8

I know I can write them as metrical modulations but I prefer to write them like this: enter image description here

...which Sibelius can't do. I have to switch to a graphics program and it's time-consuming.

The main advantage of writing them this way is that tuplet-relationships are familiar: you don't have to stop and work them out. You just have to remember to stop tapping your foot! Using metric modulations makes the page untidy.

Any suggestions? There's something for Mac called NoteAbility that seems to know about these time signatures, but it doesn't look very sophisticated otherwise.

Old Brixtonian
  • 5,533
  • 12
  • 24
  • 1
    Does Sibelius allow fractional time signatures? Allowing I understand your intent (a dubious proposal), could you change 7 / 6 to (2-1/3) / 4 – Aaron Jul 14 '20 at 05:45
  • I know some composers have done that, but no: Sibelius doesn't allow it. A good idea though. Thanks. – Old Brixtonian Jul 14 '20 at 05:53
  • 2
    Musescore (musescore.org) can do it. Their time signature feature allows you to overlay the "actual" time signature with custom text. It's very easy to do. Just know, I couldn't discover how to create the incomplete tuplets at the ends of your 7/6 measures. – Aaron Jul 14 '20 at 07:12
  • Can anyone even read this? I mean, 6th notes don't exist afaik. There must be clearer ways to express your intent. The notation inside the 7/6 bars doesn't even make sense: there is a triplet at the end which has only 1 8th eight note and 1 8th rest in it. If it were a triplet of 8ths then it would need three 8th note values. – herman Jul 14 '20 at 08:29
  • 2
    Also, in any case these are not "irrational" as that would mean they would involve numbers like Pi or Sqrt(2) or other non-rational numbers. Don't do that :) – herman Jul 14 '20 at 08:32
  • 2
    Ok a 6th note is actually a 4th note in a triplet, I get that now. – herman Jul 14 '20 at 08:46
  • 1
    Btw do you just want it to look like this, or do you need the software to understand this semantically (like play it back correctly without hidden tempo workarounds etc.)? If not then Dorico can do this. – herman Jul 14 '20 at 08:58
  • Given that you use the notation to convey to the performer what you want done it would seem logical to write in the way that is simplest to understand. Your meaning - in terms of note duration - looks very unclear to me. A change of time signature and an indication of how the beats in the two time signatures relate would be easier to understand and make your music more approachable. The end result should be the same. Why make it harder than it needs to be? – JimM Jul 14 '20 at 11:05
  • @herman I only called them that 'cos that's what they're called! And I put the word in scare quotes to show I don't like the term! Jeez! :-) – Old Brixtonian Jul 14 '20 at 12:08
  • @herman Thanks. No - it doesn't need need to play it back. I can do that in Cubase. Dorico? Wow. Haven't looked at it for a few years. – Old Brixtonian Jul 14 '20 at 12:11
  • @JimM Thanks. For musicians this way IS the easiest to understand. As I said, you don't have to stop and work them out. They SHOW you the relationship. Metric modulations make the page cluttered, and you DO have to work them out, and they can't be sight-read because they don't show the relationship. I AM trying to make it as easy as possible! Been doing it this way in hand-written stuff for years :-) – Old Brixtonian Jul 14 '20 at 12:24
  • @OldBrixtonian I guess in tearms of readability you could also choose write the first 6 beats of the 7/6 as regular 4/4 and then add a bar of 1/6. Since they're notated as triplets in 7/6 anyway it doesn't change the notation, and everyone's familiar already with 4th triplets in regular 4/4. I guess that also depends a little on whether all instruments are playing the triplets. If some are playing regular 4th notes I'd personally write it as 4/4 + 1/6, – herman Jul 14 '20 at 12:49
  • @herman That would be possible certainly, but not in this particular case. There's a different chord on each tenuto note, so the phrasing is more like 1/6 + 3/3 isn't it? And I could make it that. That's how it would be conducted. BUT 1/6 + 3/3 is a more confusing time signature than 7/6. (No-one IS playing in regular 4/4 btw.) I could change the bar before into 4/4 + 1/6, but I don't think that's any clearer. I really appreciate your thoughts. – Old Brixtonian Jul 14 '20 at 14:12
  • 1
    @herman, your initial objection to 'irrational' got me thinking ... someone should really clarify that it's not 'irrational' in the mathematical sense; it's 'irrational' in the 'only a crazy person would ask someone to play this' sense. – Aaron Jul 15 '20 at 06:31
  • @Aaron yeah but we're talking about mathematical ratios here, so using that term with a non-mathematical meaning is very confusing. Just call them "crazy time signatures " then :) – herman Jul 15 '20 at 07:44
