I’ve tried and tried to find the resolution to my problem in previous forum discussions, as well as the help documentation, but without success.
With like instruments condensed on one staff, what can I do to make unison passages all show up with one stem and the a2 designation rather than space-wasting double stems? I have this big concert band score that is a jumble of sometimes a2 and sometimes double stems when everything in a phrase is absolutely identical.
Welcome to the forum, Richard. In general phrases can only condense to a single voice if they are perfectly identical, not only in terms of rhythms but also articulations, dynamics, playing techniques and other ornaments, and without any pitch crossings or switches between dyads and unisons. So there must be something preventing it, but without seeing the project file or a comprehensive set of screenshots it’ll be quite impossible to tell what detail might be the cause.
Thanks for the input. I could find the cause in a few cases, but not in others. I’ll keep working on it.
You just need to put some Condensing Changes at particular places in the layout. We would need to see a screenshot or project (much better) to explain where and why.
Basically, Dorico calculates condensing based on phrases. In the case of condensing, this means music between rests. If a phrase needs to be split into two voices then it will be two voices for the whole phrase - no matter if there are unison notes. Condensing Changes tell Dorico to “recalculate condensing at this point”.
Thanks again for both repies. They were helpful and I am starting to catch on to how Dorico thinks and behaves. Like learning a new language with different rules of grammar!
This is the go-to video for all questions one might have about condensing:
Let’s see if there are still open questions after watching it 
Sometimes, it is tiresome as I end up with condensing changes on nearly every page . Sometimes every 2-3 bars to match a given paper score. This is most probably also related to the paper scores following other rules (or none?) not covered by the Dorico settings.
The key really seems to be what Dorico considers as a phrase. Although you might have a number of exactly identical bars Dorico might not condense to a2 because it considers the bars as part of a phrase with bars before the identical bars which are not identical for the targeted staffs. Putting a condensing change at the start of the unison will correct this.
A bit more AI would be welcome here.
AI makes results unpredictable and non-deterministic, you would really want that in music notation?
Oops I opened a can of worms with this last statement. I agree that AI is sometimes unpredictable. I use Newzik since it was launched and it claims to uses AI to go from a pdf to a MusicXML. The results remain completely unpredictable. Sometimes it is Okish but often not even the instrument in a simple part score is recognised and more often than not a number bars with rests, even in a part, simply disappear.
We might open a new thread from hereon on the use of AI as a tool to get faster to the engraved score we want. We can discuss if it is possible to add such AI supported, or maybe better formulated learning by example functionalities in further developments of Dorico.
I have the impression that Dorico already has a lot of what used to be called expert systems build in. The proofreading function is a good example of such a system but also condensing and anything related with the notation options could be considered an expert system. These are all rule based. So one might say it is no AI yet but it is a first step towards a learning system. It even “learns” in a way as we can tell it to ignore some of its proofreading rules but it does not learn from other user interaction yet. For notation and condensing one can only change the rules in the options and in many cases it does not work fully as desired as not all rules used are covered by those options.
An example for notation; How many times did you have to split a doted quarter note or a doted quarter rest by using forced duration (or the other way around dependent on the score at hand) in a common 4/4 piece because it did not reflect the score you are copying. This even though you thought you had the appropriate presentation settings in the options. This would at least for me be a time saver. Presently I from the start often write all notes I want to show in a tie separately and tie them with forced duration to avoid having to correct them afterwards.
It is maybe too complicated but if I change, let us say 5 times, dotted notes or rests in a piece of 500 bars with 50 or more players maybe it would be possible that Dorico recognises what I want and that it would set the appropriate internal rule to (I think it would need a rhythm mark) to get the presentation which was already changed manually 5 times in the rest of the project. That would be a learning system and therefore one could consider it AI but it will be predictable.
For condensing it is most certainly more complicated to learn by example.
IIRC This topic of AI in Dorico has already been discussed. I think you should use the search function to find that thread, read it to see what has already been said, and then continue your conversation there so we don’t repeat arguments already been made.
deleted my response
@Derrek has a point
Thanks Derrek. There are several threads but I think this is the most recent one of April 30, 2025.