Composers' Forum

Music Composers Unite!

Someone said it was OK to post another piece already, so I'm doing that.

This recording features very raw midi (I'm (as yet) on Linux with no good tools), and (referring to another discussion here) a total lack of diminished chords despite extensive modulation.

There is as yet no real orchestration. String orchestra should be sufficient, no need for any winds, brass or percussion here.

I think of this as either usable as film music, or in a televised "easy listening" performance (i.e. a flamboyant pianist in the style of Richard Clayderman, pulling his fingers through his rich hair; the camera clipping to romantic scenery with mountain tops etc., ). I have no realistic chance of getting it performed that way, but some days I dream of myself in the role of such a pianist...

I'm content with: the third theme (beginning in bar 83), and its smooth modulations. I'm not so sure about: the modulations when the orchestra takes over the melody in bars 26-34 and bars 59-68 (They sound a little more Hollywood old school there, than the rest of the piece, which pop-habituated listeners might find troubling (hope you understand what I mean)).  

I would like comments on: The style I went for here. General issues of composition, like large scale form and so on. Is it too cliché? Or, in contrast, too complex?  

I would not like comments on: the quality of the midi rendition (sound fonts, panning, mixing etc.) I know it is really bad.

UPDATE: There is now a simple full score for piano, strings and electric bass guitar. Would love comments on my "orchestration" since I'm a beginner at that.

Views: 257

Attachments:

Reply to This

Replies to This Discussion

(I'm on Linux too, but I do get by quite well. It took quite a bit of effort, but I did get a basic orchestra going with (barely) tolerable midi output. But at least it's at a level where it won't make my ears bleed and I can at least get some idea of what things would sound like played on a real orchestra. (And btw, I'm also using Lilypond. Finally, a fellow lilyponder!))

I disagree that string orchestra would be sufficient. The themes you have are very reminiscient of a Rachmaninoff concerto, which calls for full orchestra!! Well, I suppose in some sense it doesn't really matter because Rachmaninoff uses the orchestra mainly as accompaniment anyway, so in some sense it doesn't matter what instruments are present. But I still think some parts would be greatly enhanced by judicious use of a flute or oboe doubling, or perhaps some background horn harmonies.

Another reason I recommend full orchestra is ... I feel this piece has so much potential, but it's undeveloped! (Or under-developed). And in the process of scoring for full orchestra, one would, at least I hope, be naturally brought to consider the larger-scale structures and how to arrange the musical material in the most effective way.

You asked for comments on large scale form: I feel that it would do a lot of good to the beautiful themes you have here if you stuck to, or at least tried to stick to, one of the classical forms. While I found the themes truly beautiful, I also had trouble understanding the overall structure of the work.  It seemed to have sudden, unexpected jumps from one theme to another, seemingly without purpose or direction. Or at least, none that were obvious to me on first hearing -- I'm sure there is some logical structure to it, but there seems to be what I'd like to describe as a mismatch between where the music "wants" to go, and where it was "made" to go. In other words, the sequence of themes lack inevitability; the sections don't sound like that naturally lead to the next section.

Wow what I wrote sounds so abstract and unhelpful. Maybe I could illustrate by pointing to specific passages:

- m.1: The first 2 notes (F#-Ab) sound like they belong in C minor (which is also the key indicated by the initial key signature), but the way the first chord G is scored makes it sound like the tonic chord of G major. This clash makes the opening sound very awkward to my ears.  It starts making sense as mm.2-3 unfolded, that this is actually the dominant chord; but only in retrospect does the G make sense.  I'd recommend using a G7 (so that it clearly establishes that this is not the tonic chord), or something like a 6-4 chord (i.e., Cm/G) so that it's clear this is a kind of harmonic anacrusis that only resolves later.  Either that, or you could trick the listener by inserting a pure G major chord before the discordant F#-Ab, so that what is initially heard as a tonic chord slowly reveals itself to be actually the dominant chord of C minor.

