< November 18	November 20 >

The following discussion is an archived inquiry of the possible unfree file below. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the media's talk page or in a deletion review). No further edits should be made to this section.

The result of the debate was: Delete; deleted by Explicit (talk · contribs · blocks · protections · deletions · page moves · rights · RfA) AnomieBOT ⚡ 04:03, 2 December 2012 (UTC)[reply]

File:Garbage - Automatic Systemic Habit.png (delete | talk | history | logs).

this looks like it might be unfree. I'm not sure where the thrash hold for a derivative work lies, but it seems derived from the original recording. 217.17.138.246 (talk) 12:49, 19 November 2012 (UTC)[reply]

I created and uploaded the image myself. It was inserted to illustrate points raised by several commentators in the Critical Reception field of the song's parent album. And I don't understand the point raised about it being derived from the original recording. It clearly is an unedited screen grab of the recording as officially released (it would need to be, in order to meet its intended purpose). In any case, an Audacity screen grab of a waveform surely isn't intruding on a band's ability to sell a copy of a CD. Homeostasis07 (talk) 03:02, 20 November 2012 (UTC)[reply]

Yeah, that does sound like a copyright problem to me. (same anon as above btw), I can't quickly find the proper guideline, but the article on Derivative work makes that clear IMO. Weather or not that affects the ability to sell CD's or not is not at all important for the copyright status. What it is important for, would be a fair use claim, which IMO could be justified in this case, and in line with our non-free content criteria. That would require a licence change on the description, and a fair use rationale in its place. 62.194.104.217 (talk) 22:02, 20 November 2012 (UTC)[reply]

A waveform is not a copyrightable piece of work, as it fails Commons:Threshold of originality on all counts. The image also clearly removes any identifiable pieces of content that could interfere or intrude on the company's copyright. Homeostasis07 (talk) 23:28, 20 November 2012 (UTC)[reply]

The first claim (that a waveform doesn't meet the threshold of originality) is problematic because the waveform itself is a representation of the sound of the song. If the sound of the song meets the threshold of originality, then the waveform does too. As If the waveform doesn't meet the threshold, a simple transformation to sound wouldn't either. As a second note in that regard, if the waveform itself doesn't meet the threshold, why would an image of a graph of the waveform meet the threshold? Since you licensed the image, I assume you do believe that the image is copyrightable. As far as the second claim goes: the entire premise of the use of the picture is that it illustrates clipping in the song. The (fragment off) the song is the subject of the image. If it is not identifiable, then how is it valuable. To be clear here, let me repeat, I do think it is a valuable image, and I do think that - in case we decide the image is a derivative work - the image should be used as fair use. 62.194.104.217 (talk) 20:08, 21 November 2012 (UTC)[reply]

Musical copyright = publishing, lyrics, music, the individual instruments and sounds put together to make a recording. No one has ever or would ever be able to copyright a waveform. Homeostasis07 (talk) 00:44, 22 November 2012 (UTC)[reply]

Sound recordings are copyrightable. This is basically just a form of encoding the sound recording. --Stefan2 (talk) 00:46, 22 November 2012 (UTC)[reply]

There is some confusion as to what a waveform actually is here. Waveforms are viewable frequency signals. Nothing more. And certainly nothing to do with encoding. Anyone can record their own voice and view it as a frequency spectrum. Any individual waveform is in no way a copyrightable object. Homeostasis07 (talk) 01:36, 22 November 2012 (UTC)[reply]

There indeed seems to be some confusion as to what a waveform is. Waveforms are not viewable. They are simply descriptions of waves. You can plot them, and you could view the plot. You could play them through an amplifier, then you can hear them (that is, hear the sound that the waveform represents, which is the (fragment of the) song in this case). This is only tangentially related to the spectrum of the sound, which can be obtained through a Fourier transform of the waveform. This image is not an image of a plot of the spectrum of the song, but of a visualisation of the waveform itself. The waveform is certainly copyrightable. A plot of the waveform where there is no possible procedure to regain the original waveform, thus the sound may be copyrightable (hence this discussion). If it could be copyrightable then it is a derivative work, and we should move forward to make it FU. If it's not copyrightable, then the picture may not be copyrightable at all, and should be marked as such (if it's not copyrightable, placing it under CC-BY-SA is nonsense, as there is no copyright). 62.194.104.217 (talk) 17:45, 23 November 2012 (UTC)[reply]

Waveforms are indeed viewable. Viewable descriptions of frequency signals that's often times completely unrelated to any individual song, that is. Clipping is clipping. At this point, you're simply attempting to bog this discussion down with technical jargon that the lay person couldn't understand. The simple fact that a viewable frequency spectrum is uncopyrightable still stands. Homeostasis07 (talk) 03:17, 24 November 2012 (UTC)[reply]

They're not viewable. You can render them viewable as easily as you can render them audible. If it's copyrightable in audibale form, then why not in viewable form?
The image is not of the spectrum of the song. The waveform of the song and the spectrum of the song are very distinct things.
Please don't assume I'm attempting to bog down discussion in some underhanded way to come out on top.
If it is non-copyrightable (which I don't think it is), why do you claim copyright? 62.194.104.217 (talk) 10:54, 24 November 2012 (UTC)[reply]

And here we go again: the basic premise of your argument is that waveforms may be copyrightable. But viewable frequency signals are not copyrightable. Please feel free to resubmit the work as an item for which no copyright can be claimed. Homeostasis07 (talk) 02:19, 25 November 2012 (UTC)[reply]

