No, you cannot share it, because the music you think you wrote probably isn't your music. It belongs to the music publisher who published the particular sequence of four notes before you did. Under the "substantial similarity" standard used by United States courts, there are fewer than 50,000 possible distinct melodies in the Western musical scale, and there are hundreds of thousands of copyrighted songs published by major music publishers who have cross-licensing agreements with one another. Do the math. What's the probability of avoiding a lawsuit? What's the probability of winning if you can't afford legal representation?
It's more or less the inverse, this bug enables the referer to know where they refered you to.
Grandparent was talking about the CGI scripts used to track users who click an outward link on a web site. (Some Slashdot users abuse those scripts to create a link that appears to go to Yahoo! but really goes to Goatse.cx.) However, this bug in Mozilla gives a site's scripts access to a clicked bookmark or to a URL entered in the location bar.
However, shouldn't "prior art" then be able to be argued in this case?
You said "argued". It's true that copyright protects only the original portions of a work, but if you don't have much money on hand, you can't afford to hire somebody to argue anything. Those without sufficient income to afford legal representation must steer clear of performing any action that anybody else could conceivably think of as infringing or defamatory.
Yea. So should Referrer be removed from existence.
I respectfully disagree. Without the Referer: header, how is a developer supposed to know whether or not somebody else is leeching his bandwidth by linking directly to an image or to a large zip file, so as not to run into problems with metered bandwidth?
The best advice is just to use your best judgment when sampling
That may work for sampling, where it's possible to completely avoid liability by not using a particular recording, but it won't work for songwriting, where even if there are more than a million possible melodies, the sheer number of existing songs means that you are still likely to infringe accidentally.
The standard for proving copying under U.S. case law includes two components: substantial similarity, which we've discussed, and access to the original work. The "He's So Fine"/"My Sweet Lord" case destroyed the "independent creation" defense: the fact that a song was played on the radio during the defendant's lifetime constitutes proof of access to the work.
Won't help much. The 46,656 figure already accounts for basic rhythm by allowing a note to be short, medium, or long. Do you think any finer gradations would make the melody no longer "substantially similar" to the original melody? And even if you do change the rhythm enough to create a different melody, now you're infringing on somebody else's copyright.
how much pressure applied, using the pedal, sharp or soft and probably more I can't think of.
All of which would be considered inconsequential, resulting in a melody that's still "substantially similar" and still infringing.
Actually, 3 symbols in an alphabet of 11 is 11^3 not 3^11.
Even then, what about a repeated attack (think beginning of Beethoven's fifth symphony)? I originally had 11 there, but then I remembered repeated attacks and added the unison interval for a total of 12.
But then, every copyrighted music out there may not be copyrightable
Anyone see the embedded Windows.NET ad in the full story page of this article..?
No. A line in my G:\winnt\system32\drivers\etc\hosts file redirects DoubleClick ad requests to a WinApache virtual host on my machine that puts up a PNG image reading "DoubleClick blocked."
No, I'm not trying to cheat OSDN out of ad revenue. That's actually the only ad site I currently block because 1. it gets rid of most of the Java and Flash ads and 2. it gets rid of a lot of potential privacy invasions.
if it was impossible to trade mp3's(or.ogg or such), i'd be listening to.mods,.xm's , sids, and maybe midi's.
Turning a recording into a.ogg file and distributing it infringes both the songwriter's and the performer's copyright.
Turning a song into a module file (mod, s3m, xm, it, mid+sf2) won't draw any fire from RIAA labels but will still infringe the songwriter's copyright. You still need a license from BMI or Harry Fox, depending on the intended use.
Writing your own music is harder than it looks because it's nearly impossible to avoid "substantial similarity" to the millions of songs out there.
Don't the major labels record rap music and sell it at a profit
Yes.
without giving a cent
Cross-licensing agreements exist among the major record labels. These are analogous to what happens in other industries, where the big players pool their patents to keep newcomers out.
How many notes do you have to copy before it stops being fair use and starts being plagarism?
