Feature Request + Improvement

[FredGUnn]

@mducharme

Even when working in film, if a bar is temporarily incomplete due to writing in the way I described (as one example), you would want to make sure these were all fixed before it was handed off to musicians, because otherwise it would be possible to miss one

Yeah, I’d think that having everything metrically correct would almost be more important in film music as the costs are so high. Even a smallish orchestra of 30 musicians x $120hr + $300 hr studio works out to $65 a minute. Congrats, a 5 minute clusterf**k to figure out why Dorico only has 3 beats in a 4/4 bar just cost the producer $325 and probably got the copyist fired.

Perhaps there could be several automatic options to select from by default. With Dorico’s grid, I don’t think there is an easy way to overfill bars like the Finale option to maintain barlines, but maybe there could be a few other Preferences or Notation Options to select from:

1. Current handling - Rebar overfilled bars up to next meter. Good luck playing Where’s Waldo? with finding underfilled bars. An option to throw a signpost here sure would be nice, but doesn’t currently exist.
2. Complete bars as specified by meter - Rebar overfilled bars up to next meter. Complete final underfilled bars by padding with rests at the end.
3. Complete bars preserving duration (film mode) - Rebar overfilled bars up to next meter. Automatically change meter of final underfilled bar so it becomes metrically correct and matches whatever leftover duration exists after rebarring.

I would think most users would prefer to use options 2 or 3 rather than the current option 1.

[mducharme]

The problem with option 3 as I said earlier is that if you change the time signature repeatedly then you could potentially have a bunch of random autocompleted bars that start piling up. For instance, if you have music written starting bar 30 with proper sync to picture, and now go back to the beginning to work on the first half of the cue and connect the two parts together, you change bar 1-29 from 4/4 to 3/4 and now suddenly you get an incomplete measure as a new bar 30 or whatever and now you get say a 1/4 time signature to complete that. Now you finished writing bar 1 and you want to change bar 2 to 7/8 and now you get another automatic bar in bar 30 with 3/8 time or something. If you keep doing things like this, then when you finally get up to bar 29, you’re going to have a whole bunch of extremely short bars of nothing with weird random time signatures that you end up having to clean up and amalgamate before you can write the last bit. Trying to amalgamate them could be risky as these changes could easily end up shifting the later material so it is out of sync - you would have to write things down on paper and do the math to make sure that as you deleted bars you would create the same number of beats to keep things in the same spot. With the current behavior, because it doesn’t add explicit time signatures automatically, you won’t end up with like umpteen garbage time signatures as you are writing up to that point, instead you will always just have a single measure at the end that might be complete or incomplete as far as beats.

For something like your Option 3 to really work, it would have to use some kind of implicit time signature (the equivalent of a cautionary) as @PjotrB suggested, which doesn’t currently exist in Dorico, instead of the current explicit one. This theoretical implicit one could dynamically change as you made further edits and got closer to bar 30, then you would always end up with just a single bar as the “autocompleted bar” and the time signature of it would change as you work from bar 1 up to 29. It could be displayed as something like a “ghost” time signature, that you could just click to make into a real time signature that would appear on the printed score. But if it was designed like this, it might as well be implemented as a new option 1 instead of option 3 as there would be no reason to not use this.

[Michel_Edward]

ok, I have to ask those who specifically work in film scoring:

if Dorico does this screwing around with beats so that specific spots later in a score remain at the exact same point time-wise as far as video is concerned,
how exactly does inserting a rit or accel affect this?
I can’t see a ritenuto being applied from bars 2 through 4 NOT displacing where specific “hits” are placed at, let’s say, measure 50.

[Lillie_Harris]

It’s (at least the way I understand it) not about time, but about beats: Dorico doesn’t count in bars, it counts in beats. It’s this approach that means you can change the time signature and the barlines move around the notes you’ve already written – something that in many, many situations is very, very helpful.

To preserve the continuous stream of notes and their beat positions, Dorico won’t insert beats unless you tell it to. You can do this, by activating Insert mode prior to changing the earlier time signature.

However, this thread and other discussions indicate that an alternative behaviour where you change the time signature earlier than another time signature (whose position is also beat-locked, essentially) is desired by some users.

You can safely consider this topic within the team’s remit of awareness.

[mducharme]

Hi Michel,

As I mentioned before, a common way of working for film composition is actually to construct a tempo map in advance of doing any writing. This means that any tempos and tempo changes are already in the empty 4/4 (or whatever) bar structure even before any music has been written. So you start off with something like this (imagine something like this but much longer - stretched out in terms of bar length):

As I said, perhaps with dummy notes entered for certain points you want to hit in the on-screen action, just to test the synchronization. But there is no music at this point, although the tempo and tempo changes are fixed. I have a few different tempi and there is a 75% rit in bar 2 lasting 4 beats and ending at the downbeat of bar 3.