"viewable frequency signals are not copyrightable."^{[citation needed]} Can you back up this statement somehow, preferably in policy of enwp or commons, or otherwise by precedent in case law, because that is exactly the portion I'm in disagreement on. I haven't found anything on commons that indicates this would be true, nor on en.wp. 62.194.104.217 (talk) 12:29, 25 November 2012 (UTC)[reply]

Here you go: [1] You may need to read the entire paragraph to comprehend the original intention, but it's too long to copy and paste here. But the relevant portion is: "On the other hand, a computer program is defined under the 1976 Act as a "set of statements or instructions to be used directly or indirectly in a computer in order to bring about a certain result." Because [frequency] samples are merely data and do not "bring about" a result, they are not programs under the Copyright Act, no more than are word processing files or musical notation files. Furthermore, the congressional policy regarding sound recordings dictates against a court finding that a digital sample is [comparable] to other digital representations of literary works (such as a word processing file containing a short story.)"

So there you go. Not covered by any Copyright Act legislation, and a court case was explicitly rendered limp by congressional policy. Homeostasis07 (talk) 15:48, 25 November 2012 (UTC)[reply]

I don't think the document you present makes your case very strongly. The document deeply discusses the murky nature of copyright in the use of samples in other pieces, and this situation is quite distinct. The document doesn't assert (at all) that a waveform can't meet the threshold of originality. However, there may be other relevant pieces her. I don't fully understand the relevance of the part where a sample doesn't constitute a computer program, or the relevance of that part on the issue at hand. The part you quote "Furthermore, the congressional policy86 regarding sound recordings dictates against a court finding that a digital sample is analogous to other digital representations of literary works (such as a word processing file containing a short story.)" might be relevant though The note there (note 86) points to "See, e.g., H.R. REP., supra note 57, at 56." but unfortunately I have no idea what that refers to, so I can't research further. Further, I find you are quoting very selectively here. You are leaving out very important other parts:

Under the "genesis test" suggested by Barry, if the data "began life" in the analog world as a pattern of sound waves, the data will be considered a sound recording for copyright purposes.

or

A literal copy of a sound recording must be made in order to constitute infringement. Samples are literal copies.

. None of these support or disprove my point that a the waveform of a sample should be treated like the sample itself. However, I think you and I are not coming to an agreement on this point. Maybe it's better to leave it to others here: I'm not seeing a path forward where we are going to agree on the issue. 62.194.104.217 (talk) 18:26, 25 November 2012 (UTC)[reply]

It is indeed a murky issue. I still contest that a simple waveform of a piece of music is not copyrightable. A waveform can have absolutely no baring on any part of a recording that is copyrightable - melody, vocals, instrumentation, etc. The image of a waveform depends almost entirely on engineering and mastering levels - ASCAP, BPI, etc. do not copyright mastering levels or even make distinction between differing audio formats (CD, vinyl, cassette - if you took a waveform of the same song from these 3 different versions of a release, you would have 3 completely different waveforms). The song itself is copyrightable, its waveform is not. Homeostasis07 (talk) 01:43, 26 November 2012 (UTC)[reply]

I think I'm starting to understand what you are getting at: rather than the waveform itself, you seem to be talking about the envelope. If you are, then I could understand a claim of de Minimis, though I'm not sure if I agree with that. If you were, the talk about the spectrum threw me off completely, as there is nothing to be learned about the spectrum of the original by looking only at the evelope of the waveform, which would explain us talking past eachother. As I think I understand you now, you are comparing this to - for example - taking a picture of a (copyrighted) painting, and presenting a plot of the colors used. I'm entirely unsure if de minimus would be applicable to this. I believe that obviously it should, but courts have had a history of disagreeing with what I think about copyright. 217.17.138.246 (talk) 08:46, 26 November 2012 (UTC)[reply]

De minimis means that the copyrighted work is irrelevant to the image. For example, if you take a photo of a building, and the building contains a copyrighted advertisement, then the advertisement is de minimis because the interesting thing is the building, not the advertisement. However, in this case, the waveform (i.e. a part of the copyrighted work) is the main purpose of the image, so the waveform can't be de minimis. Also, the waveform fails WP:NFCC#8, so fair use doesn't apply. --Stefan2 (talk) 10:46, 26 November 2012 (UTC)[reply]

My reasoning for de minimis is as follows: the work is a transformative work from the original song. The transformation is such, that the original is no longer derivable from the new work, nor is it easily recognisable. This is especially true because through the clipping illustrated in the image, information is irreversably lost, making it less distinguishable from another song with heavy clipping. On the other hand, the use of clipping is an artistic choice, so part of the original work. I'm on the fence on that one. Concering the NFCC, I do find that the image facilitates the understanding of the amount of clipping greatly. The problem with NFCC8 is the phrasing "significantly increase": how much is significant? Wikipedia has traditionally allowed some leeway with criterium 8, for example, when using album or book covers as primary means of identification. Can one really say that an image of a book cover contributes more to the understanding of the book than this image does to the clipping in the audio? I for one wouldn't think so. btw, still same annon 217.17.138.246 (talk) 12:32, 26 November 2012 (UTC)[reply]

Delete Derivative work of the song. Also a derivative work of the software used, but that part is likely below the threshold of originality. --Stefan2 (talk) 22:10, 20 November 2012 (UTC)[reply]