Swing Ratio | Hacker News Reader

Swing Ratio

nld-intern.ds.mpg.de

142 pointschesterfield32 comments

This is a misleading interface, because it doesn't actually support continuous variation of swing ratio. I was confused for a while because I couldn't detect anything until 1.06, but it turns out it's just playing pre-rendered mp3s, rounding to the nearest 0.1. I couldn't hear 1.05 because it's the same mp3 as 1.0.

1.06 is rounded to 1.10. It's clearly audible, and IMO a big improvement over 1.0. I think 1.2 is even better, and 1.3 also good, but beyond that the effect becomes too exaggerated for my taste.

Sub 1.0 ratios are interesting. I don't think they sound very good, but they make it feel like the tempo is slower despite being the same.

I prefer Groove Pizza for exploring this [1]. Makes it very clear visually, IMO.

There's also a good video explaining it [2].

  [1] https://apps.musedlab.org/groovepizza/
  [2] https://www.youtube.com/watch?v=T4EtsPf3X-I

Wow, super cool. As a musician, I find it oddly unintuitive to think of it that way. I guess I approach it more as a bouncing ball, linearly from x=0=>, left to right.

Clickable links:

[1] https://apps.musedlab.org/groovepizza/

[2] https://www.youtube.com/watch?v=T4EtsPf3X-I

Apparently the swing ratio (as measured by the ride cymbal) used by professional jazz musicians varies with the tempo:

Friberg & Sundström, "Preferred swing ratio in jazz as a function of tempo", SMH-QPSR (1997) http://kth.diva-portal.org/smash/get/diva2:1246291/DATASET01...

That’s awesome that someone put the data together. It makes total sense to me, since the slower the tempo, the more blunted the impact (less emotional, less jarring) of a low-ratio

Fig 2 in https://www.nature.com/articles/s42005-022-00995-z.pdf shows an analysis of swing in solos as function of tempo. At 160 bpm, soloists swing the most.

3.0 isn't swing.

3.0 is a time-traveling swing band stuck in 17th century France, forced to play courtly music at Versailles until Lully agrees to smuggle them out an underground passageway in exchange for a copy of that smoking hot basso continuo part

In conclusion, 3.0 is my favorite because it reminds me of that old timey Versailles swing

The most interesting part of this to me is seeing that this is made with brython, and spotting a text/python script in the wild. Fun stuff, people are really creative.

The brython appears to only have been used for SVG manipulation. I would have been more interested in seeing the audio being generated dynamically based on the swing ratio, but instead it just plays pre-rendered mp3s.

This is the least swing-y phrasing of this theme I've ever heard. The swing ratio is only one aspect, arguably not really an important one. You can create swing feel also by emphasizing off-beats, which this thing absolutely doesn't do. No matter 1:1 or 1:2 or 1:3.

A reasonably skilled jazz musician can get swinging at any of them. In modern jazz it's quite common to actually do 1:1 with a quite lazy/delayed on-beat, so it feels like 1:2 even though if you'd measure it, it's close to 1:1.

Max-Planck-Gesellschaft funds this? Sorry to be blunt, but this satisfies all German stereotypes. It's the furthest from music you could ever get.

Reminds me of this old Python experiment, which will ‘swing’ any MP3:

https://musicmachinery.com/2010/05/21/the-swinger/

Unfortunately most of the examples seem to be gone, but I did find this great one on YouTube:

https://youtu.be/9lBP-KvujRY

Some common ratios since I spent a few minutes figuring these out:

* 3.0 - Dotted 8th note + 16th note played as it should. Many classical musicians don't play this pattern "sharply" enough IMO.

* 2.0 - Triplets (quarter note + 8th note inside a triplet).

* 1.5 - Beat subdivided by 5 with the first note taking 3 counts and the second note taking 2 (dotted eighth + eighth note in a 5:2 polyrhythm). Not uncommon in swing and fast jazz music.

I was at a Christmas church service on Monday and the quartet of singers was making two classic errors in a particular piece, one in the first measure and the other in the second measure of the refrain:

The first was singing a triplet rhythm (even quarter note triplets) as a dot-dot-whole (dotted eighth, dotted eighth, eighth).

The second was singing a dotted rhythm (dotted quarter note, eighth) as a triplet (i.e. swung eighths).

It was amusing

When danced, watch their feet for the triple-step rhythm that hangs on the swing:

https://www.youtube.com/watch?v=kvh1dYcPlPc

https://youtu.be/tnPfJFFYAcw?t=397

Swing ratio and tempo are related. Large swing ratios sound better at a slower tempo.

http://kth.diva-portal.org/smash/get/diva2:1246291/DATASET01...

I'm not sure why, but to me a ratio of about 2.5 sounds the most pleasingly rhythmic. I wonder if it has some interwoven relation with audio perception or if it's just a matter of taste!

That's interesting - 2.5 was the same number I landed on that sounded best to me.

Interesting, usually the ratio is 2 as it makes triplets!

2.5 is sort of the over-the-top classic laid-back sound. 1.25 also works very well and is typically close to what you hear on faster passages in a lot of jazz. It would be cool if this had multiple sections of music to show where different swing ratios are appropriate, like some faster and slower stuff, Bebop and other styles.

To my ears 2.5 sounds both too rushed and a little too regular - I think ~1.9 or ~2.1 sounds best.

Somebody studied a bunch of recordings and found a mean ratio of 2.38:1. That sounds good on this record too. Further in the article are some observations of a correlation between swing ratio and tempo.

http://kth.diva-portal.org/smash/get/diva2:1246291/DATASET01...

None of these really swing, because swing isn’t (just) about note duration; where the accents are placed, and which notes get partially swallowed are at least as important.

Setting it to less than 1 sounds horrible!

yeah, like speech played in reverse - satanic!

It has it's place: https://www.youtube.com/watch?v=h9YyazDg9YM

I don't think it's <1 anywhere in that song, not sure why you linked to this. It sounds a bit less than 2 mostly, the super-common 12/8 shuffle feel, as in e.g. Rock Around The Clock (1955) https://www.youtube.com/watch?v=ZgdufzXvjqw and a zillion other songs.

ah, ducks, I flipped the notation in my head; you're correct sixteens are swinged, but it's 1.5, not 0.75

I'm not a musical person so I couldn't tell any difference. What is it meant to demonstrate?

putting it below 1 reveals one thing swing does- it orients your ear to the beat. as i think lots of musical devices do, it enhances perception. it sounds "backwards" when you have it less than 1

Anyone know what the just noticable difference of swing ratio is?

That would depend on the tempo, the listener, and a host of other things.

One thing I learnt, funk feels like this: https://youtu.be/gLjXWiZVCrs?t=297 - Clyde Stubblefield's amazing drum feel on Give It Up Or Turnit A Loose - rely on a very subtle swing ratio, very slightly >1. I've found delaying the offbeat quavers (1/8 notes) by something like 1/32 note triplets gives something like this feel. It doesn't sound swung, but if you play it evenly (with a 1:1 swing ratio), it sounds horrible, dead, the groove is lost.

Reply on news.ycombinator.com