As easy as ABC?

[DrPhill]

... and just because it there is K: D in the ABC doesn't mean that it is D major !!??!

No, sure it does. Uppercase notes in the K: field default to major unless otherwise specified.

Maybe he was referring to people like me who wish to represent a tune with two sharps, but cannot tell it is in E Dorian, so just call it D. Or there may just be folk less knowledgeable than I who do not realise that two sharps may not mean D Major.

So while you are technically correct MTG, highwood may be correct in a pragmatic sense.

[hans]

It is what the author means it to be. Of course he could have got it wrong and it is actually E Dorian or some other mode which is covered by two sharps in staff notation. Abc notation allows to specify the proper mode, whereas with staff notation one is left to work it out oneself. Would be nice if abc software which renders staff notation would offer the option to print this mode information, like it can print info from other abc fields.

[cross posted with Phill]

[MTGuru]

Cross-post with Phill and hans. I expanded my previous post, saying basically the same as the both of you.

(And Phill, the bit you quoted above has changed.)

[DrPhill]

Of course, if the tune really IS in E dorian, then technically speaking your ABC is wrong; you've written the wrong key/mode. But your software doesn't care a whit, and will give you the proper 2 sharps anyway.

Does it make a difference to the automatic chord generation software - I keep meaning to experiment but might accidentally choose a misleading subset of combinations.

So people sometimes get lazy about this, and just write K:D when they want 2 sharps regardless of the true mode. You see this sometimes in the session.org transcriptions. Or maybe you're genuinely not sure of the mode, only the sharps. So you write K:D as a kluge.

But if you play by the rules all tidy and proper, especially if the ABC itself will be viewed/shared, you'll code the tune's key/mode correctly. And you'll write K:D if and only if you really mean D major.

So if unsure write

Code: Select all

%Two sharps, unsure of mode
K:^C^F?

[Peter Duggan]

The K: field maps unambiguously to a key signature (and clef and staff). And instead of stating a key or mode you can enumerate sharps and flats if you like. For example "K:^f ^c ^g" instead of "K:AMajor".

So good for those quirky Bartok key signatures (yes, I've tried)!

But (while it's coping with one at a time) you'd still be struggling with something like this...

[MTGuru]

But (while it's coping with one at a time) you'd still be struggling with something like this...

For your crossed hands example, try running this ABC through the concertina.net converter (or abcm2ps).

It's close, except you can't force the second key signature to the bottom of the staff.

X:1
M:3/4
Q:"Lento, " 1/4 = 72
%%score [T1 | T2]
K:_e
[V:T1]+mf+(+0+CD E3F|ED C3B,)|
[V:T2 clef=treble][K:^f^g]+p+y"_sempre legato"+1+c3 B cB|c3 B AB|

[Peter Duggan]

It's close, except you can't force the second key signature to the bottom of the staff.

Gosh, I'm quite impressed you've got that far (so must study your example carefully in conjunction with the ABC spec.)!

X:1
M:3/4
Q:"Lento, " 1/4 = 72
%%score [T1 | T2]
K:_e
[V:T1]+mf+(+0+CD E3F|ED C3B,)|
[V:T2 clef=treble][K:^f^g]+p+y"_sempre legato"+1+c3 B cB|c3 B AB|

But I'm curious about your fingering (when that RH '0' could be useful to me!)...

X:1
M:3/4
Q:"Lento, " 1/4 = 72
%%score [T1 | T2]
K:_e
[V:T1]+mf+(+2+CD E3F|ED C3B,)|
[V:T2 clef=treble][K:^f^g]+p+y"_sempre legato"+1+c3 B cB|c3 B AB|

[MTGuru]

But I'm curious about your fingering (when that RH '0' could be useful to me!).

Hmm, I thought I was just copying the fingerings in the original, unless I'm reading them wrong (am I?). Heck, if Bartok says I have a zeroth finger, who am I to object?

Gosh, I'm quite impressed you've got that far (so must study your example carefully in conjunction with the ABC spec.)

I reckon Bartok in ABC is like teaching a dog to dance the Tango. As Johnson says, it's not done well, but you are surprised to find it done at all. Still, for a system that started with scribbling tunes on napkins, it's fairly capable as long as you can live with the (mostly reasonable) defaults. It's when you want finer control over placements that things get dicey. You can always manually edit the PostScript output (which I've done) or manipulate it with something like Perl or Python (which I've done), but that's pretty extreme and geeky.

For my limited purposes nowadays - mostly IrTrad tune transcription and an occasional multipart - ABC is great. If I wanted fancier output, I'd import the ABC into e.g. Sibelius or Finale and gussy it up there.

[cboody]

I asked the question about the Bartok on the abc list. Here's a solution that places the key signature correctly

X:1
T:Hands Crosing
M:3/4
L:1/8
Q:"Lento," 72
V:1 clef=treble
V:2 clef=treble
K:none
%%score {1 | 2}
[V:1] [K:none _e] !mf! (CD E3 F | ED C3 B, |
[V:2] [K:none ^F ^G] !p! "_sempre legato" c3 B cB | c3 B AB |

Apparently you can place the key signature items by upper and lower case letters, something I thought might be true but did not know and could not find in the little digging I did.

[MTGuru]

Apparently you can place the key signature items by upper and lower case letters, something I thought might be true but did not know and could not find in the little digging I did.

Ah, very clever, Chuck. Yes, uppercase in the explicit key sig is apparently undocumented in the 2.0 spec.

