Dorico 3 feature: Condensing

Reviews

This is part of a series of in-depth reviews about each of the major features in Dorico 3, in addition to our general review.

First things first: Do you already have Dorico 3? If you do, go launch it right now and open the full score layout of one of your orchestral projects. If you don’t have one at the ready, download this excerpt of Brahms’ fourth symphony. Choose Edit > Condensing, and watch what happens. Go ahead, we’re not going anywhere.

Ready? Picked up your jaw? Off we go then.

Introduction

What is condensing? In a nutshell, condensing is the process of combining the parts of multiple players on a single staff. This is a crucial step in the preparation of usable conductor’s scores because it helps save precious vertical space and improve legibility.

Condensed wind sections in the score of Brahms’ fourth symphony

OK, you say, this doesn’t sound overly complex. What’s all the fuss about? You simply take Horn 1 and then you take Horn 2 and then you print both voices on the same staff, right?

You see, the beauty in Dorico’s solution is that it makes it look like there’s really no more to it than that.

But don’t let the apparent simplicity fool you. With an undertone of understandable pride Dorico product manager Daniel Spreadbury states that making this happen has taken years of research and development. Indeed, the editorial decisions that Dorico has to make on its own are of staggering complexity:

Can I condense the flutes in this phrase? Wait a minute, what is ‘this phrase’? What chunks of music should I consider to begin with? … OK, looks like I can condense something here. How far can I go, separate voices, dyads, unison? What about dynamics, slurs, articulations, playing techniques, lyrics, breath marks… and look, things look different altogether three bars later… OK, that’s it for the woods. Next, the horns. Oh crap, there’s six of them! 3+3, 2+4, 4+2,…??! I’d better find a good way to indicate who plays what. Which reminds me, did I take care of the staff labels for the flutes? … [fades out in the distance]

Let’s do it the Dorico way

Dorico is not the first scoring software that tackles condensing. But it is the first to offer a viable solution. Why is that?

It’s a matter of direction. Technically speaking, the reason is that Dorico’s architecture and data model are fundamentally different from its predecessors, allowing it to… well, do things the right way.

Source music and condensed view of a clarinet snippet from Mozart’s KV 550

In older notation software, the full score, condensed or not, is the primary musical data. Linked parts were invented much later, so they came to be derivatives of the full score. This relation leads to vexing problems when a bit of music needs to be represented differently between score and parts. It also turns the source-derivative relationship between normal and condensed music upside down. After all, the ‘natural’ notation of the music is the uncondensed part that players see on their stands, not that squished conductor’s score.

Groaning under the weight of decades-old design decisions, the established scoring applications are, we’re sorry to say, really unable to condense anything. The best they can offer are halfway-decent algorithms for the other direction: disentangling their users’ condensed scores during the creation of linked parts or leaving the job to third-party plug-ins — useful in certain circumstances, but leaving lots to be desired.

On the other hand, scores and parts are hierarchically equal in Dorico: independent renderings of different vertical excerpts of the music, called layouts. Therefore Dorico is not forced to approach condensing from the wrong side altogether. The user can freely input the source music just as the players need to see it in their parts, without ever worrying about squeezing things together.

Dorico will take care of the rest because, ultimately, condensing is just another way to compile abstract musical data and transform it into beautiful notation — precisely Dorico’s strong suit. But enough of the theory. Let’s get our hands dirty and do some condensing, shall we?

The basics

As already hinted at the start of this review, enabling condensing in an existing Dorico project could not be simpler. It’s easily done via the quick toggle Edit > Condensing, and to more easily toggle back and forth you can assign your own shortcut to this item in Preferences > Key Commands (Dorico does not assign one by default). You must be in Page View to see the condensed results; Galley View is always non-condensed.

You can also open Layout Options, choose your Full Score layout, and select the new Players > Condensing section. There you’ll see a single checkbox that will enable condensing in your score.