  • I see you’ve already accepted an answer about using Dorico, but I’ll add that I have definitely done this in Finale, and I suspect similar methods would work for Sibelius, although I don’t use that program. The Maestro font includes time signature-style version of every number, so you just build the signature as an expression attached to the beginning of the measure. Then, you go into the measure attributes and add an appropriate amount of extra space at the beginning of the measure. Again, I don’t know Sibelius specifics, but I’d be surprised if it doesn’t have similar flexibility. – Pat Muchmore Jul 15 '20 at 14:13
  • Pat, I really appreciate your help. Haven't got time to try it till later, but I'll get back to you. I 'm sure I can find whatever font Sib uses for time signatures and, hopefully, build one as you suggest. There's no 'bar/measure attributes' exactly, but I can create a bar, hide its rest, drag the time sig into it and hide the end barline. I think so anyway. It sounds feasible. Thank you. Speak later. – Old Brixtonian Jul 15 '20 at 15:19
  • @Pat Muchmore I've cracked it. You were right. Sibelius CAN do it, AND play it back correctly. Enter the notes as normal triplets into a 4/4 bar followed by a quaver into a 1/4 one, then hide the barline and delete the time sigs, so that it looks like a single bar. Add space at the start of the 4/4 bar using a hidden grace note. Make 7/6 time sig with 'special text tuplets' font. Use an extendable triplet bracket with a hook at only its left end, and drop a '3' on it with 'magnetic layout' switched off. Use metric modulation changes to make the invisible 1/4 bar last only 1/6, and hide them. – Old Brixtonian Jul 19 '20 at 02:09
  • @Pat Muchmore I know you're not a Sibelius user but I thought I should answer my own question somewhere here in case anyone ever needs to do it. (Yes, very unlikely!) Thanks again for making me try harder! – Old Brixtonian Jul 19 '20 at 02:11
  • @OldBrixtonian Sure, no problem. Actually, you can provide a full on answer to your own question, and I think this might be a time where it’s appropriate. You don’t need to supplant the accepted answer, but future people searching might find it helpful, but are less likely to look deeply into the comments. – Pat Muchmore Jul 19 '20 at 02:14
  • It's very confusing for a musician, so he or she would spend a lot of time trying to figure out the meaning. There are 7 quarter note triplets in the bar which could relate to the number 7 in the time sig 7/6. But there is nothing that relates to the number 6. How can 7 quarter note triplets be the same thing as 7 sixth notes? Doesn't make sense. Let alone what is a sixth note? There is no indication of something that could be labelled as sixth notes, well maybe the eight notes as a subdivision of the quarter note triplets but that wouk give 14 eight notes. Big mystery. – Lars Peter Schultz Jul 19 '20 at 22:53
  • @Lars The number at the top - as it always does - tells you how many. The number at the bottom - as it always does - tells you of what kind. So, in the example, there are seven sixth notes in a bar. Seven is how many. Sixth note is what kind. You have always called them quarter-note triplets, but if you think about it you'll realize they could also be called sixth notes. In other words, it follows the existing rules of music. Eighth-note triplets could also be called 'twelfth notes'. Just ask yourself 'how many of these are there in one whole note?' and I think you'll understand it. – Old Brixtonian Jul 20 '20 at 01:43
  • @Brixtonian Ok, I see what you mean. I strongly suggest that you include an explanation in the start of the score what you mean by a sixth note. – Lars Peter Schultz Jul 20 '20 at 13:40
  • My question was only to do with finding music notation software that would handle them correctly. I didn't think I would end up explaining them and defending their existence. I learned about them nearly half a century ago but thet seem to be better known in Britain than elsewhere. The composers Brian Ferneyhough and Thomas Adès both use them, and their music has been played in the US, so there are musicians there who may well have come across the things. But If I am ever concerned that musicians might not know what they are, I will indeed explain what sixth notes, twelfth notes etc. are. – Old Brixtonian Jul 20 '20 at 17:08
  • 1
    My experience with NoteAbility is that it can do just as much as the others but trying to adapt to the workflow slowed me down too much. There’s not much of a community to help you. The developer used to have a purely graphic mac only music typesetter called Notewriter which was incredible, but was discontinued with Os X. It’s the inclusion of midi with NoteAbility (its successor) that hampers it. –  Sep 18 '20 at 09:35
  • 1
    I personally like pi/e as a time signature. Flooring both numbers gets you 3/2, and it's awfully close to 1/1. –  Sep 18 '20 at 16:27
  • when you write "score writer" do you mean "notation software?" – Michael Curtis May 25 '22 at 17:14
  • @Michael Curtis: Yes I do. I see I started calling it "notation software" three comments up from here. I believe I use both terms interchangeably, even dipping a toe - on special occasions - into "music-engraving software". Are you suggesting I change the question? I've no objection if so. – Old Brixtonian May 27 '22 at 11:03

4 Answers4

4

I discover it CAN be done in Sibelius. AND it will play back correctly with a couple of hidden metric modulations.