In legal theory, you infringe copyright when you copy enough to create a "substantial similarity." In practice, that's about four notes.
Is it plagarism if the individual notes of your composition are sampled from some other song rather than played anew in a studio?
They are "played anew" on a digital audio workstation, and they still infringe the songwriter's copyright.
Is a song a "copy" if a stock riff common to many songs of the form happens to be sampled from a commercial recording rather than played anew in a studio - and this can be identified by computer processing but NOT by a human ear (even a well-trained one)?
That's why the developers of the first PC clone BIOS made sure to employ and document clean room methods when writing a binary-compatible replacement for the IBM PC BIOS.
These are not rhetorical questions. Some of them have already been litigated.
It's odd how the RIAA are stamping down on P2P but not on TAB and sheet music.
Published non-recorded copies of musical works are the NMPA's domain (national music publishers association), not the RIAA's (recording industry association of America). And yes, the NMPA does tag along when the RIAA sues P2P service operators.
So there are 50000 melodies and thousads of ways of playing them
The courts aren't looking for "identical". They'll easily overlook the "thousads[sic] of ways of playing them" when applying the legal standard of "substantial similarity." My figure of 50,000 melodies was a statistical attempt to capture what it takes for two melodies to be "substantially similar."
sometimes I wonder how Weird Al and others get away with some of their works, unless they have contracts/agreements with the original artists?
Al Yankovic makes a point of signing such contracts with the original songwriters. He doesn't have to pay the original performers because he typically doesn't sample; he covers (re-performs) the songs. For example, that's why the guitar at the beginning of "She Drives Like Crazy" sounds so much different from the one at the beginning of "She Drives Me Crazy" by Fine Young Cannibals.
When you the MP3 of the song in a few filters, it gives the DeCSS source code.
I have done this. Along the lines of what Aphex Twin used to hide his face, I wrote a program that converted a.bmp of the efdtt source code (efdtt is a small DeCSS program, available at the Gallery of CSS Descramblers) into a waveform (using an inverse fourier transform of sorts) and mixed it on top of some song.
but is there not a certain degree of freedown allowable in reference parodies?
Under United States copyright law as interpreted by the courts, parody is only parody when the parody ridicules the original work itself. That's why The Wind Done Gone is legal but The Cat Not in the Hat isn't.
As the second article you link to states, the test is for the tune being "substantially similar."
However, even though I know that "substantial similarity" is strictly not a statistical measure, four notes is the best statistical approximation of "substantially similar" that I have ever found. Do you have a better one? And how does a songwriter determine whether or not a work that he or she created is "substantially similar" to at least one of the million or so musical works still under copyright?
While i believe there is/was at least one startup that was working to match music using a beats & tone analysis method that could match to songs that had been shifted or obscured in some way
Why don't these people put their time to some constructive use and learn how to write actual music on their own
Could it perhaps be because songwriters either are close to running out of unique melodies or already have run out of unique melodies? (There exist fewer than 50,000 possible melodies; read this article to see why.)
Slashdot Mirror
[Allowing links from outside GameFAQs.com to link only to HTML pages is] a solution that doesn't involve Referrer.
Without the Referer: header, how is the GameFAQs server supposed to know if a link is from outside or not?
What if it's MY music? I cannot share it?
No, you cannot share it, because the music you think you wrote probably isn't your music. It belongs to the music publisher who published the particular sequence of four notes before you did. Under the "substantial similarity" standard used by United States courts, there are fewer than 50,000 possible distinct melodies in the Western musical scale, and there are hundreds of thousands of copyrighted songs published by major music publishers who have cross-licensing agreements with one another. Do the math. What's the probability of avoiding a lawsuit? What's the probability of winning if you can't afford legal representation?
Well, that's why you put things on the web (a public forum)... For other people to point their browsers at.
But when other people link directly to non-HTML files, your advertisers don't pay you. That's why GameFAQs.com allows linking only to HTML pages.
I happen to run junkbuster - you get no stinking Referer from me.
