Almost all sample libraries have immanent deficiencies, most notable in the cheap ones is the lack of round-robins, and the fact that articulations are not level matched within the same instrument, and worse still, not level matched between different instruments. In order to approach even a remote degree of realism in playback, extensive level matching of instruments is a required and time consuming operation. (As both Thomas Bergersen and Arne Wallander (of Noterperformer) have stated in interviews well worth reading.)
Amen! I've found several libraries that require low level edits to 'fix' these instruments; and, some players/libraries don't have a way for the user to dig into patches/presets to make such adjustments. I.E. Sonic SE only lets us tweak what's given in the macro page, and on the 8 Quick Controls, while Halion 6 lets us dig deep into the patch and alter every single sample and parameter, either hard-set, or via real time automation to our liking. I.E. While ARIA doesn't give this control for Garritan Instruments, we can at least pull up the sfz files and make tweaks there (even inject new CC controls and such for custom automation, which could then be driven by the expression maps in real time).
The great thing about NotePerformer is that it takes this approach many steps further. It contains a full (and growing) set of level matched instruments, and it will determine and transmit technique changes based on the musical context. It may not always top dedicated sample libraries timbrewise , but in terms of pure musicality , this is nearly always compensated for by superb phrasing.
As has been stated many times on this forum and elsewhere, Dorico is not yet in a state of playback that will allow a port of Noteperformer. Personally, I hope that Dorico will prioritize development of the audio engine so that this becomes possible soon...
No argument on the quality and efficiency of Note Performer. Please don't take my Halion Hype as an attempt to be condescending about NP. There is little doubt in my mind that they can begin releasing very nice versions of NP for Dorico in a relatively short time-frame if that is a goal. I do not agree that Dorico's playback engine is 'not yet' in a state to work with NP. For dynamics and phrasing, it simply boils down to well designed instruments that get things like interpretation of note velocity, and subtle dynamic and filter changes as notes go higher or lower right, and that part of NP is already done, no matter what sequencer plays it back. To manage the library (call up presets), it'll use similar, if not the same VST protocols as Halion Sonic. Furthermore, if NP really does live internal analysis of what is coming into the MIDI buffers (up to one second latency for this purpose I understand) to make interpretive decisions (instead of relying on soundsets/filters/maps from the DAW to send the right instructions), then it would not need very complex expression maps from Dorico at all...
From what I can surmise, with the exception of fancy groove engines, the Dorico dynamic interpretation (velocity or CC dynamic control), and expression map system is already very close, if not surpassing that of Sibelius, and NP works in that. What little may be lacking to 'maximize' NP's possibilities, I suspect is merely missing in the current Dorico UI for building/altering expression maps, but not deeper down in the actual playback engine. Case in point, we can already open up a Dorico expressionmap, or an exported library of them in an xml or text editor and see quite a few tags that aren't yet available in the UI. While I wouldn't recommend messing with/relying on them in a text editor just yet (they are subject to change since they're not yet documented and in the UI), the current state of Dorico's readiness for a good NP experience should not be underestimated given the right information about Dorico from its developers.
Sorry to drone on with the techno babble, and many apologies if my excitement about how far Dorico has come in such a short time comes across as being pushy, disrespectful, or rude-condescending. I'm just excited about the potential I see in Dorico and Halion for a superb polish in future releases, so my real point for the 'here and now' is: I've just found that there are already some methods to get much better solo strings (at least in my opinion) out of Dorico using only the bits that come with Dorico (no third party plugins/libraries required). It's not the default setup, and at first there are a number of manual steps to change things around, but it doesn't have to be this way as future versions of Dorico get released (they can always select a different set of default string sounds from existing content, or even include brand new ones).
I also have a theory that the solo strings in HSO can be improved a bit as well. Right off the bat I find that removing the legato keyswitch from the expression map is better to my ear (Currently Dorico's default string map calls up a 'slow legato' sample with a key-switch on the 'legato' technique [all the slurred notes in the score], and it's not a smooth transition from the faster bowing sample that's used as the 'natural' technique). I think we might can at least make them blend better, and have a nice smooth legato interpretation with some alternative presets (Those of us with H6 can dig in and reshape every detail in the preset, and when SE 3 is released, such presets should be possible to share with fellow Dorico/Sonic SE users).
