I think it'd be better to leave that requirement quite open, and for people to make whatever improvements they think are appropriate during the nomination process. --YFB ¿ 00:59, 30 January 2007 (UTC)
(Edit conflict) Yes, I saw the "sufficient to explain..." line and wondered if it was really necessary. The simple description, "three violins and cello playing a rare rendition of..." or similar, would be fine to at least inform users what to expect. Any details would probably be in the article. --Tewy 01:06, 30 January 2007 (UTC)
OK, I'll agree to hold off on a strict score requirement, for say, three months, when things mature and I can master GNU LilyPond to do it myself. But even if we don't have the score on Wikipedia, we should make every effort to link to the score if it exists somewhere else online. I agree with YFB and Tewy re: the depth of captions, at least for now.--Pharos 01:14, 30 January 2007 (UTC)

[edit] Moved from candidates page

The text below from WP:FSC was essentially covering 'criteria'. To cut down the space of the intro I have removed it to this page.

Featured sounds may be music, "sound portrait"-type illustrations of articles, or possibly historical recordings, but Spoken Wikipedia articles are not appropriate as featured sounds as they are basically text-based (though of course they would benefit by the incorporation of these sounds).

Sounds listed here should be either in the public domain or covered by the GNU Free Documentation License or a similar license. Since a sound archive is of limited educational value (a requirement for fair use) fair use sounds are not appropriate candidates for inclusion in the featured sounds gallery.

In general, recordings should be of high fidelity and technical quality. However, exceptions can always be made for recording taken under extenuating circumstances.

It looks like we cover most of these conditions in the existing criteria, but some of this wording might be used in place of what we have currently. --CBD 12:37, 31 January 2007 (UTC)

I originally wrote this stuff on the model of the Featured Pictures candidates page back in 2005, when that page itself was somewhat more elaborate. I think most of the current criteria we've worked on recently are probably more specific and medium-relevant. I guess the most important thing to be incorporated on criteria now would be the Spoken Wikipedia thing.--Pharos 13:40, 31 January 2007 (UTC)

[edit] Question

What is a sound portrait? >Radiant< 14:51, 31 January 2007 (UTC)

  • It was the best term I could think of for a sonic "illustration"; i.e. a recording of the ambient environment– a bird singing, a locomotive running, an avalanche crashing. The term is used somewhat regularly on National Public Radio in the US. If you know of a better term, please share it.--Pharos 15:09, 31 January 2007 (UTC)
    • I suppose it's a reasonable term but you should probably either add a description on the main page here, or create a brief article on it and link to there. >Radiant< 15:18, 31 January 2007 (UTC)
      • Hmm... on second thought, do you think perhaps I should replace this with field recording? It's a more academic term, anyway.--Pharos 15:34, 31 January 2007 (UTC)
        • If google is any judge, yes, that'd be a good idea. >Radiant< 15:36, 31 January 2007 (UTC)
          • My understanding, and the article seems to bear it out, is that a "field recording" is generally used for recordings of people, in the field of ethnomusicology, mainly. I have not heard it used for, for instance, birdsong, so "sound portrait" may be more appropriate, as it could include field recordings and other recordings of non-musical sound. Mak (talk) 15:48, 31 January 2007 (UTC)
            • After a little research and a little reflection, it seems to me that the term "field recording" is primarily used in academic contexts; i.e. both Alan Lomax-type stuff and biologists' recordings of animal sounds. "Sound portrait" may be more common in certain radio contexts, especially for some more purely evocative sounds (there is generally little scientific demand to do recordings of locomotive engines). "Field recording", from my very cursory examination, also appears to be a popular term of art among internet hobbyists (potential contributors). On balance, I'm leaning toward some qualification like "environmental field recording", but we should improve field recording on the more general definition in any case.--Pharos 17:24, 31 January 2007 (UTC)