As for the other differences (e.g. Q:, V: and K:), most are simply facultative, except

o The trailing space in "Lento, " is required for a space between the comma and the quarter note symbol.

o The invisible y spacer is needed between the +p+ symbol and the "sempre legato" text if they are to be offset from each other as in the original.

o The !! notation for decorations is now deprecated in favor of ++, since the exclamation is now accepted as a line break symbol.

[MTGuru]

Does it make a difference to the automatic chord generation software

Sure, it certainly should. The chord patterns and cadences inferred by the software analyzing a given melody could (and likely would) be quite different depending on the key/mode assigned.*

Bit IIRC, the only time I played with one of those auto-chord things, the result sounded like a drunken pianist anyway.

* Every ITM accompanist has probably had this experience: You hear an unfamiliar tune for the first time, and after the first repetition you craft and play a lovely chord progression. Only to be told afterwards that the tune is actually in a completely different mode or key from what you were playing (and, obviously, hearing).

So if unsure write

Code: Select all

%Two sharps, unsure of mode
K:^C^F?

An honest approach, to be sure. But frankly, I doubt most casual ITM users of ABC are even aware that you can assign key sigs like that. So it could be confusing. Very useful though for notating Balkan modes, maqams, etc.

[Peter Duggan]

I reckon Bartok in ABC is like teaching a dog to dance the Tango.

For sure. But, as something that just grew out of your comments re. key signatures, it's interesting!

As Johnson says, it's not done well, but you are surprised to find it done at all.

Think you edited your post to add that bit? But, again, you started it with that isomorphism stuff and its ability to do more than it's usually asked!

Still, for a system that started with scribbling tunes on napkins, it's fairly capable as long as you can live with the (mostly reasonable) defaults.

Happily agreed.

It's when you want finer control over placements that things get dicey. You can always manually edit the PostScript output (which I've done) or manipulate it with something like Perl or Python (which I've done), but that's pretty extreme and geeky.

For sure, but sounds just like you from what I know of you so far?

For my limited purposes nowadays - mostly IrTrad tune transcription and an occasional multipart - ABC is great. If I wanted fancier output, I'd import the ABC into e.g. Sibelius or Finale and gussy it up there.

Yep, I'm really just looking at it as this 'trad standard' I'd previously been unaware of... but worth learning properly even for that!

I asked the question about the Bartok on the abc list. Here's a solution that places the key signature correctly

Thanks for that (so at least we've got an elegant canine tango now!).

Every ITM accompanist has probably had this experience: You hear an unfamiliar tune for the first time, and after the first repetition you craft and play a lovely chord progression. Only to be told afterwards that the tune is actually in a completely different mode or key from what you were playing (and, obviously, hearing).

Or the flip side, which is that (as a player of tunes) you work with accompanists who basically play the same thing for every tune in the same key (and I'm not sure I'd really expect much better from software)!

Anyway, to get back to my OP for a moment before I forget, I've been meaning to point out that my reel An Tilleadh gu Colla should be played as what (for want of a better term) I think of as a 'bounce reel'... which means slightly swung in the vein of Phil Cunningham's Cutting a Slide, Jim Sutherland's The Easy Club Reel or perhaps a pipe reel, and never at breakneck speed.

[DrPhill]

Does it make a difference to the automatic chord generation software

Sure, it certainly should. The chord patterns and cadences inferred by the software analyzing a given melody could (and likely would) be quite different depending on the key/mode assigned.*

Bit IIRC, the only time I played with one of those auto-chord things, the result sounded like a drunken pianist anyway.

Thanks for confirming that. But sometimes a drunken pianist is better than none. Altogether now......"My old man said follow the band......."

* Every ITM accompanist has probably had this experience: You hear an unfamiliar tune for the first time, and after the first repetition you craft and play a lovely chord progression. Only to be told afterwards that the tune is actually in a completely different mode or key from what you were playing (and, obviously, hearing).

That is why I allow myself to be a bit pedantic about the mode/key when transcribing to ABC, since I do not want to mislead.

So if unsure write

Code: Select all

%Two sharps, unsure of mode
K:^C^F?

An honest approach, to be sure. But frankly, I doubt most casual ITM users of ABC are even aware that you can assign key sigs like that. So it could be confusing. Very useful though for notating Balkan modes, maqams, etc.

Yes, probably better to let people criticise my ignorance than mock my flaunting of obscure geeky stuff. [crossing with Pete Duggan. I was aiming that geekiness thing at myself.]

[cboody]

Apparently you can place the key signature items by upper and lower case letters, something I thought might be true but did not know and could not find in the little digging I did.

Ah, very clever, Chuck. Yes, uppercase in the explicit key sig is apparently undocumented in the 2.0 spec.

o The !! notation for decorations is now deprecated in favor of ++, since the exclamation is now accepted as a line break symbol.

If I recall the discussion !! is about to come back in 2.1 but I might be mistaken about this. Lots of discussion of lots of things right now...

[david_h]

Is the problem that the K: field is being used for two things, one essential and unambiguous (what the notes on the staff are) one additional information that may or may not be unambigious (the mode) ? Unambiguous if the composer wants to instruct players that he is thinking of the tune in a particular mode, or a collector of a melody could hear that it was being harmonized in the that mode, or the editor of a collection is passing current convention for the tune. Ambiguous if there are differences of opinion or the composer intended players to come to some agreement amongst themselves.

From the information being coded two fields would make more sense, much in the way as the time signature and rhythm can be coded in separate fields. Is that on option in the specification ? Not having it seems untidy to the programmer in me.