- Also, I feel that the opening arpeggios in the piano part is a bit weak for an opening.  What about an all-out 32nd note arpeggio as an opening flair, to begin the piece with a bang?

- m.4: I feel the drama here could be much enhanced by perhaps strengthening the rise of the strings to Ab with a (possibly short or sfz) piano chord following by a descending unison (instead of descending 3rds like you have here).  What you have here works, but sounds a bit "flat".  By emphasizing the dramatic rise to Ab in m.4 with a sudden crashing chord, followed by a sudden reduction to a unison (or perhaps even solo) line in mm.5-8, you would heighten the sense of anticipation on the part of the audience. The descent in mm.5-8 is also slightly on the long side; if you were to follow my suggestion to use a unison/solo descent instead, you could maintain interest by keeping the 3rds in m.7, to soften the starkness of the solo line a little before the full string section returns in m.8-10.

- m.10: the open 5ths on Bb here seems to be begging for a susp4 chord resolving to Bb (i.e. add Eb 1/2 note + D 1/2 note). Could just be my biased ears, but I thought I should throw this out there in case.

- m.12: the sudden dropping of accompaniment sounds a bit too sudden. What about switching to pizzicato basses here?

- mm.11-17: brilliant progression, IMO, very much reminiscient of a Rachmaninoff piano concerto. The melody in mm.16-17 is absolutely beautiful. However, I found the sudden change in key in m.18 to be very awkward.  I'm not sure how to remedy this... Perhaps one solution could be to develop the beautiful melody in mm.16-17 to a fuller form that cycles back to repeat itself, but upon the second repetition gets interrupted by what you currently have in m.18.  The F major chord in m.18 actually doesn't sound all that bad, actually; it's just that leaping to A immediately afterwards sounds too sudden at this point in time.  Maybe you can have the F major appear in m.18 but quickly draw it back to Fm or Bb, then repeat the melody at m.15 last beat to m.17, then have the F -> A that you currently have in m.18. This would give a better flow that eventually lands back in C minor as you have in m.22. (The darkening of the triumphant F major in m.20 is also brilliant, btw, and very much in the Rachmaninoff idiom.)

- m.22: the sudden open 5ths kinda bother me... i was expecting at least an Eb somewhere in the chord. But it could just be my classically-biased ears...

The following sections marked D and E flow quite well; I feel that if you addressed the concerns on the preceding points, this first part of the piece would work very well.  I'll leave the rest for later, perhaps, there are a few other places I felt the structure could be clarified more.

Thank you for this long and detailed answer!

May I ask you about the specifics of your Linux setup?

Yes, I can see that flutes and horns doubling could be useful. I also used an electric bass in the midi file. But I will have to do a lot of reading before attempting a real orchestration, if by that you mean an actually performable one.

About large scale structure: I agree about the episodical character and "non-unevitability", although part of the idea (which may be misguided) was to have something with only the "candy" of a traditional piano concerto -- so, beautiful melodies and climaxes all the time, more or less. Also, I know my limitations -- I'm no Beethoven. I have made a few attempts in traditional sonata form and none of them made me completely happy.

That said, here is how I perceive the structure of this piece:

Section A. Intro.
Section B. Main theme (in C minor).
Section C. Continuation of main theme.
Section D. Modulation (orchestra recalls the main theme and continues the melody differently). This parts actually ends in a Eb7 chord, going into Ab major, not to Eb major where we will land for the Side Theme. But I liked that.
Section E. Orchestra introduction to the side theme. Ends in Ab minor.
Section F. Piano plays the side theme, in a half-sudden Eb major.
Section G. A lead-over to C minor, where a very abridged "repeat"-like thing happens in the orchestra, but goes into another version of the modulation thing again, which at the end puts us in A major. Since the piano plays from the start, I think more in terms of sonata form than concerto form here.
Section H. Piano plays a heated up version of the Side Theme, which can be seen to serve as both Side Theme for the repeat, and a kind of development section if you squint your eyes.
Section J. Third theme. This comes from nowhere, but I liked the effect.
Section K. Modulation back to C minor.
Section L. Recapitulation of main theme, with some added voices in the strings.
Section M. Continuation of the recapitulation.
Section N. A one-bar piano arpeggio leads us over to the side theme presented by the orchestra (initially in C major), leading up to a climax and a solo cadenza by the pianist (here I forgot to put in a letter). Instead of returning to the orchestra after the trill, piano surprisingly (hopefully) recapitulates Third Theme as a Coda. The Third Theme here gets a closed 16 bar form for the first time, and this new form of that theme brings us back to C minor for some final hint of the main theme by the orchestra, and closing with a bang.