[edit] My recent copy-edit of the criteria

  • I think "Featured sound" is an awkward title for musical recordings.
  • I've separated caption and the info page into 6 and 7, since they're quite different issues.
  • Criterion 5: Accuracy of a sound? Doesn't work. This needs to be reworded.
  • Criterion 7: "gives a sense of the file for users who cannot play or hear it". Is this practicable or achievable for a musical work? Why is a linguistic description of the actual music required? If it's going to be required, it needs to specify what aspects of the music/recording need to be included: frequency response? Reverb? Intonation and rhythmic precision of the players? The macro-form of the musical work? I'm surprised that more useful information isn't required, such as recordists, date and venue of recording, performers, composer and years. No mention that if a musical score/transcript is linked to (as required where a digital form exists), it should be free. There should be a requirement that the info be well-written in clear, plain language. Tony 11:02, 4 May 2007 (UTC)
  • I suggest another criterion: that the file be of appropriate length (for non-musical recordings). See the current nomination of Pseudacris-crucifer-003.ogg. Tony 11:13, 4 May 2007 (UTC)
  • Should there be a requirement that the item be used in at least one WP article? Tony 11:32, 4 May 2007 (UTC)
  • Another suggestion: if a recording has been edited for the purpose of its use on WP, this should be noted. For example, if the noise levels on an historical recording of a speech are reduced to make it more intelligible/easier to listen to, this should be noted on the info page. I'd support a requirement that the duration be given in the caption. Tony 11:41, 4 May 2007 (UTC)
  • Yes, it's an awkward title, but this process is not only for musical recordings, but also for field recordings and speeches.
  • "Accuracy of sound"... I'm not sure how to fix that without making it an incredibly long and involved guideline. The idea is that if you're recording a Chopin Étude, it should have the right notes, and should preferably be played on a piano, not a harpsichord. I'm sure there are more ways a recording to be inaccurate that you can think of (really out of tune... rhythmically ambiguous, etc.)
  • I thought we had rejected that as 7, but anyway. I think the image description page should give important information such as performer, time of recording, composer, specific piece, possibly edition. If it is an electronically created file perhaps the method of creation. I'm not sure how much other specific recording information should be required. It might be nice to put together a template, similar to that which was used on Commons for uploads, including fields for what is decided to be necessary.
  • Recordings should definitely be used to enhance an article, in my view.
  • I definitely think changes to historic recordings should be noted, but I'm not sure whether every single edit to a standard recording should be noted. Mak (talk) 21:38, 7 May 2007 (UTC)

[edit] Revamp of criteria

Here's a quick job. Please provide feedback below. I propose that the criteria be upgraded in, say, a week, to allow consensus to form. Tony 01:58, 8 May 2007 (UTC)


A featured sound is a recording of a musical performance, an environmental field recording or a voice recording that exemplifies our very best work. In addition to the requirements for all sound files, it has the following attributes.

  1. Public domain / free license. The recording and its subject are available in the public domain or under a free license. Non-free recordings and musical performances that use texts under fair use are ineligible.
  2. Recording quality. The recording is of high fidelity and is free of technical faults, such as unintended noise and distortion, and sonic and compression artifacts. The balance, reverberation, frequency response and stereophony are of a high standard. Musical performances are of a high technical and artistic standard.
  3. Role. The file helps readers to understand the topic of an article. The file is used in at least one article, where it is supported by the text and, if appropriate, reference citations on the file description page.
  4. Caption. The file is displayed with an informative, well-written caption, including the duration of the file.
  5. Sound description page. The page contains an extended description of the file, including:
    • (i) a description of the subject;
    • (ii) the duration of the file;
    • (iii) the date and venue of the recording, where they are recoverable;
    • (iv) the name(s) of the recordist/producer, and for historical and field recordings, a brief description of the recording equipment, where known;
    • (v) any editing that has been applied to the excerpt by the uploader, such as noise reduction by the uploader (aside from obvious fade ups and fade downs at the start and end), and by others, where that information is recoverable;
    • (vi) for a musical performance, the name and years of the composer (and the arranger, where relevant), the year of composition (and the arrangement, where relevant), the name(s) of the performer(s) or, for an ensemble, the name of the group and, where relevant, the conductor and soloist;
    • (vii) for a musical performance, links to a musical score in digital format where available;
    • (viii) for a voice recording of linguistic text, unless inappropriate, a transcript or a link to such a transcript; and
    • (ix) where a recording has been released into the public domain or has been transferred to a free license, the name of the original owner and the date of release/transfer, if this information is recoverable.