Many popular download sites (fileplanet, gamespy, gbadev.org, etc.) happen to run a leech script - you get no stinking cool apps from them.
It's more or less the inverse, this bug enables the referer to know where they refered you to.
Grandparent was talking about the CGI scripts used to track users who click an outward link on a web site. (Some Slashdot users abuse those scripts to create a link that appears to go to Yahoo! but really goes to Goatse.cx.) However, this bug in Mozilla gives a site's scripts access to a clicked bookmark or to a URL entered in the location bar.
(Why don't < and > work when I select "Plain old Text"?)
1. Use the Preview button to avoid submitting comments with mistakes.
2. According to this FAQ page, Plain Old Text only converts newlines to <br>. You're looking for Extrans, which also escapes &, <, and >.
However, shouldn't "prior art" then be able to be argued in this case?
You said "argued". It's true that copyright protects only the original portions of a work, but if you don't have much money on hand, you can't afford to hire somebody to argue anything. Those without sufficient income to afford legal representation must steer clear of performing any action that anybody else could conceivably think of as infringing or defamatory.
Yea. So should Referrer be removed from existence.
I respectfully disagree. Without the Referer: header, how is a developer supposed to know whether or not somebody else is leeching his bandwidth by linking directly to an image or to a large zip file, so as not to run into problems with metered bandwidth?
The best advice is just to use your best judgment when sampling
That may work for sampling, where it's possible to completely avoid liability by not using a particular recording, but it won't work for songwriting, where even if there are more than a million possible melodies, the sheer number of existing songs means that you are still likely to infringe accidentally.
The standard for proving copying under U.S. case law includes two components: substantial similarity, which we've discussed, and access to the original work. The "He's So Fine"/"My Sweet Lord" case destroyed the "independent creation" defense: the fact that a song was played on the radio during the defendant's lifetime constitutes proof of access to the work.
changing the duration of each note
Won't help much. The 46,656 figure already accounts for basic rhythm by allowing a note to be short, medium, or long. Do you think any finer gradations would make the melody no longer "substantially similar" to the original melody? And even if you do change the rhythm enough to create a different melody, now you're infringing on somebody else's copyright.
how much pressure applied, using the pedal, sharp or soft and probably more I can't think of.
All of which would be considered inconsequential, resulting in a melody that's still "substantially similar" and still infringing.
Legally, I suppose the question would be the number of melodies that people could distinguish as being different.
Watch it. Under United States copyright case law, the standard is "substantial similarity" not an exact match.
People with absolute pitch consider the same melody or even the same motive played at two different pitch level to be quite different, for instance.
Different but still substantially similar.
the number of possible melodies is 3^11 == 177147
Actually, 3 symbols in an alphabet of 11 is 11^3 not 3^11.
Even then, what about a repeated attack (think beginning of Beethoven's fifth symphony)? I originally had 11 there, but then I remembered repeated attacks and added the unison interval for a total of 12.
But then, every copyrighted music out there may not be copyrightable
That issue is explored in the anti-Bono short story "Melancholy Elephants" by Spider Robinson.
Anyone see the embedded Windows .NET ad in the full story page of this article..?
No. A line in my G:\winnt\system32\drivers\etc\hosts file redirects DoubleClick ad requests to a WinApache virtual host on my machine that puts up a PNG image reading "DoubleClick blocked."
No, I'm not trying to cheat OSDN out of ad revenue. That's actually the only ad site I currently block because 1. it gets rid of most of the Java and Flash ads and 2. it gets rid of a lot of potential privacy invasions.
if it was impossible to trade mp3's(or .ogg or such), i'd be listening to .mods, .xm's , sids, and maybe midi's.
Turning a recording into a .ogg file and distributing it infringes both the songwriter's and the performer's copyright.
Turning a song into a module file (mod, s3m, xm, it, mid+sf2) won't draw any fire from RIAA labels but will still infringe the songwriter's copyright. You still need a license from BMI or Harry Fox, depending on the intended use.
Writing your own music is harder than it looks because it's nearly impossible to avoid "substantial similarity" to the millions of songs out there.