(The above was my idea, but I'm not sure it works in practice. I have a feeling that some parts could be lengthened, for example. See below.)

For your detailed comments:  

I will definitely try a C minor chord in the piano on bar 1. Hadn't thought of that.

I wanted the real start of the piece to be section B, where tempo takes up, not the intro. But I still wanted some piano participation in the intro. So those broken 16th chords were the thing I came up with. Will think again about it. Maybe tempo of intro should be still slower, to indicate that the piece hasn't started yet.

The intro is definitely one of the weakest parts of the work, so thank you for commenting on that specifically!

Piano participation in m.4: will think about it, definitely.

m.4-8: Yeah, that needs spicing up, your suggestion about starting in unison and introducing thirds later on is a really good one.

m.10: This suggestion is gold, although I would add another bar, start on the empty fifth, and then introducing a Eb coming from the F as another voice, and then resolve it?

I thought the effect of piano playing that short motif completely solo in m.11-14 could be dramatical in a concert situation, taking his/her time, but perhaps this flies past too fast to get the desired effect. Pizzicato is a good idea if I choose to add some orchestra there.  

Maybe m.16-17 could be lengthened, since you like those. Didn't think much of it. Also the F major thing coming next could possibly be lengthened, if the modulation is too abrupt.  

m.22. Here I actually think I'll keep the empty fifth as it is.

Thanks again for taking your time!

H. S. Teoh said:

(I'm on Linux too, but I do get by quite well. It took quite a bit of effort, but I did get a basic orchestra going with (barely) tolerable midi output. But at least it's at a level where it won't make my ears bleed and I can at least get some idea of what things would sound like played on a real orchestra. (And btw, I'm also using Lilypond. Finally, a fellow lilyponder!))

I disagree that string orchestra would be sufficient. The themes you have are very reminiscient of a Rachmaninoff concerto, which calls for full orchestra!! Well, I suppose in some sense it doesn't really matter because Rachmaninoff uses the orchestra mainly as accompaniment anyway, so in some sense it doesn't matter what instruments are present. But I still think some parts would be greatly enhanced by judicious use of a flute or oboe doubling, or perhaps some background horn harmonies.

Another reason I recommend full orchestra is ... I feel this piece has so much potential, but it's undeveloped! (Or under-developed). And in the process of scoring for full orchestra, one would, at least I hope, be naturally brought to consider the larger-scale structures and how to arrange the musical material in the most effective way.

You asked for comments on large scale form: I feel that it would do a lot of good to the beautiful themes you have here if you stuck to, or at least tried to stick to, one of the classical forms. While I found the themes truly beautiful, I also had trouble understanding the overall structure of the work.  It seemed to have sudden, unexpected jumps from one theme to another, seemingly without purpose or direction. Or at least, none that were obvious to me on first hearing -- I'm sure there is some logical structure to it, but there seems to be what I'd like to describe as a mismatch between where the music "wants" to go, and where it was "made" to go. In other words, the sequence of themes lack inevitability; the sections don't sound like that naturally lead to the next section.