Exceptions to Criterion 2 may be made for historical recordings, musical recordings, and recordings made under unusual or extenuating circumstances. Spoken Wikipedia readings of articles are ineligible for nomination.

___________________________________________

Link to compare with the existing version

Suggestions please:

One issue for musical recordings is that the performance, music itself, and edition, must all be free. The edition question can be a bit flexible I think, if it is not something where there is significant editorial input, but we should be cautious about it. Also, people tend to be clueless about the issue of performance copyright vs. composer's copyright. I don't know, however, whether adding that to the criteria will help the already clueless. I think the issue of editions is more likely to confuse otherwise sensible people.
Also, I worry a little about the "merely ornamental" wording. It's not clear what it's excluding, and I feel like it could exclude some very worthy and encyclopedic content. For a musical piece, where it's possible, I think adding a recording of that piece where possible is great and should be commended. I can't think of a better for Britten's Hymn to St. Cecilia than a recording of that piece, or excerpts from it. I think it's important that the recording add to an article, but could you clarify your intent? Mak (talk) 03:37, 8 May 2007 (UTC)
I have no problem removing the "ornamental" bit. (Done.) Usually it's impossible to know what edition of the music is used, so it may be better not to mention it. I think it's covered by the current wording, anyway. Tony 03:48, 8 May 2007 (UTC)
I do think the edition question deserves at least some mention, at least when, as you say, there's "significant editorial input". Especially since this is an issue we can expect most contributors to be fairly ignorant about.--Pharos 23:00, 9 May 2007 (UTC)
  • As for the technical quality criterion, what if the sound is historical? Like the Edison Phonograph sound? Perhaps it should be "in relation to the era it was recorded"? — Deckiller 02:35, 9 May 2007 (UTC)
Do you think the final paragraph sufficiently covers this concern? Tony 02:42, 9 May 2007 (UTC)
I think it's sufficiently clear with the last paragraph. Mak (talk) 02:53, 9 May 2007 (UTC)
  • It seems to me that Technical Quality and Musical Performance could be merged. I also suggest using "non-free" instead of "fair use". Pagrashtak 04:12, 9 May 2007 (UTC)
Oops, I changed the piped link. But "fair use", I think, should stay as the title of the legal doctrine at issue, shouldn't it? See Wikipedia:Non-free content.
I merged the musical performance and technical quality criteria, thought about it, and then reverted. I'm unsure, since they're such different things: musical performance involves more than technical considerations—it's an expression of cultural meaning, and some people would argue that artistic as well as technical values are at issue. Sure, artistic values can be at stake in balance and frequency response, but these recordists'/producers' tasks appear to be on opposite sides of the technical–artistic continuum from those that encompass the musicians' role. It's no big deal, so if people feel they're better merged, no problem. Anyone else got an opinion?
I've made a few trivial edits. Tony 04:44, 9 May 2007 (UTC)
They would both fall under "quality", but it's quality coming from two different angles. I do think they should probably be merged, though, maybe as (a) and (b) subsections. In any case, it must be noted that the musical performance criterion would of course not apply to non-musical recordings.--Pharos 22:54, 9 May 2007 (UTC)
Changed title of Cr. 2, merged musical performance into it, and explicated your last point. How's that now? Tony 23:09, 9 May 2007 (UTC)
Yes, I like that better. Thanks.--Pharos 23:12, 9 May 2007 (UTC)
Agreed. Suggest replacing "In a recording of a musical performance, that performance is of a high technical and artistic standards." with "Musical performances should be of high technical and artistic standards." Pagrashtak 23:44, 9 May 2007 (UTC)
I've replaced it as you suggest. That was my first inclination, but I avoided it because it involves the sole pluralised form in all of the criteria. The less attractive previous wording at least retained the singular. Tony 10:18, 13 May 2007 (UTC)
Also, I don't think I like "featured sound file" too much. Why shouldn't we have a parallel name with "featured picture"?--Pharos 23:22, 9 May 2007 (UTC)
  • Should the duration of the file be explicated? As it stands, Criterion 4 insists on it in the caption. Should it be in the info page instead, as well as, or not at all? Tony 23:37, 9 May 2007 (UTC)
    • The image description page should definitely have the duration. The examples at {{Listen}} and {{Listeninbrowser}} don't give duration, but instead file size. On this matter I'll defer to those who work with sounds more often. Pagrashtak 23:42, 9 May 2007 (UTC)
      • I agree with Pagra, duration on the image description page is an absolute must. I also think duration should be included in the article captions, for the simple reason that sometimes one has time to listen to a ten-minute recording, and sometimes one doesn't.--Pharos 23:51, 9 May 2007 (UTC)
        • Exactly my thinking. In addition, it's a download capacity issue. Shall adjust wording now, in anticipation of consensus. That makes it nine points, so I changed to lower-case roman numberals to avoid (i) as the ninth one Tony 02:29, 10 May 2007 (UTC)