Next you would actually compose part of the cue and many composers may work on some important spot first and then lead up to that, so lets say our important spot that we work on first begins in m. 4 (in an actual score this would be more like m. 40 or so but that would be too big of an image):

Now I’ve got the ending to my cue and everything synchronizes up, and I want to write the beginning that leads to that glorious musical moment. But now that I’m actually coming up with ideas, I decide I want the beginning to be 3/4. So I make this change from bar 1-3:

I did not move the rit, it moved automatically when I changed the first bar to 3/4. I just selected it to clarify the location, as the rit that previously fell on the downbeat of bar 2 and lasted four beats now falls on beat 2 of bar 2 and lasts four beats until bar 3 beat 3, because it stayed in the same spot in terms of the overall musical time in beats. This means that even though I have changed this time signature, everything I wrote before in bar 5 (previously 4) and later falls at the correct time in seconds, because the rit still happens exactly 4 beats from the beginning (as it originally did) and is 4 beats in length.

This is exactly the way things work in DAWs and is a big part of the reason why Dorico is so viable for film scoring compared to other notation programs, an area where DAWs have generally reigned supreme.

Obviously making any musical rits or accels after the fact will throw things off, and so when these are added later for musical reasons, it involves some time and trial and error to get later events syncing up again the way they should in the last bit. Also you may decide that you want the rit to happen a bit differently so it begins on a downbeat in the final music for easy readability, assuming that moving the rit slightly won’t affect any hitpoints, and then you’ll similarly have to do some extra futzing with it, but this is the same as in a DAW.

The important thing is in both DAWs and Dorico, composers can feel confident that they can change time signatures, and thereby change where the bar lines fall, without impacting the sync of anything.

[PjotrB]

Your example is slightly ‘too good’: changing three 4/4 measures into four ¾ measures will not leave you with an incomplete measure.

[mducharme]

Yes, I was going to mention that in the text but couldn’t find the best spot. In this case the new bar 4 created by the change to 3/4 in bar 1 happens to be a complete bar, but it could be an incomplete one and cause proofreading challenges. If it was incomplete I would manually add a time signature to it after I got up to it in the music, as I may still add other time signature changes in bar 2 and bar 3 that are not reflected in those images above.

So no, my particular example doesn’t demonstrate the issue brought up in the initial post of this thread, but it was really just meant in response to Michel’s question, which wasn’t necessarily directly to do with that.

[Derrek]

Okay. here’s an alternate approach. Shows the same thing but this time results in an incomplete measure. Not sure it makes much difference to the point being made.

[mducharme]

And sometimes the issue can naturally correct itself out through later edits. For instance, if you then changed bar 2 to 3/4, which results in yet another bar being added between bars 2 and 3 of your image (and now three blind mice begins in bar 5), you still have the incomplete bar as bar 4, but then if you change bar 3 to 2/4, then the incomplete bar of bar 4 becomes complete. This doesn’t always happen of course, but may happen naturally in the course of writing. You’ll figure that out by the time you actually write back up to three blind mice, you might have to set the time signature manually in that bar if it is still incomplete by that point.

[mducharme]

This example allows me to demonstrate exactly how Dorico could make a bit of a mess of things if it were to have the behavior that it automatically set the time signature for incomplete bars by auto-creating an explicit time signature marking. I’ll start with 4/4 in this case, everything is correct:

Now I make the first bar 5/4 like @Derrek did to result in an incomplete bar 3, but now lets say Dorico had a hypothetical behavior where it automatically added a time signature to that to ensure that the incomplete bar was immediately notated correctly. You would get this:

That by itself is OK, but now suppose I want the next bar to be 3/8 so I change it to 3/8, and now I have another new incomplete measure lasting one eighth beat. Auto adding a time signature onto that immediately would result in something like this:

Now, suppose I want a bar of 2/4 next. That would leave an incomplete bar with a quarter rest, which the hypothetical behavior would add a 1/4 time signature onto:

Now I’ve written up close to “three blind mice” but I’ve run into these garbage auto-generated time signatures that have sort of piled up just before it. I’ve got an empty bar of 1/4, an empty bar of 1/8, and an empty bar of 2/4, all of which I would have to amalgamate somehow (and write music in the amalgamated bar(s)) so that I wound up at “three blind mice” at the same exact moment it should arrive. At this point, even though Dorico is generally good about this sort of thing, I would also become slightly worried about the possibility of accidentally doing something that did actually change the timings in the process of fixing this, and messing up the timing of the arrival of “three blind mice” and other specific hit points later in the music.

