Humanization

Dear Member of the alternative tuning igest,

Altho this is slightly off topic, and all though it seems like a pretty
perdantic question, I have looked everywhere for the awnser to this question
and I simply cannot find it.

My question is about drum machines and what they call a software tool called
a "humanzier", -I don't know if anybody has ever heard this term.

It seems pretty self descriptive this name, and pretty obvious, at least,
and I feel that more people should be using it these days, due to the fact
that there is too much, and in my honest opinion -àFAR TOO MUCH robotic,
tight, stiff, starchy, and mundane, mechanical techno-funk about these days,
and I would have thought that the mainstream would have cottened on to this
fact earlier.

I feel, when listening to techno-funk, at the raves that I have been at,
that I'm going slightly out of my mind, and the feeling is very, VERY
different when I look at a live drummers, and see a live band,-for some
reason, or other, I tend to concentrate of the drummer….

My idea is about applying some sort of stochastic algorithm to loosen, and
VERY SLIGHTLY randomize tight drum tracks, by manner of SLIGHTLY
increaseing/diminishing the pressure of a drum sample, and by SLIGHTLY
offsetting the timing, either forward/backward.

I feel that this could well be done on most music these days, and when
creating a drum line on my music, (when, of course, I get that darn sampler
and music sequenceing software), I intend to go through even if MANUALLY,
and "hand loosen" the tracks of drumming and bass up.

Altho this is by all means a small touch, I feel that it would be a great
pshcyological, advantage when listening to music, at least over a long term.

It greats on my nerves after a while.

But my question is this:

HAS THIS BEEN THOUGHT OF BEFORE IN A DRUM MACHINE?

Also, ---are there any practiceing drummers on the alternative tuning list?

And why not go even further than this, -look, I realise that the individual
effect of drummers being individuals, and having their own style is as
aparent to me-minimal, at least I feel this intuitively, and, at least when
compared to the guitar, or even the piano, but I'm sure that this could be
implimented, as in DEGREE of error in a really good drum machine.

Any comments?

---Sarn.

I would like to bring up a slightly different issue regarding the
"humanization" of rhythms: Is there some sort of order to the
"randomness" in which a rhythm played by a human drummer possesses? Is
there a rhythmic equivalent to the difference between a just third
(5:4) and a pythagorean third (81:64)--in other words, do our ears
tend to prefer particular rhythmic relationships that are slightly
uneven, not because they are uneven due to human error but actually
because there is a more complicated rhytmic relationship than we tend
to think? It would be fascinating to explore the applications of
alternative tuning systems to rhythms, especially if it results in
rhythms that are truly more "humanized" and more wholesome to listen
to than the computer-happy drum machine rhythms one hears at a rave.

John N. Engelmann
xjakvvuz.tripod.com

Dear Sarn,

Humanizing functions have been included in both hardware and software for quite a while now. Basically, they started by loosening up the rhythms by moving notes away from their quantized values by an arbitrary (and usually alterable) amount. Later, a number of software sequencers (Cakewalk is the first to come to mind) allowed for "groove templates" to be applied to tracks. An easy example would be to have the backbeat be the furthest behind the beat (late to "2" and "4" by some amount of ticks). You could even use the algorithms to analyze tracks done in realtime and not quantized to mimic a groove.

It's been done. Maybe you might want to wonder why all those people are dancing *in spite* of the rigidity of the beat. Must feel good to some of them...

Cheers,
Jon

`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`
Real Life: Orchestral Percussionist
Web Life: "Corporeal Meadows" - about Harry Partch
http://www.corporeal.com/

--- In tuning@egroups.com, engell69@g... wrote:

http://www.egroups.com/message/tuning/17616

>do our ears tend to prefer particular rhythmic relationships that are
slightly uneven, not because they are uneven due to human error but
actually because there is a more complicated rhytmic relationship than
we tend to think?

My guess is that it is just due to human error, and there is a human
"empathy" to that kind of thing... leading to the concept of
"humanization..." (Not that that can't be done with a complicated
algorithm...)

