Steve

My wifey is calling, need to sign off. We're not fully reconciled but at least we each know a little about where the other is coming from.

I think afer all this back and forth, we may be getting somewhere. Beauty, usefulness, cohesiveness, longevity, value, and melody to some degree is in the eyes and ears of the beholders. Humans, not definitions determine value. But some definitions are better than others. In the engineering world, someone coined "All models are wrong, but some are useful".

If you could build one house out of air and the owners moved in they would quickly find that (yes it may be a house as you point out) but it doesn't work. It quickly thaws and then boils. Like I say, millions of similar "houses" have been built out of "air" and all of them were failures and are lost to history. They didn't follow the rules.

Ok, I listend for about 3 minutes. From zero to about 1 minute I hear phrases (no melody) then I begin to detect a melody. Around 2 minutes the melody becomes stronger and I can hum it. Overall, I like it. Whatever rules that are out there, I'd say they followed them.

Put another way. Who cares if it's theoretically possible to build a house out of air; yes air. It IS possible to chill it into a liquid and then continue chilling until it freezes. Now you have a solid building material. But here's the catch, nobody CARES because you can't phisically do it. It's not useful, consistent, viable and it doesn't work in practice.

Imagine for a minute if Beethoven messed up big time on the melody of Fur Elise. Can you not imagine such a messup that we would not have that great classic today? Conversely, how many pieces of music have been lost to the dustbinof history because of such rules that wer broken? I say millions.

If the "melody" is inconsistent with everything else, how can that be a true melody??? Music is a language. All languages have rules. If you break enough of those rules your audiance will not understand you and you are transmitting garbage. Few people will tolerate garbage.

FWIW, I am becoming more "liberal" in my understanding of melody. In other words, I recognize that just because I can't detect a melody doesn't mean that one doesn't exist. That said, there has to be limits on what is and what is not a melody. If not, then why does the word exist? It exists to descibe something other than harmony and rhythm.

Maybe this can help. Should I qualify "melody" with, useful? Understandable? Consistent with the chord progression? Consistent with the tempo? Consistent with the genre? Etc. Melody for me is not some abstract technical spec that doesn't care whether it works in the song or not. Rather, melody is a physical, important aspect of the piece. And music worth anything, especially music that survives generations had to have followed non-random rules.

Yes, I am going off topic because the original topic involved SFM to which I learned that there was little agreement on "melody". Then to make sure we have ANY overlap in understanding I need to confirm that we have a common understanding of what MUSIC is. And if we fail there, the next level down is SOUND.

Let me go one step deeper and more fundamental. Music (in my mind) is designed, recorded and performed to serve a purpose; battle cries, patriotism, entertainment, therapy, money, message-sending, etc. It's designed (in my mind) for a the purpose of someone seeing value in it. Music is not noise.

My main point is that music theory and associated definitions go back a long time. Way before you, me, StackExchange or even Beethoven. It's a mature, tested, tried and improved over the centuries for the purpose of creating music for human consumption. I'm not aware of randomness or noise playing a large role in it's development, for good reason.

Yes, it is an individual thing and Beethoven, McCartney and thousands of others have figured that out. I agree, randomness CAN result in a melody, but only if you're lucky. I'm talking about design and composition not luck.

OK, here is the Oxford def. "A series of single notes arranged in a musically expressive or distinctive sequence". But a series of random, nonunderstandable notes is NOT notes arranged in a musically expressive or distinctive sequence because they are random and nonunderstandable. Can you understand this? "Musical expressiveness" requires design, composition and arrangement. Randomness does not.

There has been a lot of miscommunication on this thread because of a lack of common vocabulary. That's why definitions are so important. What appears to be apparent is there is a technical definition/context of melody (you/"Spanish") and there is a common definition (AI/dictionary/me/"French"). I stated my (common) definition in the original question. I'm still waiting for who told Dekkadeci that any random, chaotic, nonunderstandable set of notes can be an acceptable melody.

@Dekkadeci This is absolutely bizzare to me. I hear you saying that any stream of notes whatsoever, no matter how unpleasant, discordant, unmusical, random, noisy, nonrepeatable, nonunderstandable and downright bad that in your mind it still is a melody. Who told you this? Would Beethoven, Mozart, Mark Knopfler and Paul McCartney agree? Would Music Schools around the world agree? Please, do tell, who told you this?