Wow what I wrote sounds so abstract and unhelpful. Maybe I could illustrate by pointing to specific passages:

- m.1: The first 2 notes (F#-Ab) sound like they belong in C minor (which is also the key indicated by the initial key signature), but the way the first chord G is scored makes it sound like the tonic chord of G major. This clash makes the opening sound very awkward to my ears.  It starts making sense as mm.2-3 unfolded, that this is actually the dominant chord; but only in retrospect does the G make sense.  I'd recommend using a G7 (so that it clearly establishes that this is not the tonic chord), or something like a 6-4 chord (i.e., Cm/G) so that it's clear this is a kind of harmonic anacrusis that only resolves later.  Either that, or you could trick the listener by inserting a pure G major chord before the discordant F#-Ab, so that what is initially heard as a tonic chord slowly reveals itself to be actually the dominant chord of C minor.

- Also, I feel that the opening arpeggios in the piano part is a bit weak for an opening.  What about an all-out 32nd note arpeggio as an opening flair, to begin the piece with a bang?

- m.4: I feel the drama here could be much enhanced by perhaps strengthening the rise of the strings to Ab with a (possibly short or sfz) piano chord following by a descending unison (instead of descending 3rds like you have here).  What you have here works, but sounds a bit "flat".  By emphasizing the dramatic rise to Ab in m.4 with a sudden crashing chord, followed by a sudden reduction to a unison (or perhaps even solo) line in mm.5-8, you would heighten the sense of anticipation on the part of the audience. The descent in mm.5-8 is also slightly on the long side; if you were to follow my suggestion to use a unison/solo descent instead, you could maintain interest by keeping the 3rds in m.7, to soften the starkness of the solo line a little before the full string section returns in m.8-10.

- m.10: the open 5ths on Bb here seems to be begging for a susp4 chord resolving to Bb (i.e. add Eb 1/2 note + D 1/2 note). Could just be my biased ears, but I thought I should throw this out there in case.

- m.12: the sudden dropping of accompaniment sounds a bit too sudden. What about switching to pizzicato basses here?

- mm.11-17: brilliant progression, IMO, very much reminiscient of a Rachmaninoff piano concerto. The melody in mm.16-17 is absolutely beautiful. However, I found the sudden change in key in m.18 to be very awkward.  I'm not sure how to remedy this... Perhaps one solution could be to develop the beautiful melody in mm.16-17 to a fuller form that cycles back to repeat itself, but upon the second repetition gets interrupted by what you currently have in m.18.  The F major chord in m.18 actually doesn't sound all that bad, actually; it's just that leaping to A immediately afterwards sounds too sudden at this point in time.  Maybe you can have the F major appear in m.18 but quickly draw it back to Fm or Bb, then repeat the melody at m.15 last beat to m.17, then have the F -> A that you currently have in m.18. This would give a better flow that eventually lands back in C minor as you have in m.22. (The darkening of the triumphant F major in m.20 is also brilliant, btw, and very much in the Rachmaninoff idiom.)

- m.22: the sudden open 5ths kinda bother me... i was expecting at least an Eb somewhere in the chord. But it could just be my classically-biased ears...

The following sections marked D and E flow quite well; I feel that if you addressed the concerns on the preceding points, this first part of the piece would work very well.  I'll leave the rest for later, perhaps, there are a few other places I felt the structure could be clarified more.

My Linux setup is somewhat complicated, especially since Lilypond, being a notation program, isn't exactly catered to production, especially for pieces for large orchestras.  But since you asked:

Firstly, the most effective use of Lilypond I've found is to always separate the score for layout (IOW the conductor's score, the one that you'll actually print) from the "score" intended specifically for Lilypond's rather rudimentary midi capabilities.  Also, if you ever want your work to be performed by a real orchestra, you'll also need to print instrument parts. So I settled on keeping the actual music in a separate file (or you could use multiple files -- but I found keeping everything in one place easier for keeping everything in sync), basically as separate variables for each instrument part (e.g., fluteIPart, fluteIIPart, clarinetIPart, clarinetIIPart, etc., the string section is just violinIPart, violinIIPart, violaPart, celloPart, bassPart).  Then in a separate input file, specifically intended to produce the conductor's score, I write the \score{} block and create various StaffGroups, Staffs, etc., which \include the actual music content and places them in the right place. Similarly, I have a separate file for producing the "midi score" that is never printed, but only used for driving midi production.  And a third file specifically geared for generating instrument parts.