Wim Hoogewerf listened to some of my electronic pieces and one of the
things that he "liked" was the fact that the rhythms seemed slightly
"off" or "humanized."

Actually, I had just "played them in..." but it seemed to add to the
piece that way...

________ _____ ____ _
Joseph Pehrson

Joseph Pehrson wrote:

> --- In tuning@egroups.com, engell69@g... wrote:
>
> http://www.egroups.com/message/tuning/17616
>
> >do our ears tend to prefer particular rhythmic relationships that are
> slightly uneven, not because they are uneven due to human error but
> actually because there is a more complicated rhytmic relationship than
> we tend to think?
>
> My guess is that it is just due to human error, and there is a human
> "empathy" to that kind of thing... leading to the concept of
> "humanization..." (Not that that can't be done with a complicated
> algorithm...)

Well, have a look at this article:

<http://www.newscientist.com/features/features_22708.html>

Uneven, but very precise, rhythms. It also doesn't scale with tempo, so
speeded-up Drum & Bass loops will likely sound more uneven than they
should.

> Wim Hoogewerf listened to some of my electronic pieces and one of the
> things that he "liked" was the fact that the rhythms seemed slightly
> "off" or "humanized."
>
> Actually, I had just "played them in..." but it seemed to add to the
> piece that way...

Yes, you'll probably have a good idea what you want a piece to sound like,
and playing it into the sequencer is a good way of realising it. But
slowing down the tempo for recording seems to be trouble.

I'm sure there are similar rules for governing intonation, which haven't
been discovered yet. But narrowing leading tones, and singing sharp but
fusing with the cadence, must be part of it.

Graham

Many, many pieces of hardware and software have had these elements
incorporated for some time. As but one example, check out the "groove
quantize" and "session drummer" features in the latest versions of Cakewalk's
ProAudio.

Rather trendy-ish names for these features, I know, but, what one can do with
them and the ease of interface and user input is QUITE extensive. One can
work with these features on anywhere from an intuitive level to that of a
rhythmical exactitude that would make Nancarrow proud.

Also this tip: if you're working with a drum machine with MIDI capability,
most likely the keypads are all set to a velocity of the maximum value of 127
thus inhibiting any accents from taking place. MIDI a velocity sensitive
keyboard to the drum machine and put the parts in that way as according to
the mapping of drums to keys (most likely to follow a GM set up of the manual
will have some kind of chart). Way cooler and as expressive as you want it to
be. Sounds like a real percussionist layed down the parts this way.

MotoMusic wrote:
> thus inhibiting any accents from taking place. MIDI a
> velocity sensitive
> keyboard to the drum machine and put the parts in that way as
> according to
> the mapping of drums to keys (most likely to follow a GM set
> up of the manual
> will have some kind of chart). Way cooler and as expressive
> as you want it to
> be. Sounds like a real percussionist layed down the parts this way.

This may sound obvious, but one trick I use to make drum tracks sound better
is I actually play a few measures into cakewalk using a keyboard controller,
and then I just copy paste those measures to create the whole drum track.
Most drum parts are pretty much one beat for the verse and another for the
chorus, so this works pretty well. Lots of times I go back and add in a few
crash cymbal hits or trangle stuff or what-not with the mouse, and just
synchronize it with my human entered notes.

For really tricky parts, I record the high-hat first, then the kick, snare
and toms, then I go back and do the wierd cymbal stuff and maybe re-do the
high-hat to add complicated open/closed patterns. Then once again I just
copy past and then maybe make a few changes to sound more human. It takes a
little longer, but it sounds much better.

TOdd

--- In tuning@egroups.com, "Todd Wilcox" <twilcox@p...> wrote:
>
> For really tricky parts, I record the high-hat first, then the
kick, snare
> and toms, then I go back and do the wierd cymbal stuff and maybe re-
do the
> high-hat to add complicated open/closed patterns. Then once again I
just
> copy past and then maybe make a few changes to sound more human. It
takes a
> little longer, but it sounds much better.
>
> TOdd

Todd,

Do you perchance have any mp3s of your handywork out here in internet
land? Would like to check'em. The above sounds interesting!