This might help. Take the Row, row, row your boat song. Which we all agree is a melody. Enter that sequence of notes into a computer program and then one by one randomly change those notes. At some point you will reach a point where you have destroyed the melody and produced garbled noise. How does your definition of melody identify the cross-over from melody to noise? Or could it be that you consider noise a melody?

OK, I now know this is a definition thing. Clearly I can hear the loudest thing and for a short time I can actually hum along with it. But then the regularity breaks down and I'm out of sync as it proceeds on a note much different than what is expected from the prior phrase. To be sure, this does NOT happen to me with hundrends of songs across several genres. Lawrence alluded to why this is with his answer. I am hearing something that you say is the melody. But I'm still left without a crisp working definition that would prevent random noise from being classified as a melody. Make sense?

@Todd Wilcox. I don't think there's anything wrong with my hearing or my speakers., this a definition thing. If I adopt musicmante's definition, almost any sequence of notes could be a melody; random or otherwise and that doesn't make sense in an established theory. The AI bot and Merriam-Webster makes much more sense to me. A friend just contacted me who has a copy of Music Theory Remixed by Hudson and he says they talk about "singability" in regards to melodies which agrees strongly with the dictionary and what the AI came up with. Something isn't adding up. But thanks anyway. We can end.

@musicamante. I understand your definition is "a sequence in time of single (possibly different) notes, normally more recognizable from the other pitch-based components of a music piece (harmony) due to the fact that it's normally the highest pitch, is played by the most prominent and distinctive instrument (by timbre or dynamic), usually has a more complex rhythm than the harmony and varies more frequently than it. "That is quite broad. Do you have a reference source?

Here is another thought. Is it possible that to an American ear with an American understanding of melody can disagree with a Scandinavian ear who has a Scandinavian understanding of melody? If so, can anyone consult with a Scandinavian musician and get his/her definition of melody? Just a thought . . . Another thought is does no one have an American text book on music theory that can be consulted for a definition? What if the phrasing is not just nonlinear, but purely random? Would that still be a melody?

@musicamante. Duly noted on the subjectivity of "sweet" and "agreeable". Sometimes subjectivity is all we have until it's replaced with something objective; which is where it appears we are. Of course Wiki and today's AI are not flawless. But I'm still not hearing YOUR definition of melody. And I'm thinking "linear" is somehow useful.

. . . the reason I think "linear" is interesting is that Lawrence points out that it doesn't fit nicely into 4 or 8-bar phrases. So does that make it nonlinear? I'm thinking yes.

@Edward, Another increment of progress! Your/Wiki's definition is a good starting point for English-English communication. I think we both agree that more flesh is needed for those bones. Here is what my AI assistant tells me "The word melody refers to a “sweet or agreeable succession or arrangement of sounds”. It is a linear succession of musical tones that the listener perceives as a single entity ." But even this needs more "flesh". But I find the terms "sweet" and "agreeable" on the right track. If it can't be sung [easily] is it sweet? "Linear" is also an interesting descriptor.

@Edward, Bear with me as I'm not sure I follow. Depending on the tradjectory of this question I could be interested in the broader question of how melody is or isn't used in SFM. But we seem to have our hands full with just restricting the discussion to the 1st 60 seconds of my example. So I'd like to remain restricted, if possible.

@PiedPiper That's unfortunate. Crisp definitions are everything when it comes to communications. Otherwise you're speaking French and I'm speaking Spanish :) And isn't music chock full of definitions?

@PiedPiper. I gave my definition of melody above. Can you provide yours?

@Edward: I sharpened my question to hopefully clarify. The original was too broad. We seem to have differing views on the defintion of a melody. Can you state yours?

@PiedPiper I think we are making progress on this question. I agree, that the 1st minute is danceable. But so is a drum groove. Can a drum groove be a melody? I say no. When you say "not something most people would want to sing", I also agree. So are we concluding that the 1st minute contains no melody?

@Andy Bonner OK, I listened to the first 2 minutes of Väsen - 30-års jiggen (12) and I can easily detect the skeleton of an explicit melody and can even roughly whistle it. But not the first minute of the video I provided. I find these genres much different.

Yes, I have listened to Scandinavian music outside of this video but clearly I am not an expert; hence my questions. I chose this specific video so that we would have a specific piece of music to refer to and not be overwhelmed by the various genres and sub-genres that exist.

So someone from that culture would find it predictable and therefore (for them) it would contain an "explicit" melody?

