Perfect Layout

[Tj Shredder]

I am a very happy user of the LinnStrument, playing all the time the rosa Dulcimer of the Equator synth from Roli...
When I got it, I stuck to the standard 4th tuning layout and could come along fine. Then I was thinking, as my original instrument is violin, and even Robert Fripp put some thoughts into tuning his guitar in fifths, to go to a fifths tuning. This had two advantages over the 4th tuning: The LinnStrument would roughly span one more octave and playing melodies felt very harmonically in terms of the intervals across "strings". But it came with a price - as the fifth is the most common interval in harmonic chords, playing chords felt quite uncomfortable, as I would often have to play pads directly over each other (this is an advantage on a guitar by the way, as you would play it as barre chord).
Today I tried the mode in tritones across "strings". This made a wow effect. Not only chords are easier to play, the whole layout is now very symmetric, octaves are simply two strings apart. Its easier to navigate and it seems all harmonic theory jumps right into your face on its own. Playing any scale "upwards" will keep its fingering the same as it repeats exactly every two strings. Playing a unison trills with one hand seems to be on a perfect distance for my hand and so on. I can't find any disadvantage...
I guess I will stick to this to get deeper into the LinnStrument...
I wonder how others think about this and would love if you shared your personal experience here...

[docsunset]

TL;DR: Its difficult to say what the perfect layout is, but I think it would be interesting to try to empirically test different layouts, perhaps by trying to geometrically predict how easy or difficult it is to play certain chords or scales on them.

I think this is a very interesting question. I would state it, "what is the best way to play notes on a rectangular grid?" This question is naturally incomplete, since it's not necessarily clear what it means for a way to play notes to be "the best way", nor is it necessarily clear what it means to "play notes" (What temperament? Does using an arpeggiator count as playing notes? How about triggering a clip in Live?), but if one nevertheless plunges onward, the question is still interesting. It's particularly interesting as rectangular grids become increasingly common (LinnStruments, Push, Launchpad, MPC, Blocks, Geoshred, Monome, etc). As this style of interface becomes more common, there's the potential for people to start to really master playing music on it, but only if certain mappings are also adopted as common.

Right now the default arrangement on the LinnStrument is probably the most common layout, also seen on the Push and the Launchpad when they are used to play notes. Roger calls this "Fourths String Layout", which I think is a very appropriate name. "String" implies, among other things, that moving left to right on the grid will increase the pitch continuously and chromatically from cell to cell, and that moving up and down between rows will increase the pitch by a certain interval (in this case fourths). From here we could generalize, and say that a "string" note layout is any layout where the pitch changes chromatically from left to right, and by a fixed interval between rows. The chromatic feature of string layouts is a definite advantage for a lot of music, making it easy to do things like pitch bends between notes and land in a certain place consistently, and to play in all twelve keys in equal temperament.

If one restricts their inquiry to string layouts (which by definition excludes many other isomorphic note layouts like Janko and Harmonic table, and any other note playing strategies one might consider, such as samchillian-style for example), then the question simply becomes "what interval between strings makes the best string layout for playing notes on a rectangular grid?" This is arguably a clearer question.

Brief experimentation will show that all string layouts have many features in common: 1. They are all isomorphic, meaning that the fingering is the same* for all keys. 2. Pitch slides are always uniform. 3. Arguably, music theory concepts should be easier to learn because of the isomorphism.

Some layouts are clearly disadvantageous, such as semitones or tones string layout, where there is such a small difference between rows that there might as well be only one. But for all the rest, it's less obvious which, if any, is superior. What criteria make one layout better than another? Is there any way to empirically compare two string layouts? It seems uncontroversial to say that a layout is better if it is easier to play common pitch structures (chords, scales, melodies, etc.) in that layout than in another.

In Roger's article, he argues that 4ths layout is the best layout because it's possible to play many chords in one hand, including chords spanning several octaves, but the layout still provides a very good pitch range. He specifically compares 4ths layout to major thirds and fifths. I've no doubt that he extensively tested all of these layouts, and probably other, while designing the LinnStrument. Another important factor in favor of 4ths layout is that it is already seeing widespread adoption. It's as likely as not that there's no really critical difference between different string layouts, and that practicing on the layout is much more important than the layout itself.

Despite all this I still wonder if there might be more empirical ways to measure the relative merits of different layouts, rather than simply relying on experimentation and good musician-ly judgment (not to say that this isn't a great and important kind of measurement). Is it possible to geometrically predict how easy or difficult a chord is to play based on its shape, for example? If so, one could write a program to automatically generate ease/difficulty scores for a library of common chords, and empirically estimate which layouts are the best for playing which chords. Similar strategies could perhaps be applied to other note structures.

It's something I'm interested in. Is anyone else interested to empirically compare different string layouts?

*This is true and not true. In any given isomorphic layout, there's usually at least two ways to play any chord or scale. In tritone string layout, for example, there are as many as six ways, depending how you count. To find this yourself, take two pieces of paper or two books large enough to cover part of the LinnStrument from top to bottom. Play one book somewhere in the middle of the grid. Play the other on on either side so that there are six columns uncovered (or five if you prefer to play in 4ths layout). Starting from the lowest note you should be able to play a scale between the two books. If you now move both books to the right one column, you can play the scale still, but now with a brand new shape! Repeat 4 more times (fewer if you prefer to play in 4ths). 6 is still a big improvement on 12 on the piano, and you can probably ignore some of them, but I'm just saying: there's usually more than one play to play anything on an isomorphic layout. You're not totally free of the tyranny of learning new fingerings.

PS. @TJ Shredder, I don't mean to hijack your thread. These are just my personal thoughts.