Jacky

Graham wrote,

>Well, have a look at this article:

><http://www.newscientist.com/features/features_22708.html>

excerpt:

>In his latest research, Friberg went back to the same recordings and looked
at the timing of >soloists, such as Miles Davis, to see if they used the
same swing ratios as the drummers. He >found that the soloists' swing ratios
also dropped as the tempo increased. More surprising was >the fact that the
drummer always played larger swing ratios than the soloist they were playing
>with. Even at slow tempos soloists rarely had swing ratios greater than 2
to 1.

>The difference helps to explain why a soloist can seem to be so laid back
on a particularly >toe-tapping number. When playing a note that nominally
coincides with the basic quarter-note >beat, the soloist hangs back
slightly. "The delay can be as much as 100 milliseconds at medium >tempo,"
says Friberg.

>This tendency to hang behind the beat goes back to the musical ancestors of
jazz. In the >introduction to the 1867 book Slave Songs of the United States
Charles Ware, one of the >editors, observed that when they were rowing a
boat, the oars laid down the basic beat for the >slaves' singing. "One
noticeable thing about their boat songs was that they seemed often to be
>sung just a trifle behind time," he said.

>Members of the audience synchronise with the band by tapping their feet to
the basic beat. But >musicians have a more subtle strategy. "If you generate
a solo line with a computer and delay >every note relative to the cymbal it
sounds awful," says Friberg. "The funny thing," he adds, >"is that there is
a distinctive pattern that most musicians are not aware of. They synchronise
>on the short eighth note."

This is related to a common way of playing straight but behind the beat in
jazz that has often been described as playing exaclty 1/6 of a beat late,
and makes the line swing even though it's played in perfectly even eighths.
The idea is that, if the drummer were swinging in a 2:1 ratio, the short
eighth would be at 2/3 rather than 1/2 way through the beat; 2/3-1/2=1/6, so
a soloist who played straight eighth notes 1/6 of a beat behind would be in
sync with the short eighth note, as Friberg found to be essential.

Many, many pieces of hardware and software have had these elements
incorporated for some time. As but one example, check out the "groove
quantize" and "session drummer" features in the latest versions of Cakewalk's
ProAudio.

Rather trendy-ish names for these features, I know, but, what one can do with
them and the ease of interface and user input is QUITE extensive. One can
work with these features on anywhere from an intuitive level to that of a
rhythmical exactitude that would make Nancarrow proud.

Also this tip: if you're working with a drum machine with MIDI capability,
most likely the keypads are all set to a velocity of the maximum value of 127
thus inhibiting any accents from taking place. MIDI a velocity sensitive
keyboard to the drum machine and put the parts in that way as according to
the mapping of drums to keys (most likely to follow a GM set up of the manual
will have some kind of chart). Way cooler and as expressive as you want it to
be. Sounds like a real percussionist layed down the parts this way.

> >do our ears tend to prefer particular rhythmic relationships that
are
> slightly uneven, not because they are uneven due to human error but
> actually because there is a more complicated rhytmic relationship
than
> we tend to think?
>
> My guess is that it is just due to human error, and there is a human
> "empathy" to that kind of thing... leading to the concept of
> "humanization..." (Not that that can't be done with a complicated
> algorithm...)
I would doubt that you feel the same way about harmony--that
microtonally inflected intervals sound better to our ("our" referring
at least to those who follow this list) ears because they are more
"humanized". The general trend I see among those who support alternate
tunings is one of devoting a huge amount of energy to calculating
precise ratios between frequencies. To say all this effort is in vain
because microtones are governed by human error undermines most
microtonalists' work.

In regards to the discussion on swing rhythms, I suspect that there is
a biological origin in the taste for swing. This origin is the human
heartbeat. Like the human heartbeat, the distance between the second
eigth note and the next quarter note cannot be too short, as a heart
can only switch directions (from contracting to relaxing) so quickly.

-john
equal temperament=equal tamperment