Converting 12/8 to 3/4

Hi! I need to convert 38 bars of 12/8 orchestral music (4 beats in the bar) into 3/4 .
My first attempt looked lie an tornado of disasters.

  • Doing a section a a time, setting stop on bar 16
  • Select all (15 bars)
  • Edit > Double Note duration


  • Tuplets
  • Notes out of place
  • Lines out of place

What is the best approach?
Bar by bar?
Small sections per instrument?

3/4 has three beats in a bar, unless you consider it the compound time of 1/2.
Or am I not understanding something?
Can you share a picture of a page of the original music?

Well, it might be 3/4 counting 3 in the bar. Anyway, that doesn’t matter.
It’s a conversion q.=h.

Ah! I get it: you want that each original beat of the 12/8 (thus each 3/8) becomes a full 3/4 bar, correct?
I suggest you play with this by copy-pasting the music into an empty project to avoid disasters.
I don’t know if this changes things, but what about changing meter to 3/8, then engaging insert mode and doing double duration?
Can you share this specific excerpt of the project with us so that we can play along with you?

Yes. Sure. Here are a couple of pages
Coming from 2/4 ca. q=75, attacca

I believe this is how I’ll do it:

  • Make a backup
  • Convert to 3/8 in the backup
  • Set 3/4 in the new project
  • Copy phrases by phrase from old to new project

(I agree with @Michele_Galvagno1 , Copy your passage to a new flow (or project)…)

Handling Tuplets. Mark the start note of each (add a fingering 3). Insert mode on. Select the passage, filter select tuplets. Delete.
Double the the note values. Change the meter.
Recreate the tuplets (Select a fingered note. Hit ; followed by enter)

Note, if any tuplets have incoming or exiting tied notes, cut them before deleting them!.

Any grid attached items (Lines, dynamics, text etc.) must be moved manually.


In all honesty, this shouldn’t really have to be more difficult for the user than a dialog box and typing
“q.=h.” and “3/4” in the appropriate places.
Tuplets and what-have-you should just work…


P.S.: Lovely team, consider this post a feature request… :wink:


Here is an example of what to handle

Double note value:

I just have to rewrite all in such occasions.
Btw. Note-edit lock duration does not work for an instrument like woodblocks.

1 Like

Thanks! I’ll see if I get the hang of working with tuplets this way.

It’s usually a case of take a deep breath - don’t panic - be methodical.

(Bu this method does not work on percussion kit instruments)

1 Like

Managed to do a section of 15/60 bars. Puh
The tuplet teqnique worked ok. But, lot’s of errors with slurs, breath-marks, dynamics.
For us composer having a sleek way of changing metrics is just very important.

Doing a large tutti I see that the Dorico-way of doing this is painfully slow and error-ridden.
I’ll try exporting to Sibelius, convert the import back.

It may prove to be a bit more complicated than that, but I certainly endorse the idea of a built-in dialog to deal with situations like this. They come up more frequently than I might have expected.

Just now, I am starting an arrangement where I can’t decide if it will be best to write as 3/4, 6/8, or 12/8. Because of the difficulty in changing later, I really must make up my mind before starting. It would be much better if I could dive into the arrangement knowing that I could easily transform it to a different meter if I change my mind later.


That is definitely the best way to go.

When you have tuplets it can be more effective to double note values using a sequencer such as Cubase. Then when you import a midi file Dorico will figure out the new tuplets.

Errmmm, this does look like you did not activate insert mode before doubling the note values, have you? Instead chord mode which results in all those overlapping notes.
I experienced problems doubling passages, too, because parts went out of sync during passages of rests, so you have indeed to be careful. There are some insert mode options now as well as a breakpoint that should help, though.
This really should not be a problem, because there are no tupletconversions that need to be done… we often had the problem converting 12/8 to 4/4 with triplets or vice versa, that really does not work automatically yet, because you need a factor of 1.5, but in your case, you don’t need that… I‘ll do some experiments and report back if it seems to work for me.

OK, I definitely was too optimistic here: Dorico screws up every tuplet when doing the math… no idea why (should be as easy as dividing/multiplying the values inside and outside, instead it’s screwing up the rhythms). It screws up many attached objects’ positions as well… maybe we just need a new feature “metric conversion” which really just multiplies a passage as a whole leaving everything as it was.

So converting the meter should currently work like this:

  1. export/print/duplicate your current version as a reference since it will get complicated…
  2. At the end of your conversion, add in at least as many bars as your conversion is long, add some spare for conversion mistakes and temporary changes to prevent loosing notes
  3. Double the meter (in your case 12/4), barlines make visuals unnecessary complicated
  4. set Insert mode break point after your spare bars to make sure everything behind it remains untouched
  5. activate insert mode (global) (the one with the globe…)
  6. select all notes within the passage
  7. force note durations (will help to keep every voice at the correct position)
  8. double note values
  9. deactivate force note durations (probably better for the next steps, but not sure)
  10. set the new meter
  11. set insert mode to voice
  12. correct from beginning to end what has gone wrong
    That last part includes (at least)
  • tuplets ( make sure to have activated Insert mode voice to not lose notes, then delete brackets, correct note values, add brackets again) - DISCLAIMER: Make sure that parts after the point you are working on behave as expected; if not, try force note duration or set insert mode break point
  • slurs (seem to remember their start note, but also don’t change their length)
  • text positions (roughly go with their notes’ position, but not exactly in my experiments)
  • dynamics (just remain where they were before the doubling; probably easiest to completely delete and reenter them)

Finally, delete the spare bars at the end of your new passage. And… hopefully you are ready. :partying_face:

1 Like

Your honor, I rest my case.


Yes, tuplets are a big problem in these situations, as are explicit versus implicit rests IIRC.

I consider this requirement very similar to tonal transposition. The transposition tool works well once one gets used to the not-completely-intuitive UI.

If our friends at Steinberg would consider a “time transposition” tool in the future, I hope it would have the ability to create, remove, and adjust tuplets as needed. This comes up frequently if one wants to change from 4/4 to 12/8 or vice versa. That’s the most obvious case, I suppose, but such a tool could be useful way beyond that.

Another case I have wished for a tool is hemiolas. Let’s say I am writing in 4/4, but have a 3-bar hemiola of 4 groupings of 3 beats. It is perfectly normal for that to continue in 4/4/ throughout. However, after seeing the full context of the arrangement, maybe I think the band will play the effect better if those 12 beats were turned into 4 measures of 3/4. The process today is essentially as describen in the posts above. What I would love to do is:

  1. Select the 3 measures
  2. Invoke the time transposition tool
  3. Indicate “to” time sig as 3/4 and beat equivalency q = q
  4. Press OK and voila.

There might be other options. For example, if I selected partial measures at the beginning or end, maybe there should be an option to either adjust the time sigs in the measures before and after to keep continuous beats, or alternatively for those measures, keep the original meter and fill the voids with rests.

It would also be nice to have the option to select measures from a subset of staves and have the other staves retain their original meter.