Edited to add TL;DR

[Roger_Linn]

Great discussion! Here's a hex layout I considered along the way, same as LinnStrument's grid but with alternate rows shifted over by 1/2 column then squished down to form hexagons. I rejected it because the grid seemed more intuitive:

http://www.rogerlinndesign.com/assets/l ... layout.svg

[Tj Shredder]

I just came across this, and of course whatever layout you choose, it will need some time to get used to it to be able to judge on a very personal level what will fit best for you. So far I can tell, though the +6 layout I still have the least experience, it feels more natural to me already, compared to 4th or 5th. Even visually it seems easier to grab... Suddenly a standard scale looks much more symmetric than on a normal keyboard. Playing a minor or major chord is much more relaxed for me...
The diagonals are either a circle of fifths (to the right up) or 4ths (to the left up) both do have an harmonic importance.
But I am in the very beginning of learning this, of course there is a lack of "schools". I probably will use some piano lessons and adapt them to the LinnStrument. But such things like fingerings I have figure out myself. For such a new instrument we have to collect a lot of different experiences, and only time will tell if there will be a "best" or common way to do it. Kudos to Roger, Tim, Geert and the whole team to allow us the freedom to set it to different layouts at all. This is an important design decision. ( I am also glad to have the big version, as the 128 version seems to have physical navigation aids for the 4th layout...;-)

[Roger_Linn]

Very interesting, TJ! Please keep us abreast of what you discover about the tritone (6 semitones) Row Offset. Your journey is our reward.

[Tj Shredder]

It came to my mind, that string instruments might also be tuned preferably in fifths or 4ths, because they can play open strings, that way more open strings will fit in a scale for most common keys.
For the LinnStrument that does not apply. Or if you'd want something like that, you'd rather solve it in software and pluck those open strings differently and would choose pitches independent of any layout. If my suggestion for an additional row controller would be implemented, one could easily mimik open strings...

[CaseyD]

( I think it must be frustrating for designers such as RL to know that flexible, clear OLED-like displays are just a few years away.)

[Tj Shredder]

Don't know what this has to do with the topic, but some years ago we had that Lemur, which now entirely moved to tablets. But most instruments need a focus. Too much flexibility is not necessarily an advantage. I predict a bright future for the LinnStrument. It is much more inspiring than the Lemur (and its successors) ever had been...
I also prefer real faders over any displayed fader. Nothing can beat a tactile user interface...

[OSC85]

After a good while experimenting with different layouts, I also think the tritone layout yields the best results for my efforts.

Everything makes more sense and you have less choices for different fingerings which I consider a great advantage in building muscle memory.

Now if the rows were slightly offset like a computer keyboard to account for natural hand atonomy .... we'd be looking at the perfect layout and instrument

[Roger_Linn]

You mean like this?:

http://www.rogerlinndesign.com/assets/l ... layout.svg

[BobDog]

Are you thinking of doing a hex one Roger?

I still play my AXiS every so often, a linstrunent version would be pretty neat.

[Roger_Linn]

Sorry, no. That picture has been on my site for a couple of years as an example of layouts I considered but few wanted. You can get to it from the LinnStrument page, then click the link "more about piano vs. string layout", then on the next page click the link "hex layout".

People overwhelmingly favored the grid over the hex layout, and the fourths tuning over all other tunings. For an isomorphic layout, there are really only 2 options-- grid and hex, and hex is merely a grid with alternate rows shifted over by 1/2 column then squished down to form hexagons.

You might consider lobbying the makers of the Joué or Sensel Morph to make a hex overlay for their devices. Both of these devices have a general-purpose pressure-sensitive multitouch surface underneath and therefore permit different overlays, at the expense of playing surface size and lit notes. By comparison, everything in LinnStrument is designed specifically for the grid layout. I could make a hex version but it would require changing everything inside, and there's not enough demand to justify the development cost.

[OSC85]

Hi Roger,

I'm definitely not referring to the hex layout. What I meant was having the starting position of each string slightly offseted to the one below it so grabing 2 notes below one another naturally falls under the fingers.

Essentially if you look at a computer keyboard and disregard all keys except for the letters, you can see the offset pattern that I'm referring to.

Of course I could be totally wrong in thinking that it could be beneficial but it works really well for the computer keyboard so There might be a merit to the design.

Thanks,

[BobDog]

Sorry, no. That picture has been on my site for a couple of years as an example of layouts I considered but few wanted. You can get to it from the LinnStrument page, then click the link "more about piano vs. string layout", then on the next page click the link "hex layout".

People overwhelmingly favored the grid over the hex layout, and the fourths tuning over all other tunings. For an isomorphic layout, there are really only 2 options-- grid and hex, and hex is merely a grid with alternate rows shifted over by 1/2 column then squished down to form hexagons.

You might consider lobbying the makers of the Joué or Sensel Morph to make a hex overlay for their devices. Both of these devices have a general-purpose pressure-sensitive multitouch surface underneath and therefore permit different overlays, at the expense of playing surface size and lit notes. By comparison, everything in LinnStrument is designed specifically for the grid layout. I could make a hex version but it would require changing everything inside, and there's not enough demand to justify the development cost.

I was getting all excited for a while there

"Changing everything" doesn't sound good though so I get your point.

Andy

[Roger_Linn]

What I meant was having the starting position of each string slightly offseted to the one below it so grabing 2 notes below one another naturally falls under the fingers.

Essentially if you look at a computer keyboard and disregard all keys except for the letters, you can see the offset pattern that I'm referring to.

I think you mean that instead of the hex keyboard arrangement in my link, in which alternate horizontal rows are shifted left or right by 1/2 column, you'd want each of the rows to be shifted over--in relation to the row above or below it--by a distance less than 1/2 column? if so, I don't see any advantage, but perhaps you'd be willing to draw what you're thinking and post it here for me and others to better understand the perceived advantages.