Hello all, first post here I think. I am pretty much a summer-only (northern hemisphere summer, that is) composer as I teach full-time during the academic year, a grueling schedule. I just thought I would get my completed works up on the new board here: all two of them! (three counting the strings-only version of my Op. 2, a symphony whose ending has given me no end of headaches over the past 4 years. But that will be for another post.
I began this Quartet as a student in July 1975 at University of Michigan in Ann Arbor under William Albright. As the end of the summer approached, I couldn't figure out how to end it, so I gave it a throwaway ending and shelved it. After graduating in 1977 I returned to my original major, physics, and stopped composing altogether after a year or two. I knew nothing of notation software until late 2019 when I discovered MuseScore and used it to resurrect my old score and try to fix some errors and work out the ending. I quickly realised that it didn't need an ending, at least not quite yet. Instead I added an intense development section, an ultimately tender recapitulation, and a searching coda, mainly for solo viola. I made a couple of changes after abandoning MuseScore in favor of Sibelius with NotePerformer, but I consider it finished at this point.
All that to say that if you heard this piece in the last couple of years on the old board, there is no need to re-listen, I don't believe I made any changes since then. The score and rendering are from mid-July of last year. They're on Google Drive, which in my experience is an awfully unreliable playback device. I recommend downloading the MP3 and playing it back on the player of your choice. Playing time is about 16 minutes.
Replies
a deepy serious and compelling work -- if anything even more impressive than my initial memory of it. The only problem I have with it is -- as you may have guessed - the mechanical nature of the NotePerformer solo strings which is really the worst aspect of that library. You may not share that view but if you do, I could try it out with other solo strings libraries if you were to post a MusicXML export.
Hi David, thank you for your feedback, and I'm glad you find it compelling! No, I'm not fond of NP's solo strings, though I'm not sure why you call them "mechanical" - the problem I have with them is the timbre. Basically, they sound MIDI-ish and synthetic (which, after all, they are). NP's renderings are to my ears more musical and expressive - less *mechanical*, in other words - than anything I've heard with libraries supported directly by notation software.
But now, with NP4 supporting certain libraries, it might be worth trying that option. Here's a MusicXML export:
https://drive.google.com/file/d/1OQKR5AVZmoNJMP5zxKDrP8sbSDpVPopf/v...
I suspect "mechanical" was the wrong choice of word here as it is indeed the timbre which is the problem. Given decent samples to begin with (which above all means the optional NP Performance Engines), NotePerformer can produce outstanding results.
Anyway, the obvious starting point is to use the Cinematic Studio NPPE. I have simply exported the results from Dorico with no changes whatsoever. If you don't like this library (I tend to use it for "sweeter" music than this but it seems not bad), I can spend a little more time using a different one. If you hear anything which seems obviously wrong, it could be an issue with reading the MusicXML file -- in which case first check the attached with the Sibelius original -- but at first glance, it looks quite good.
https://app.box.com/s/wc7u00eu15iy9lzw9k8xvby0a4jzfuik
01 - Full score - String Quartet No. 1.pdf
Thank you for doing this, David! As you may have guessed, to my ears Cinematic Studio really isn't the right library for this piece. It tends to attack nearly all notes very gently, gradually - and especially in the faster parts of the development, the attacks shouldn't be gentle - much of that music should sound pretty ferocious. But I'm concerned because in places the four instruments seemed not very well synchronized, almost like the old problem with NP3 and solo strings where the entrances were timed with a random "human error" factor. But not quite - this sounds more as if each line was being played rubato independently of the other instruments, sounding like poor ensemble. In places, even sostenuto notes followed by a staccato note were played as if the first note was tied to the staccato note. This was especially noticeable after letter D. There is also a tendency for sostenuto notes to be allowed to sound beyond their written time, so that plus the gentle attacks might explain the missing staccato notes.
I'm guessing at least the ensemble issue is NPPE still having some bugs. If you're game, I'd be interested to see how a different solo strings library does with my piece - from your earlier description it sounds as if VSL might be a better match (as I recall, VSL is not supported by NPPE). But first, I have to correct some errors in my notation of certain tremolo passages - for some reason, I transitioned in the middle of the development to what sounded to me like a more rapid tremolo, notated with a 'Z' on the stem. Stupid me, that was a buzz roll, and obviously it didn't export anyway. That's a lot to fix, so it could be a few days.