Wait a second or two while Dorico performs an incomprehensibly large number of calculations and optimizations behind the scenes and you’ll be rewarded with an automatically condensed score.

What if you’re starting a brand new project? In that case, simply enter your music as normal onto each individual player’s stave in a non-condensed score. Dorico takes care of producing the condensed score for you whenever you want it. Edit > Condensing will switch between condensed and non-condensed views. You could even choose to use one non-condensed Full Score layout for note entry, and have a second condensed Full Score layout for the conductor’s score.

Once you’ve got your condensed score, you can tweak it in Engrave mode: either by adding system or frame breaks as needed, or by making graphical tweaks to the appearance of individual notation elements. In this way you can make sure every “a2” label is positioned to your liking, and every slur looks perfect.

What can condense

By default, Dorico assigns each identical adjacent solo instrument in the layout to a condensing group. For example, Clarinet 1 and Clarinet 2 would form a condensing group, as would Trombone 1 and Trombone 2. This covers the majority of common condensing cases and produced good default output on a variety of orchestral scores we tested.

Finer control over which instruments condense is available through custom condensing groups, which will be explained in the next section.

Currently, section players cannot condense. So you cannot, for example, condense together Violin 1 and Violin 2 in an orchestra. Dorico won’t condense divisi that are created in a section player, either; you’ll have to manage those the old-fashioned way, for now.

Wanting to condense different string sections is unusual, but you might want to condense together sections in large wind or brass ensembles. For now, the workaround for this is to create these as solo players instead of sections. It’s easy to convert existing scores — in Setup mode just drag the section instrument to a new empty-handed solo player to take advantage of condensing.

One special case of section condensing is supported: section players with vocal instruments will condense if you add them to a custom condensing group in Layout Options. Yes, you can write for SSAATTBB choir on eight separate staves and have Dorico automatically generate an SATB reduction where Soprano 1+2 share a stave using separate voices. Rehearsal pianists, rejoice!

You can also write for SATB on four separate staves, give that to the singers, and have Dorico generate an SA+TB reduction for the conductor automatically. The only current wrinkle with this is that Dorico always uses the clef (and transposition, for that matter) of the first condensed instrument. In the case of condensed male voices you’ll end up with a tenor clef, not bass, as is standard practice. More control over this is promised in the future.

The individual staves as entered…
…and automatically condensed by Dorico 3.0. The music condenses, but Dorico incorrectly chooses the transposing treble staff for the men instead of the customary bass clef.

Once this is sorted out, automatic condensing of vocal staves promises to be game-changing for large choral works (including musical theater), as it will be possible to maintain a full score, vocal score, and the instrumental parts all in the same Dorico project, with all layouts kept automatically up to date and in sync with each other. It is hard to overstate what a revolution this will be for music preparers.

Layout Options

As shown earlier, you can enable condensing in Layout Options > Players > Condensing.  You can also use Layout Options to control the condensing groups, in case Dorico’s defaults don’t quite produce the result you want.

For example, you could create a custom condensing group if you wanted to condense trombone with tuba, or bassoon with contrabassoon. You could also create a custom condensing group if you wanted French horns to be condensed as 1+3 and 2+4.

In Layout Options you can also exclude any of the default groups from condensing. This might be useful, for example, if you wanted to condense all the woodwinds but leave the trumpets on their own staves.

Currently, the condensing groups defined in Layout Options are fixed for the whole layout and cannot be modified on a per- or mid-flow basis. Finer control is promised in the near future.

Editing condensed music

Here’s something that you should really try to grasp, particularly if you’re migrating from other software: condensed music in Dorico is a special view of the source music, not the source music itself. It is therefore impossible to change any music in its condensed state. You can’t even select it in Write mode.

If you need to edit condensed music, you are free to do so in Galley View which always shows the source music, i.e. each part on its own staff. Alternatively, and in line with Dorico’s ethos of clean separation of ‘procedural stages’, keep condensing switched off until you’ve finished the content of your masterpiece.