As part of this 'user community', given a little more time to test things (If things I've tried so far can load in the current version of SE...right now I'm not in a position to stop projects and run such tests), I hope to offer some alternative Halion presets and expression maps here in the forums that'll make things easier for someone who doesn't want to spend a bunch of time making their own. I might just have to wait until Sonic SE 3 is available...not sure yet.
The Steinberg Halion engine is very capable of giving Dorico Users an impressive 'out of the box' orchestral playback experience. It just doesn't seem as high on the priority list for the dev team as getting the scoring and engraving stuff where it needs to be, and getting the core audio engine elements into place (Complete and responsive Mixer, Play Tab options, etc.). It's also only fair to give the new English guys some time to really learn the foundational audio engine that may have largely been done by German teams some time ago. In time however, I'm confident every detail will get its deserved attention :)
Already many years ago, Finale's Human Playback system introduced a feature where related articulations were governed by HP and didn't have to be explicitly entered in the score (as long as they are defined in the preferences) Whether to send the Key Switch for sustain or détaché is conrolled (deducted) by the note duration. (And who would want to put "spiccato" in an instrumental part when it's completely obvious...?) See attachment.
Already, right now, Dorico can do this sort of thing to some degree (perhaps not based on note duration or tempo yet [never knew until now Finale could do this], but that should be in the pipeline). Instead of opening "Human Playback" and creating a 'filter', we go to the Play Tab, and open the Expression Map to create a 'technique'. I can let Dorico Know that I want notes with legato(slur).staccato to play back as martele, and legato.tenuto to be portato, etc. In time, I believe we'll also get access to add 'custom' sticky techniques, so it'll be easy to change the 'exclusion group' and thus our default score translations easily, and on the fly (I.E. change that legato.staccato translation into sautille instead of martele right in the middle of a score, simply by dropping in a single technique (and hiding it if we don't want it to print out).
I know that in Sibelius, it took me more than 1,000 lines of xml just to teach it translate scores for GPO5's Solo Violin like discussed above (just knowing that dots under slurs should be martele until told otherwise), and it still cannot make decisions based on the tempo or duration of notes, nor can it do simple MIDI channel hops from a soundset, and it's decades old, and beyond version 8.5. So, I've had to rely on custom sticky nodes....uggg...not much hair left. I got sick of having to tag every single note in the score for playback properties, so I just made my Own Soundset with enough booleen logic in the "SwitchType" list to get some auto-interpretation.
I find Dorico's Expression maps are already a bit easier to create and manipulate than Sibelius's soundword, and it's still only on version 1. While Finale gives me easier access to 'build as I compose', the documentation on Human Playback is poor to non-existent (Nothing in any of the manuals explains well how the plugin filtering works). In Dorico, it's already easy and intuitive enough to simply build it up and make adjustments as I compose music (instead of shifting into a programmer's frame of mind, and having to adjust xml, restart/reload, adjust, restart/reload as with Sibelius), and communicate to the Dorico Dev teams on what I can't figure out, or think should be added/changed.
It's pretty exciting to me, how good the core Dorico playback engine already seems to be. We already get more potential dev tools as end users (I.E. We can upgrade Halion to the fully featured version 6, and make our own libraries for Sonic...not so simple with Sibelius Sounds or the Plogue/ARIA Engine [true, we can make sfz instruments for ARIA all day long and load them manually, but if we want Sibelius or Finale to auto-manage them in their score managers as part of a library, we must 'register' such libraries with Plogue...fine if you intend to sell the library, but for single or lab/classroom user purposes it puts a hamper on things]).
When we get the UI tools and user documentation to take advantage of it all, and save our favorite setups as the default, and get features similar to styles and manuscript papers in Sibelius...the Dorico/Halion combo is REALLY going to soar :)