This just occurred to me—in addition to having the subject and recording under a free license, we should probably require the file to be in a free format such as Ogg. Easily added to criterion 1. Pagrashtak 23:59, 9 May 2007 (UTC)

  • Good idea, unless that's required of all sound files on WP. Is it? These criteria are specifically the extras. Tony 02:35, 10 May 2007 (UTC)
    • It is required of all sound files on WP. Indeed, it's required of all files on WP, including all media files and the underlying software, as this is a free content project.--Pharos 19:20, 10 May 2007 (UTC)

[edit] Additional consideration

I'm not sure where this should be added, but in the case of music files there should be some criteria about the interpretation of the music. If scored, the performance should follow the score according to an established performance practice and the style of that performance practice should be noted. For example, a recording of a Beethoven sonata should be on a modern piano using current performance practices or performed on a period piano using 19th century practices. What we should not allow is creative contributions by the performers, as this would be "original research". --Samuel Wantman 04:10, 12 May 2007 (UTC)

Although I agree with the value judgements behind Sam's comment, I think that including explicit requirements of musical performance—in essence, defining what good performances are—is fraught with difficulties. I thought about adding the simplest, least controversial matters to the criteria (in tune, rhythmically synchronous, good instrumental balance), but soon gave up. This underlines that the review process is only as good as the reviewers; at least some of them should be able to critique the performance aspect. Tony 10:14, 13 May 2007 (UTC)