Here’s a split view showing both condensed (Page View) and source music (Galley View). Activate View > Note and Rest Colors > Condensed Music to get a visual hint as to which music on the screen is editable.

Needless to say that this limitation applies to note input as well, though many of you are likely doing that in Galley View anyway. More recent converts may initially frown at the fact that they cannot compose and arrange directly in the condensed view. If you prefer, you can of course input two or more voices into single staves and — boom — explode them later for — whoosh — Dorico to re-condense them.

However, you might find this little feature more useful: it is now possible to vertically extend the caret so you can input the same notes (and pretty much everything else) on multiple staves at the same time. If you play chords via MIDI input, the notes are automatically distributed among the selected staves. Look — isn’t this just cool?

For the sake of completeness: it is of course possible to make selections and edit the appearance of condensed music in Engrave mode. As expected, most properties that you set on condensed notes there will affect the condensed view only so you can tweak the curve of a slur or the position of a dynamic freely without worrying that your changes might carry over to the single-staff representation in the part.

You’ll quickly get used to writing and editing uncondensed music only. There is, however, one severe limitation. In many cases Dorico’s automatic choices regarding how specifically music is condensed are good. In some cases though they’re not, and it is, regrettably, impossible to adjust the results at all. At present, there is a possibility that the feature will turn out useless for a particular project because you’ll find yourself stuck with unpublishable default results. Careful proof-reading is a definite necessity for condensed scores.

Seeing that condensing is a groundbreaking technological development, this is not surprising. The core mechanisms are functional, and additional bits of the interface have been promised. In fact, we’ve had word from the developers that they’re already working on ‘condensing changes’ — local events that will allow users to override the global rules for a specific region, very similar to the existing note spacing changes. We don’t have a date, but we hope it will come reasonably soon.

How music is condensed

Dorico condenses music for solo players only, and does so by using an algorithm designed to divine how the music is organized into phrases from left to right. Once these phrases are determined, Dorico looks at them separately and then chooses condensing methods from a list of four different approaches: unison (a 2), shared stem (single voice), shared staff (multiple voices) and “no condensing” (separate staves).

Additionally, when music is condensed as two voices on the same staff and if the beginning of phrases between the two voices align, notes in the down-stem voice that are the same duration as those in the up-stem voice and are lower in pitch can be amalgamated into the up-stem voice.

Bearing these rules in mind (and the release notes dedicated to condensing must be read with great care), we have put the condensing feature through various tortures in order to test it. While condensing is still in its infancy as a public feature, the results are impressive enough to use with several existing scores. More complicated projects will fare better once the ability to execute condensing changes is implemented in a future update.

If we look at the following example of an uncondensed score:

The wind and brass from Beethoven’s 7th symphony, mvt. 1, entered as separate players in Dorico 3.0

We can compare it to its condensed sibling, having done nothing but invoke Edit > Condensing:

The result of Dorico 3.0 automatically condensing the music

Here we can observe several of the choices Dorico makes according to how it interprets segments of music. The first chord, except for the oboes, shows the “shared stem” approach because the music is rhythmically identical. As the dynamics are identical, they too are consolidated into a single instance.

The horns in bar 5 use the “unison” technique and are labelled “a 2” as a result. Should that passage be interrupted by a system break, a reminder “a 2” will also appear at the start of the new system.

The clarinets have a similar passage and here one might prefer the same as the horns, but Dorico reads the longer passage as one phrase, and that longer phrase — unlike that of the horns — includes a section with independent voices using the “shared staff” concept; therefore, Dorico chooses to have the clarinet unisons display as separate voices.

Some editions have the horns displayed in that manner as well, but this cannot be altered at the moment since Dorico reads the horns as having perfect rhythmic unison, dynamics, articulations and so forth. Until we can choose condensing methods on a more localized basis, we have to live with Dorico’s choices, but those are generally quite wise.

