Beat hinting

[pensaku]

There is so called technology named hinting in font rendering technology.

That is, by hand, manually setting which pixels will be rendered on the screen, it achives more clear font rendered result.


In music, same thing happens.


When we hit a note we think it is hit in some kind of random place, or in the right place but we don't know HOW right it is or how random it is.

Notes does not strictly fit in time line. If it did, classical midi file would sounded really good.


Consider, there is, not so implicit and constantly playing, for example 1723Hz frequency tone in your music.

We COULD fit our note in that one cycle of 1723Hz frequency. That's where the PIXEL is if music was a LCD screen. And according to our musical knowledge, we manually fit notes in that kind of frequency.

If one note was hit on 300th cycle of a frequency, we don't need to hit it in 600th. It could be 584th or 601th.


I don't know why it looks good if you manually hint a font, nor do I know why we hit note in certain places but it is obvious that we hit a note in some place and that place could be one cycle of a constant tone frequency.


So, tempo is important. But notes are not hit in exact time, but in almost exact one cycle of a frequency and that's why hand played musics are so good.(Of course there are many factors but it's one of them)

[Nystul]

So, tempo is important. But notes are not hit in exact time, but in almost exact one cycle of a frequency and that's why hand played musics are so good.(Of course there are many factors but it's one of them)

I get the feeling you are searching for a mystical pattern of humanization. You should focus on the more obvious and deliberate subtleties. The human musician does things deliberately or subconsciously based on what he has heard or what he has practiced. We learn to express emotion through music similar to how we learn to express emotion through speech. When we talk, the words we put more emphasis on, the changes in tone in are voice, the cadence (rhythm) we speak with, the articulation, etc. all are used to communicate as much as the actual words we say. Some of these things are lost when reading a typed message, much like the more subtle emotions are lost when rhythms are punched into a piano roll.

[jancivil]

I don't think it's a wise idea to try and abstract these things, as numbers or what-have-you, either. Get into the process and learn from the process and take notes... these things aren't likely to ever gell to the point you'll replace the experiential with a theoretically-derived overarching philosophy.

[innervisions]

That's a really interesting discussion idea.

You're right about the human mind's 'resolution' for the placement of the notes in the timeline - what we define as exact timing is a bit looser than the mathametical definition.

Having that said, I really think that dynamics, phrasing and articulation of a performance would still give you that 'human' feeling, even in a fully quantized performance - timing is just one of the things that the mind is looking for.

[MagicGeoff]

If you're talking about placing notes off-grid, then you need to investigate the phenomonen of Swing in all Latin music (eg, samba). The whole point is that the pushing and pulling of notes against the Grid gives the music it's life enhancing character (and awesome danceabilty)

Though I wasn't sure what you were getting at with the frequencies or pixel stuff, so I may be way off.

However, Swing is awesome either way.

[jancivil]

Much of my life is spent looking at where what I, a human, play onto what I eventually configure as a grid in bars and beats. I am not real young, I have dealt with rhythm for a long time and I assure you that measuring feel vs a grid is just not so easily quantifiable/predictable; it's very complex and a slight variance in context can mean a lot of your bets are off. I am consistently surprised where things that work fall on the timeline.

A good reggae rhythm guitar part, you could groove quantize it til the cows come home, but to get the feel really groovy there's no substitute. There is no template that's really going to supplant the human feel. EG: there is going to be so much variance say between between the 2/3rds and 3/4ths point of a 4/4 bar, and a person will give you all sorts of results. Take a piece of audio that really grooves into Cubase and warp the grid to fit it and find out for yourself how steady rock steady is or not, just by the BPMs, let alone what happens in between the barlines.

The human is superior to the machine for this task, guess what. The experiences with give and take are completely alien to your computer.

Proof: a person won't hit the marks the exact same way two times in a row. That's a GOOD thing in music.

It's a deep rabbit hole to follow into. I think spending time doing it rather than thinking about it is a better use of time.