This isn’t so bad because it is just three bars, but you can imagine how much of a mess there could be if this sort of thing went on for dozens of bars and Dorico auto created 20 random empty bars with 20 different random time signatures through this function. It could also be pretty confusing for the user, who (in the case of a much longer score) may not even know this is happening until they get up to that point in the score and wonder where all these empty bars with random weird time signatures came from.

That’s why the idea of explicit time signatures automatically being added would not really work in practice as you might initially suspect, and nobody would actually want that.

[Romanos]

@mducharme

That’s why the idea of explicit time signatures automatically being added would not really work in practice as you might initially suspect, and nobody would actually want that.

That is, unless Dorico was smart enough to continue updating any “implicit” time signatures that it had itself added, so that the smartest option was employed when the editing stopped. Seems a Herculean task, but they’ve worked their magic before.

[mducharme]

Yes, I said earlier in the thread that this could work if Dorico introduced a new type of time signature as an “implicit” or “cautionary” time signature. However, you would want to have some clear visual difference for any implicit time signatures vs. explicit ones that the user themselves created so that you could tell them apart, and then the user would know that the implicit one could be changed by the program at any point. That’s why I suggested something grey/ghosted out, and maybe these implicit ones wouldn’t even display on the printed score, but you could click on it or something to transform it into an explicit time signature that would be hard set in the music from that point on and wouldn’t change by itself. Almost like a “suggestion” when you are doing a Google search and it autocompletes for you. This would make it quite easy to both spot and fix incomplete measures so that you don’t accidentally send a score with incomplete measures to musicians.

[Steve_Sensenig]

@mducharme

I’ve got an empty bar of 1/4, an empty bar of 1/8, and an empty bar of 2/4, all of which I would have to amalgamate somehow so that I wound up at “three blind mice” at the same exact moment it should arrive

I must be missing something, because you talked about the “timing” of a hit point (such as the start of Three Blind Mice’s melody) and then talked about having extra empty measures.

If, however, all you’ve done is change time signatures, the “timing” of the entrance of Three Blind Mice would not be affected at all. You just haven’t written enough music yet to fill the time you decided you needed. It’s still a total of 12 beats of music at q=120.

You would either need to get rid of the empty measures and change your tempo (slowing it down enough that it still landed you at the point of the entry of the theme, I believe q-85 would get you there) or you need to write more music to fill those beats.

Earlier, you wrote:

If it was incomplete I would manually add a time signature to it after I got up to it in the music, as I may still add other time signature changes in bar 2 and bar 3

And this seems to be exactly what OP was getting at. If you didn’t do the math in your head to KNOW that that bar was incomplete and needed a manual time signature, you’d possibly miss it, and the score would go to print for the recording session with a very confusing measure.

I actually experienced this very thing in a recording session. It was a 15-minute session, so every second counted, and a bar with only 2 beats showed up in a stretch of 4/4 music. The session stopped and they asked the composer whether it was supposed to be a 2/4 bar or if there were 2 beats missing. The poor guy was inexperienced enough that I don’t even know if he understood the question. He just froze and much of the session slipped by as others in the studio tried to figure it out based on the prelay track, etc. It was a disaster and I felt so bad for the composer.

If Dorico had automatically changed that measure to a 2/4 bar and displayed a 2/4 time signature, there would have been no confusion. I understand your example of those incomplete measures stacking up, but as it is, if you did the steps you’re talking about, you would still potentially end up with an incomplete bar and no way to know it unless you thought to count the beats yourself. And if someone ELSE is proofreading your score and don’t even know the steps you took to get there, they might miss it, too.

[mducharme]

@Steve_Sensenig

I actually experienced this very thing in a recording session. It was a 15-minute session, so every second counted, and a bar with only 2 beats showed up in a stretch of 4/4 music. The session stopped and they asked the composer whether it was supposed to be a 2/4 bar or if there were 2 beats missing. The poor guy was inexperienced enough that I don’t even know if he understood the question. He just froze and much of the session slipped by as others in the studio tried to figure it out based on the prelay track, etc. It was a disaster and I felt so bad for the composer.

If Dorico had automatically changed that measure to a 2/4 bar and displayed a 2/4 time signature, there would have been no confusion.

Yes, this is a horrible situation, and this was already pointed out as a problem, and that’s why I was also strongly advocating above for a super obvious warning like a bright red “INCOMPLETE BAR” alert that the composer couldn’t miss.

I also advocated (even in my very last post before this one) for the ability to fill in the time signature for these incomplete bars with a really simple action like a single click.