The flutes in bar 7 display Dorico’s power in choosing what gets notated and how. As the flutes now share a staff, the crossing of voices between the first and second players can be accomplished (there is a notation option giving you a great of global control over that behavior) and the dynamics are also shown as being different — in this case according to 1-2 order. Those dynamics can be flipped manually in Engrave mode if one would rather see the reverse. Also, the flutes are clearly labeled as 1 and 2 to eliminate any confusion.

It is not only dynamics that get condensed or separated when using a condensed staff. Text and playing techniques follow the same rules. As one can see, these are very beautiful, logical results that are accomplished automatically.

Be aware that Dorico can still overuse bar rests if the option of pairing inactive players with active players is chosen (note that rests cannot be manipulated on condensed staves yet); and that Dorico will at times label proximal “a 2’s” too often. In the latter case, fortunately, there is a possibility to hide unwanted labels through the Properties panel.

Another limitation at present is that players holding more than one instrument can only have the first listed pair of instruments condensed. This is to be expected for a player holding both a flute and a piccolo, but is unfortunate for any work where clarinetists have to change between clarinet in B-flat and clarinet in A. The same applies if one wants both flutes to play piccolo later in a score: each piccolo will then remain on their separate staff if they are both listed in second place on their player’s card. We have been assured, however, that this will be addressed in due course.

Notation Options

A new Condensing section in Notation Options offers control over a few key condensing decisions on a flow-by-flow basis.

Pitch crossing approach is relatively self-explanatory: it governs whether condensing should happen when two instruments’ pitches overlap.

The Amalgamation options don’t determine whether phrases should be condensed or not, but they should still be adjusted with due caution. For example, in SATB works it’s quite usual to have sopranos and altos condensed onto one stave and tenors and basses on another, but hands might go up in the rehearsal room if Soprano and Alto notes share a single stem. Thankfully, lyrics automatically prevent the amalgamation of two voices onto one stem.

The option for Inactive Players governs what should be done with resting players. At the time of writing, there are some issues with spurious bar rests appearing, but we’re told that there will be better control over this in a future version.

Labeling condensed music

As you’d expect, labeling of condensed passages is handled globally in Engraving Options, but can also be overridden in the Properties panel.

In our experience, Dorico does a good job of positioning condensing labels by default, but it’s easy to adjust these globally in the same way you’d do for general text, playing techniques or lyrics.

You’ll also find an option here for the size of background padding (switched on by default), options for whether periods and spaces should append player numbers, and custom text for the “to” (a) indication.

Note that the options in Engraving Options > Condensing only affect labels appearing as part of the staff, and not the labels to the left of the system…

… while for labels to the left of condensed staves, there are new settings in Engraving Options > Staff Labels:

If you want to change the font or size of instrument labels on condensed staves, head over to Engrave > Paragraph Styles, where you’ll find a new Condensation Labels text style.

Condensation labels can also be overridden locally from Engrave mode. You can click and drag labels, and from the Properties panel you can hide them, scale them, or override their text.

There’s also a Line Break property which does whatever it can to horizontally shrink condensation labels. In the case of a “1.2.” label it will put a line break between the two numbers, but with an “a 2 1.2.” label it sensibly inserts the line break after the “a 2”.

Summary and outlook

Condensing in Dorico will dramatically change the workflows of composers, arrangers, editors and engravers. For the first time ever, we are presented with a fully working automatic solution to one of the most time consuming tasks in our trade. What used to be a combination of intricate editorial work and potentially maddening fights against the software has the potential to be reduced to ticking a single checkbox. The labor of days, even weeks, done within seconds. For everybody involved in the production of orchestral scores this is no less than a revolution.