First input the notes as normal crotchet-triplets into a 4/4 bar followed by a normal quaver into a 1/4 bar, then hide the barline and delete the time sigs, so that it looks like a single bar.

[Turn on 'View hidden objects' if you like.]

Add some space at the start of the bar using a hidden grace note. Switch its playback off.

Click the R. button in space above the staff, and from the menu that appears select text|special text|time signatures. Then click the L. button where you want the time signature and enter '7' then '6'.

Now to make the final incomplete-triplet's bracket.

Select the final quaver in the bar and press 'L' to open the Lines menu. Scroll through it and select the bracket without a final hook. It's named 'bracket above (start)'.

Position it where you want it, and lengthen or shorten it by dragging the end of its line with the mouse. Now right click above the bracket and navigate to text|special text and click on 'tuplets'. Then left click and type "3". The 3 appears and you can drag it into position on the bracket line.

[Depending on your set up it's possible you may need to switch magnetic layout on or off.]

Finally use a new tempo for what was the 1/4 bar to make it last only 1/6, and another one to set it back to the preceding tempo. Hide these tempo changes. Sibelius will now play it correctly.

As Sibelius still sees the 1/4 bar as separate, the bar-numbering will need to be corrected.

[Turn off 'View hidden objects' if you like.]

Old Brixtonian
  • 5,533
  • 12
  • 24
3

Lilypond handles this without fuss if you can do without a GUI: you just write \time 7/6 and enclose the notes in \tuplet 3/2 { ... }

\version "2.20.0"
\relative c'' {
  \numericTimeSignature
  \time 4/4
  e4 b8. c16 cis8 cis gis4 |
  \time 7/6
  \tuplet 3/2 2 { 
    bes4 bes f8 r 
    g!4 d8 r e4
    b!8 r
  } |
  \time 4/4
  cis4 fis8. f16 e8 e a4 |
  \time 7/6
  \tuplet 3/2 2 { 
    g4 g c!8 r 
    bes4 ees8 r des4
    ges8 r
  } | 
}
  • So \tuplet 3/2 {\time 7/6}? – Aaron Sep 17 '20 at 22:16
  • No, eg the OP’s second bar would be \time 7/6 \tuplet 3/2 {bes’4 bes’ f’8 r8 g’4 d’8 r8 e’4 b8 r8} –  Sep 17 '20 at 22:37
  • Added minimum working example. And it works, tuplet brackets should be adjusted (span too small) but otherwise effective. –  Sep 17 '20 at 23:34
  • 1
    yes, to get the tuplet brackets and spacing as desired you have to add the usual types of formatting instructions - my general point to the OP being that the use of a time signature like 7/6 doesn’t add any significant difficulty when typesetting with Lilypond. –  Sep 18 '20 at 08:02
  • Thank you, Damian. Lilypond looks really good. I don't know if you saw my own answer, but I did eventually work out how to do it in Sibelius, and although the solution was complicated to describe, actually using it doesn't take long at all. So I'll stick with the devil I know. – Old Brixtonian Sep 18 '20 at 14:40
2

With some workarounds, Dorico can do this. The workarounds are necessary for the incomplete triplet at the end of the measure. There is a custom line tool with which you can create this by hand. It won't play correctly though, unless you add hidden tempo changes.

See this forum post.

herman
  • 203
  • 1
  • 4
  • That's really useful. Thanks. Yeah - hidden tempo changes. No problem. As I said I don't need it to play back, but if it did that'd be a bonus. I'd better spend a few hours in Sibelius/Dorico hell (as I do every few years) and see if Dorico has caught up with Sib enough to be useable, or if there are any plans for Sib to incorporate these time signatures. (Very unlikely I think.) – Old Brixtonian Jul 14 '20 at 12:32
  • Well - in the end I found Sibelius CAN do it. AND it plays back correctly. It's a complicated workaround but at least it can be done. Thanks again for your input. – Old Brixtonian Jul 19 '20 at 02:17
1

Since you've given no instruction as to the relationship between a quarter note in the first measure, and a (no-such-thing) "sixth note" in the second, this notation is useless to a performer in the first place. You have nothing to indicate how to map the standard icon for a quarter-note and eighth-note into "sixth-note" timing.