(Also, I think some metronome markings didn't export or weren't read correctly, as some passages were played at the wrong tempo, most noticeably after letter M, where in the original there is a Poco piu mosso, crotchet = 72. In your score, the metronome marking occurs 3 bars later. Whether the error was in the export or the read into Dorico, it's probably not worth your while to fix as you'd have to do it manually.)
in case someone else wants to listen to the Cinematic Studio version, I've redone it with the tempo corrections made for the VSL one. The original link no longer works (as this forum doesn't allow you to edit posts after a short time, I can't do anything about the original)
https://app.box.com/s/zd7c2b3lk62jmu4j6fsdbg04nx6cjdt1
Sorry David, but my impression of Cinematic Studio hasn't changed: it's just wrong for this piece. Not enough clarity of articulation, the instruments all seem to be playing independently of each other with maximum rubato (in Sibelius, I have all rubato and espressivo turned off). The duet in the recap comes off the best, not surprisingly: it's tender and lyrical. I was a little surprised by the last few bars: in your original CSSS rendering, they were the best part of the piece: everything gradually sinking into silence, and holding the tension all the way down. The VSL rendering was very similar in this passage. But in this rendering, there's a sudden increase of dynamic level just as the other three instruments enter to cadence into A minor over the viola. Any idea about what happened there? Maybe no definition in the Dictionary (or whatever Dorico's equivalent is called) for pppp and ppppp? (But then why did the earlier renderings have the correct dynamics?)
The VSL version required some manipulation of CC11 at the end as the dynamics were not going quiet enough (I don't use anything more than ppp myself as a rule) but this unfortunately had a completely different effect under NP. I've now eliminated all that so with luck you should find it as before. I'm not going to try and convince you of the merits of the library here as you've made your view clear. I rather like it on the whole but really it's just in case others listening to your fine quartet would like to try out different versions and decide for themselves.
Cinematic Studio is not ideal when you need tight timing because, as you say, a combination of slightly woolly attack and a fairly reverberant sound means it's not always suitable although I like it a good deal for lyrical romantic music. As regards tempi, the problem is you've used in a number of situations a combination of a description and a fixed tempo. Dorico has probably a quite different descriptive dictionary than Sibelius but anyway, it clear that Dorico in the first instance tries to interpret the description when present. I've asked in the forum if the behaviours could be changed to ignoring the description when a specific tempo is also given but I don't know whether Dorico is simply following MusicXML rules or has a mind of its own.Certainly the safest is just to use fixed tempi for export.
Anyway, I have corrected all the errors I could see and created a new file with VSL which may be more to your taste. The short con sord section is ignored as I don't have the sord. solo strings library (and trying to integrate the Dimension strings would be more trouble than it's worth here). But with luck, most of the rest should be there or thereabouts.
https://app.box.com/s/82uyk79agr28fa7ryfng5rrpdnywunck
Hi David,
Thanks again for the two demos. Of the two, yes, VSL is very clearly the more suitable for this piece. Whether it is superior, artistically, to straight NP is another question. Many of the attacks are still too soft even with VSL. There is also the question of how much of my own notated attacks, phrasing, and dynamics are specifically to work around quirks in NP itself, that don't work with VSL / Dorico (and might even have to be adjusted for human players). I did notice that the change to faster tempo at letter M is in the correct place now - I assume you made the change manually in Dorico?
I'm not sure what you mean that my descriptive or text-based tempo markings could be interfering with Dorico's setting the correct tempo. Are the text-based markings even present in the MusicXML file? I'm not sure, but they're not present in the score you posted - all I see are metronome markings. I've assumed that the text-based markings were not exported. If they were, then yes, the semantics of mixed tempo markings could be different in Sibelius vs. Dorico. In Sibelius, the metronome marking always takes precedence if present, overriding the default value set for the text version.
Given the cost of Dorico + VSL Solo Strings, I'm still inclined to stick with my current system and just live with the poor synthesized solo strings in NP. (I think it's interesting that even though NP's string sections are made up of those awful solo strings, the sections actually do not sound too bad - even when reduced to 3 or 4 individual instruments.) Anyway, when push comes to shove, I can tolerate weird timbres. The musicality is always more important to me.
Liz
if you want a particularly strong attack, then you can always notate marcato for instance or whatever is appropriate -- I don't remember seeing many such markings. Dorico automatically tries to adjust dynamics according to the melodic flow (it's called pitch contour emphasis) but won't add extra attack unless specifically notated.
I changed all the erroneous tempi in the score. MusicXML does in fact export text based tempi markings and Dorico interprets these according to its own dictionary (which you correctly assume does not correspond to Sibelius) but omits them from the score . Daniel Spreadbury has already responded that he's not sure why these overrides replace metronome marks where specified and it's something he'll look into when he has the time. Dorico has become fairly good at MusicXML import but it's still not an exact science and of course also depends on correct export in the first place.
-
1
-
2
-
3
-
4
of 4 Next