The real revolution though was a silent one. It happened some six years ago in a scarcely furnished office in central London, where a dozen highly proficient software developers sat down to devise the architecture of a new scoring program. They discussed for weeks, hardly ever touching a computer. In the end they turned away from many previous concepts and modeled their future application closely around the real world: human players holding instruments, playing music. Technically speaking, each player would produce a continuous stream of ‘sonic events’ on a global timeline. Naturally, notation would be the primary representation of (and interface to) these streams. The program would generate scores and parts from them: on command, for any thinkable combination of players, and in accordance with a myriad of user defined settings that would govern the tiniest details.

If creating beautiful notation from multiple streams of musical events is Dorico’s core ability, then condensing is, in a way, no more than that ability applied to its most advanced use case so far. At last, it is now fully evident what the people behind Dorico meant when they boldly called it ‘the new gold standard’ at the time of its first release. The inclusion of condensing as a fundamental feature of modern notation software is now a marker against which the rest of the field must be measured.

To temper the excitement a bit, there’s no denying that with this first iteration of the condensing feature they’re ‘not quite there yet.’ There are hiccups and a few frustratingly insurmountable limitations. We know, however, that the most needed addition, a means to adjust Dorico’s automatic choices, is already in the pipeline, and the Steinberg scoring team has a track record of making good on their promises.

As for the rest… well, this is new territory. As stated in the version history (do read it, by the way), the developers have ambitious plans, and they look forward to discussing the feature with their users. Since experience shows that they do listen to us, we may all get the chance to shape the future of this most important innovation in our field since the introduction of automatic collision avoidance. These are exciting times!

Alexander Plötz helped edit this post.

Comments

  1. Waldbaer

    This is extremely interesting, thanks for explaining this in detail! I really wonder, though, why they did not (yet) include a feature to change the automatic mode in manually selected regions – this will fix most complaints on the automatic output and I consider it as essential as you do in your review… probably there were bugs occuring at the selection’s borders and it would have taken another week to fix them.
    So maybe I’ll stay with Dorico 2 until the update including this. Automatic Condensing definitely is a unique selling point, but at the moment it’s more of a big tech demo in my eyes than a useable functionality (“useable” should imply doing something with it, I think). Maybe some Yamaha managers could not wait to put this out, but I can. ;-)

  2. Alexander Blank

    In the demonstration video (https://youtu.be/H2-1JTfRLrc, at around 1:14), I noticed Dorico’s label for the first flute and oboe essentially get lost against the baseline (the tiny amount of masking wouldn’t be nearly enough to make it stand out). Is it possible to change the default horizontal placement of these labels, or tell Dorico to add an extra space of padding between the barline and the start of the label? (I.e., without having to do it manually)

    All in all, it seems like a nice feature, but I worry that without precise control of how the label is placed contextually, I don’t think I would save much time beyond placing the labels manually.

    1. Claude G Lapalme

      All manners of typographical controls over the labels are part of engraving options including vertical and horizontal placement, the amount of padding etc … I have set my own preferred defaults already. They controls are very comprehensive.

      1. Alexander Blank

        Good to know! Thanks.

  3. Dennis Larson

    This is the first iteration of an incredibly complex notational process… and it is utterly remarkable. Yes, there are (and always will by) XYZ complaints about the present limitations. But I marvel at the vision of the development team (and their quite rapid development rate). I have not doubt that the limitations and overall options–when fully developed and implemented–will be stunning.

    This is v3.0. I can’t wait to see what v6.2 has to offer… it will probably just read my mind and put the notes on the page. ;-)

  4. Kenneth Gaw

    This is excellent. It is a really significant breakthrough in the world of music notation software. I used to have to write out all the parts on separate staves in Sibelius and them produce a separate file containing the condensed score. This will save a lot of time and also keep the score and parts dynamically linked.

    Between this and cues, there are can be lot of vertically linked notes in Dorico. I wondered if there are any plans for horizontally linked ones, ie passages that are copied to another location and remain linked to the source, so that any changes in the original passage are reflected in the copied version.

Leave a Comment

Your email address will not be published. Required fields are marked *