Don't the major labels record rap music and sell it at a profit
Yes.
without giving a cent
Cross-licensing agreements exist among the major record labels. These are analogous to what happens in other industries, where the big players pool their patents to keep newcomers out.
How many notes do you have to copy before it stops being fair use and starts being plagarism?
In legal theory, you infringe copyright when you copy enough to create a "substantial similarity." In practice, that's about four notes.
Is it plagarism if the individual notes of your composition are sampled from some other song rather than played anew in a studio?
They are "played anew" on a digital audio workstation, and they still infringe the songwriter's copyright.
Is a song a "copy" if a stock riff common to many songs of the form happens to be sampled from a commercial recording rather than played anew in a studio - and this can be identified by computer processing but NOT by a human ear (even a well-trained one)?
That's why the developers of the first PC clone BIOS made sure to employ and document clean room methods when writing a binary-compatible replacement for the IBM PC BIOS.
These are not rhetorical questions. Some of them have already been litigated.
Links would have been useful.
It's odd how the RIAA are stamping down on P2P but not on TAB and sheet music.
Published non-recorded copies of musical works are the NMPA's domain (national music publishers association), not the RIAA's (recording industry association of America). And yes, the NMPA does tag along when the RIAA sues P2P service operators.
So there are 50000 melodies and thousads of ways of playing them
The courts aren't looking for "identical". They'll easily overlook the "thousads[sic] of ways of playing them" when applying the legal standard of "substantial similarity." My figure of 50,000 melodies was a statistical attempt to capture what it takes for two melodies to be "substantially similar."
sometimes I wonder how Weird Al and others get away with some of their works, unless they have contracts/agreements with the original artists?
Al Yankovic makes a point of signing such contracts with the original songwriters. He doesn't have to pay the original performers because he typically doesn't sample; he covers (re-performs) the songs. For example, that's why the guitar at the beginning of "She Drives Like Crazy" sounds so much different from the one at the beginning of "She Drives Me Crazy" by Fine Young Cannibals.
When you the MP3 of the song in a few filters, it gives the DeCSS source code.
I have done this. Along the lines of what Aphex Twin used to hide his face, I wrote a program that converted a .bmp of the efdtt source code (efdtt is a small DeCSS program, available at the Gallery of CSS Descramblers) into a waveform (using an inverse fourier transform of sorts) and mixed it on top of some song.
it matches the signal from the music at certain points
Clarification: Gracenote's CDDB 2 may do this. CDDB 1 (used by FreeDB) was based solely on track lengths.
but is there not a certain degree of freedown allowable in reference parodies?
Under United States copyright law as interpreted by the courts, parody is only parody when the parody ridicules the original work itself. That's why The Wind Done Gone is legal but The Cat Not in the Hat isn't.
As the second article you link to states, the test is for the tune being "substantially similar."
However, even though I know that "substantial similarity" is strictly not a statistical measure, four notes is the best statistical approximation of "substantially similar" that I have ever found. Do you have a better one? And how does a songwriter determine whether or not a work that he or she created is "substantially similar" to at least one of the million or so musical works still under copyright?
Fortunately there's an easy solution. Write and sell music of your own
Do you guarantee that the solution you mention is even possible to perform? See my other comment.
While i believe there is/was at least one startup that was working to match music using a beats & tone analysis method that could match to songs that had been shifted or obscured in some way
That was Relatable.
i'm not sure this technology has ever been in real use.
Napster 10.x used it. MusicBrainz uses it.
11,000 albums heavily compressed to 160kbps still takes approximately 600gb
Relatable claims that its tech can identify songs down to 16 kbps.
Why don't these people put their time to some constructive use and learn how to write actual music on their own
Could it perhaps be because songwriters either are close to running out of unique melodies or already have run out of unique melodies? (There exist fewer than 50,000 possible melodies; read this article to see why.)
I remember something about copyright law and music: you are legally allowed to use 4 measures or 10?
You're allowed to use three notes, as long as they aren't a sample. For more information, read my other comment.