And no, I do not have to accept all what people are telling me here. In order to accept something, it must pass a reasonable test of plausibility. ojs' wrong understanding of me talking about non-music is far from plausible. I can never accept that falsehood. Would somebody please message him and tell him I am talking about true music. Reference Pandora, greatest hits albums and award winning soundtracks.

@Tetsujin. What can I say to convince you that I do understand the complexity here. But that doesn't preclude us from discussing it as adults and recognizing that today's complexities are tomorrow's reality. And yes you can quote me on that. You don't seem able to understand that we, as a race, are on a technological spectrum. At one point of this spectrum digital music was science fiction. Now it's commonplace. Was it you that said ""Surely someone must be working on this" is a logical fallacy. " ? How can you say that is a fallacy? Are you all-aware of what MIT is doing? . . .

@ojs, this is going nowhere. I have read up on these subjects and they have nothing to do with you wrongly thinking that I'm talking about non-music when in fact I'm talking about true music. as explained above. The bottom line here is that you are unaware of such and AI. And guess what? You're in good company, neither am I. That is why I posted the question in the first place. I'm just amazed at how easily you have misunderstood what I have been saying from the beginning. Perhaps if you re-read from the top you will see where/how you missed the message.

@ojs, Respectfully I still don't think you understand. There is no "may or may not be music" involved here in the least. I am talking about music; full stop. And it's not just my personal perception or interpretation only but that of society's. The audio samples within a genre I am talking about would be universally accepted as true music; taken from a Pandora playlist, a greatest hits album, a highly acclaimed movie soundtrack or similar. Why on earth would I be interested in noise or non-music?? And more importantly, what did I say to lead you to this misunderstanding??

@ojs, you are correct. "take audio samples" is in the question. As opposed to sheet music, chord charts, TAB, written descriptions, etc. These audio samples may be audible to us when played thru a speaker, but internally they are a digital representations of music. I am most definately talking about music and musical genres here; not random sound. Not only that, I'm talking about emotionally pleasing music such as rock, bluegrass and rockabilly. Is this clear now what I'm referring to?

@Tetsujin Ahhhh, now we may be getting somewhere. There is no logical fallacy on my part here. If in the course of developing such an AI that a mathematically, or waveform or emotional crisp definition of a genre needs to be produced . . . then so be it! I see no problem with solving 2 problems in parallel or in serial. It's done all the time.

@ojs Where did this idea that I am talking about sounds come from? I'm talking about musical genres; NOT sounds.

@Tetsujin. No. I am concluding that YOU are concluding that it is currently not possible to develop an AI that can fuse 2 different genres and thereby produce a new genre . . . and you may be right :) Remember, I'm simply asking the question.

@ojs Perhaps we all misunderstand (or don't understand) how genres work?

However, due to the brevity of this discussion and my own study, I'm convinced that much more research into this than would be indicated here has been done. There is a lot going on at MIT, Berkley and elsewhere. And just because we are unaware of a certain technology today does not mean that it doesn't exist or isn't being seriously worked on. This problem will solved, it's just a matter of time and effort. That said, I do not know how to re-phrase this question to be acceptable to this site. If someone knows how, I encourage you to do it, otherwise it will remain closed. Thanks again.

I suspected that this question was too large to be fully answerable here due to our current lack of understanding of what music (and more specifically what genres and fused genres) are. Based on most of the comments here the 59 people who have viewed this question believe we are still, metaphorically speaking, clanging rocks together in a cave. So I have my answer; you are aware of no such AI. Thank you all for attempting to answer this question, I know it was a challenging one . . .

Hello All, any statistics enthusiasts here that can help answer a question I have on statistical power ?

Hello folks, by chance could someone comment/answer my question in CrossValidated on the proper interpretation of statistical power ? Thanks a bunch :)

Hi, this is my first attempt at chatting here so please be gentle. Might someone here be interested in confirming my bootstrap/confidence results using R, Matlab or some other tool ?

I agree, but there are awfully smart people out there that I'm sure have dealt with this, perhaps some flavor of Monte Carlo. Anyway, I'll let you go. And thanks again for the help, I do appreciate it. --Steve

Because if I run a test to say, 4 years and not fail, I'd definately like to take full credit for that.

Hmmm, maybe this is a separate question in this forum ?

Got it. And yes, make whatever edits you feel would clarify. Since binomial can't work for what I need, are there other methods (besides tradidional distributions) that might work ?