I can honestly say I do not know any musician who would want to try to read that and figure out the metric. Figure out how long your "7/6" measure will last in terms of the 4/4 measure, and notate triplets or sixths or 3-beat quads as necessary.

Carl Witthoft
  • 14,931
  • 2
  • 19
  • 43
  • 1
    It's quite clear to me, 4/4 lasts 4 quarter notes and 7/6 lasts 7 lengths of quarter triplet note (this is what division of a full note into six parts is). – user1079505 Jul 14 '20 at 19:51
  • @Carl Witthoft Evidently you haven't come across them before :-) Check out the link in Hermann's answer above. Or this one: http://www.paulsteenhuisen.org/non-dyadicirrational-time-signatures.html – Old Brixtonian Jul 15 '20 at 00:09
  • 2
    @OldBrixtonian Yes, I have, however they are purely composer-wanking and from a performer's point of view, highly obfuscatory. Kill them. Kill them dead and burn their bodies (the time signatures, not the composers) – Carl Witthoft Jul 15 '20 at 11:08
  • @Carl Witthoft It's so hard to tell what you really think :-) The excellent thing about the (ill-named, as I have admitted) irrational time signatures is this. You play that first bar, then you come to some crotchet triplets. You know how to play crotchet triplets because even composers you like used them. So you play six of them, then notice there's an extra one: a seventh one. You remember passing a funny time signature with a 7 at the top. "This must be the seventh one", you say. And you play it. Easy, for any musician worth his salt. In the taxi home, you work out what the 6 means. – Old Brixtonian Jul 15 '20 at 14:09
  • @OldBrixtonian But I know how to play triplets when I see eighth-notes marked as triplets in 4/4 time . When someone says "the beat is a sixth-note" that means I have a completely undefined meter relative to the previous measure. If you want 7/8 time then do that and put in a text marking indicating that the eighth-note is half the time period of the previous measure's quarter note. "What the 6 means" is already strictly defined: X/Y time means X instances of Y-length notes. – Carl Witthoft Jul 15 '20 at 15:09
  • @Carl Witthoft (Why did you switch to eighth-notes rather than quarter notes? any significance?) – Old Brixtonian Jul 15 '20 at 15:31
  • 1
    Don't know if this is helpful or not but: imagine the melody written out above, with no bar lines, and no time signatures. It's perfectly unambiguous how long each of the notes are in relation to each other. Adding barlines doesn't change that, and saying it means there's "a competely undefined meter relative to the previous measure" is simply untrue. That is to say, there is no such thing as "triplets in 4/4 time" as distinct to "triplets in any other time signature". That's like saying a dotted crotchet in 11/8 is different to a dotted crotchet in 4/4. It isn't. It's 3 quavers long. – Judy N. Jul 15 '20 at 20:15
  • 1
    And absent of any specification to the contrary, i.e. absent any specification that relative lengths of notes in the new bar have changed in relation to lengths of notes in the preceding bar, everything just happily continues at the same pace, regardless of what time signatures we may have passed; the time signatures express the size of the bars and indicate the principal subdivision, but they have NO EFFECT on relative durations of notes unless this is separately specified – Judy N. Jul 15 '20 at 20:19
  • @Judy Thank you. That was a good clear explanation! Sorry for the delay in replying. I've solved it now and it CAN be done in Sibelius with a bit of jiggery-pokery. AND it even plays back correctly. I've described the solution above btw. – Old Brixtonian Jul 19 '20 at 02:25
  • 1
    Mmm...I’ve been playing, directing and conducting time sigs with 6, 12, 5 etc on the bottom since the 80s and honestly have never come across a musician who needed more than a few seconds to get their head around the notation; "what’s 5/12?" "Five triplet eighths" "err...so like three triplets plus another two at the same speed" "yup" "ok I get it" –  Sep 17 '20 at 21:43
  • Exactly. Musicians are generally amused/interested to encounter an easy time signature they've never imagined. [I'm enjoying that Icebreaker track that was mentioned the other day. It's right up one of my streets! Looking forward to exploring you further when time allows!] – Old Brixtonian Sep 18 '20 at 14:48