Certainly we could add a simple criteria that states that "the performance be based on accomplished musicianship and scholarly interpretation and performance practice." Not having that criteria is like saying that articles have to follow the manual of style, look good, be spelled correctly, use good grammar, but they don't have to be accurate! -- Samuel Wantman 07:07, 15 May 2007 (UTC)
So you would insist on a certain type of performance practice? There lies the stuff of pages of debate. It's too tangled a web. That's why I believe that the issue should be covered by a general insistence on high quality. This is analogous to Criterion 1a of the FAC, which states that " the prose is engaging, even brilliant, and of a professional standard". To prescribe certain styles beyond that would lead to mayhem; yet the reviewers, if they're energetic, often manage to provide a certain discipline WRT to writing standards. It's the best we can do. Tony 08:26, 15 May 2007 (UTC)
Can't we leave this vague? Is it too much to say that there should be justification for the performance practice that is based on scholarship? This was the problem I had with the recording of The Entertainer. If we say that these things should be considered, they can be discussed in the context of each nomination. Your proposed criteria do not say that music has to be performed well, based on scholarship and not uniquely original interpretations. --Samuel Wantman 23:16, 15 May 2007 (UTC)
So Bach on the piano would be ruled inadmissable? Or a performance by a modern-instrument orchestra of a Beethoven symphony that is not based on scholarly performance practice, but nevertheless is, in its own frame, of great beauty?
I think performance practice involves artistry and technique as well as scholarship. It can raise subtle questions that are difficult to answer in such a judgemental setting as the FSC room. It's bound to be subjective and to come down to "taste" in some instances. Now don't get me wrong: I'm a performance-practice freak—I prefer all of my early music, right up to the mid-19th century, on original instruments. But here, legislating for a scholarly basis would be a recipe for disputes, and would risk disadvantaging worthy files, I think. Performance practice and musicianship are too entwined and subjective for hard and fast rules, I believe.
If someone nominated a recording of a Telemann chamber work on a hideous synthesised sound, I think there's scope to object on the grounds that it doesn't exemplify "our very best work". Can you provide some examples that you think are undesirable candidates but that might be difficult to argue against on the basis of the new criteria? Tony 23:48, 15 May 2007 (UTC)
You've just replaced what I am asking for with "our very best work". I'm just looking for something that covers the situation we had with The Entertainer. I want to add a criteria that explains what we mean by "our very best work". I don't want Wikipedia to become a repository for quirky unique interpretations. If a Telemann recorder sonata gets performed on a flute, there is a long tradition of making that substitution. It is not "original performance practice", which is analogous to our criteria against "no original research". If someone records Telemann's 12 fantasies on penny-whistle, I will be very impressed, but I'd object. Even Bach on synthesizer would be a stretch, even though it's been done many times. These recordings are meant to be educational. So we should assume that this may be the very first time someone might hear the music of Bach. I'd hope that it was a good performance on original instruments, if not, on modern instruments. I would object to Vivaldi on Kotos, a jazz rendition of the Nutcracker Suite, Bach on syntesizer, etc... even though I have recordings of these that I admire greatly. They are not appropriate for an encyclopedia. -- Sam 00:48, 16 May 2007 (UTC)
But how would you legistlate against this kind of thing? I don't believe that it's practicable. Better to rely on reviewers' assessment of tastefulness, under "our very best work"—that's what I'm suggesting. If you have a better idea, please provide the text of a criterion to cover it. Tony 11:53, 16 May 2007 (UTC)
  • "Musical performances should demonstrate accomplished musicianship and a scholarly interpretation. Performance practice should be based on well-established documented traditions." In other words. No Mozart on bagpipes. -- SamuelWantman 23:55, 16 May 2007 (UTC)
"accomplished musicianship" is implicit in the current wording, that "Musical performances are of a high technical and artistic standard.". "Scholarly interpretation" raises all sorts of hairy issues; a basic one is that it should reasonably apply only where the performance is framed thus. Mozart on bagpipes, if played superbly in artistic and technical terms, could conceivably be a good illustration of the flexibility/adaptability of the instrument (or of Mozart's music). If well supported by the surrounding text, this would be consistent with Criterion 3 "Role. The file helps readers to understand the topic of an article". I'm not being perverse here: I'm arguing that to print a requirement for scholarship and performance practice in black-letter law is potentially (1) overly prescriptive and inflexible, and (2) likely to spark fierce, complicated discourse on these pages. Better to let the reviewers decide case by case, using "exemplifies our very best work", "Musical performances are of a high technical and artistic standard", and "The file helps readers to understand the topic of an article". Can you provide an example where it would be difficult or impossible for reviewers to reject a nomination that is in bad taste, poor in standard, and/or not useful to an article? Tony 01:12, 17 May 2007 (UTC)
  • As advised at the top of the FSC room, I'm implementing the changes. Sam, perhaps this issue can be ironed out in the next few weeks. It would be nice to encounter problematic nominations, which might inform subsequent modifications of the criteria. Tony 23:28, 16 May 2007 (UTC)