[mducharme]

@Steve_Sensenig

You would either need to get rid of the empty measures and change your tempo (slowing it down enough that it still landed you at the point of the entry of the theme, I believe q-85 would get you there) or you need to write more music to fill those beats…

And yes, of course you would still need to write more music to fill those beats, but my point was that you probably wouldn’t want to keep them as a separate bar of 1/4 and the a bar of 1/8 and a bar of 2/4. You would probably want to group them together somehow differently like a bar of 7/8 or a bar of 3/8 followed by a bar of 2/4 or vice versa. But it still makes a mess of bars with different time signatures with that theoretical behavior, whereas with the current behavior it just makes a single incomplete bar just before “Three Blind Mice” that you would just have to manually add a time signature to. And if Dorico added some really apparent warning of the incomplete bar that the composer couldn’t possibly miss, and also gave a way to easily add the time signature to this to make it correct with a single or double click, and Dorico also gave an alternate mode of operation via a toggle for composers who didn’t care about maintaining these missing beats like concert music composers and just wanted to make sure all bars are complete, then there shouldn’t be a problem with this behavior.

I should also point out that are some less common cases that the current behavior allows for that you wouldn’t necessarily want to mess up, like the old fashioned way of reducing the final bar by the amount equal to the pickup. For instance, for a piece in 3/4 if you have a quarter note pickup as the first bar, the last bar in old scores would be just two beats of a 3/4 bar, so more like 2/4 but without a 2/4 time signature, in order to “offset” the opening pickup. And of course you might have repeats and things like that for pieces that begin with a pickup, etc. I never use repeats myself, or bother to offset the final bar in this way when beginning with a pickup, so this doesn’t impact me personally, but I do see a lot of music written this way (mostly older). In these cases you will have an incomplete bar on purpose that you don’t want a time signature appearing for, so there still should be a way of handling this. One way would be to acknowledge or suppress the alert for a certain incomplete bar where it is incomplete on purpose rather than by mistake. That way if you were writing something that did have lots of these incomplete bars on purpose, you wouldn’t have these warnings keep appearing because if you got used to seeing them, you could miss them in bars where it was actually by mistake.

There may also be other good reasons for having incomplete bars that I’m not even considering at the moment, like nontraditional time signatures based on hidden tuplets (ex. 5/5). Even though I haven’t written anything like this myself, I know this sort of thing is fairly common in modern concert music (mostly for chamber groups) and tends to require workarounds like incomplete bars or pickup bars in addition to hidden tuplets.

I’m not saying that incomplete bars are good, in most cases you don’t want them and the composer definitely needs to be alerted by the software if they have them to avoid a disastrous outcome in a recording session where time is money, but preventing them outright in all situations could backfire. That’s why it seems safer to me to conspicuously alert, rather than enforcing it to an extent of actually preventing the user from doing something that is wrong (but may be needed in some cases).

[Romanos]

@mducharme

Yes, I said earlier in the thread that this could work if Dorico introduced a new type of time signature as an “implicit” or “cautionary” time signature. However, you would want to have some clear visual difference for any implicit time signatures vs. explicit ones that the user themselves created so that you could tell them apart

Dorico already does this with implicit rests (partly why I used the term “implicit”), so I don’t think there would be any issue here.

[John_Ruggero]

Perhaps on a similar track…

As a newbie working from a long imported MusXML with many flows, umpteen pickup measures and measures divided by repeat marks, I sometimes have difficulty knowing how many beats are actually in a measure. So, I option-click the System Track and count up the divisions displayed so I can correct the measure. Is this the only way to know how many beats are actually in a measure?

[Mark_Johnson]

A temporary empty staff is helpful for spotting undesired incomplete bars. It will show a regular centered bar rest in every bar, except in an incomplete one it will show the rest values.

[FredGUnn]

Is the XML from Finale? Or something else? Obviously you can import it into Finale, run Check Region for Durations, then make the edits in Dorico as Finale finds the errors. If it’s your Finale file that you’re proofing in Dorico, then unfortunately Dorico doesn’t give you much help with this.

@Mark_Johnson

A temporary empty staff is helpful for spotting undesired incomplete bars. It will show a regular centered bar rest in every bar, except in an incomplete one it will show the rest values.

That’s a pretty neat trick! I’m definitely gonna use that one!

[John_Ruggero]

So there is no way to tell how many beats there are in a measure other than option-clicking the System Track?

FredGUnn, in this case, it’s a pdf from outside Finale. I cleaned it up in SmartScore first but there are always issues. Probably something I am not doing correctly, but the pickup measures went in as. as full measures in Dorico, and Dorico has its own way of handling measures divided by repeat marks.

Thanks, I’ll try that trick in the future , Mark_Johnson, but if that is what is necessary, then that might have answered my question above.