The nice thing about separating the music from the clerical score-production stuff is that whenever I start a new project, I can just copy the files for generating the conductor's score, midi score, instrument parts without even needing to edit them, and just write new music in the file that defines the music variables.

For the midi production, I use TiMidity++, which is arguably ancient outdated technology, but it lets me streamline everything I do with Lilypond because both are batch-processing based tools. So I can use a Makefile or script or some such to run lilypond, then timidity, then apply any audio processing necessary, etc..

I found that there is no single soundfont that will do everything I want, so I mix-and-match them. So I write a configuration file for TiMidity++ that basically pulls in different instruments from different soundfonts and assigns them to my own numbering system (I used to use the General MIDI mappings but eventually found it too limited for my purposes).  Then in the midi score file, I assign the instrument names Lilypond insists on (a silly restriction IMO - one of these days I should heckle the Lilypond devs to allow numerical assignments to Staff.midiInstrument) corresponding to the actual instrument numbers I mapped.

This setup worked for a while, but then I ran into the MIDI 15 channel limit.  But I wanted orchestras with more than 15 different instruments playing at the same time... so eventually I redid the midi score file to generate multiple MIDI files - one for each section of the orchestra (winds, brass, percussion, strings). Then I use SOX to join them together at the end. This way, I can have up to 15 channels per section, which greatly expanded the possibilities. Of course, invoking sox involves long-winded non-trivial arguments that I wouldn't impose on anyone, so obviously that goes into the script for building everything.  (There is also the problem of sync'ing the sections: for some silly reason, either Lilypond of TiMidity++ will truncate initial rests, so if your  brass section is silent for the first 16 measures, the brass midi file will actually start at m.17, so if you join it with the other sections, the brass will enter too early! The hack I use to work around this is to insert 1 bar of nonsense notes at a fixed tempo so that it lasts exactly 1 second -- that way, all sections will start at the right time -- and then use SOX to trim that off the first second of the final audio. It's hackish, but it works. And of course, this also goes in the build script so that I don't ever have to worry about it again afterwards - it's all automated.)

Then I discovered that I wanted different articulations too.  Some of my soundfonts have different "instruments" that provide different articulations, e.g., arco vs. pizzicato strings, muted vs. unmuted trumpet, etc..  This can be done by mapping them to different instrument numbers (another reason General MIDI is woefully inadequate for my purposes), and defining a Lilypond macros file that maps articulations to instrument numbers automatically.  For example, \vnArco will set the instrument number for arco violins, \vnPizz will set the instrument number for pizzicato violins, etc.. Having these as macros means I don't ever have to remember what numbers everything is mapped to. I can also change the mappings without needing to edit anything in the actual music.

There are also times when I want to be able to spell out certain things in the midi score that I want to just indicate with a single mark in the conductor's score, e.g., trills and drumrolls.  There are also some hacks that are sometimes necessary to make Lilypond produce a palatable midi output, but you don't want it to be seen in the printed score.  For this purpose, I use Lilypond's tag feature: stuff in the score that should only go into midi is tagged with \tag #'midi; stuff that should only appear in the printed score is tagged with \tag #'layout. Then in the file for producing the conductor's score, I use \removeWithTag #'midi to delete all midi-specific stuff, and in the file for generating midi I use \removeWithTag #'layout to delete all layout-only stuff. This lets me write different notes for the score vs. on the midi, so that I can, for example, write \tag #'layout { c4\trill } \tag #'midi { c32 d c d  c d c b } to precisely spell out a trill. (I've tried using articulate.ly's trill-expanding feature, but it rarely does the right thing in non-Baroque pieces, so these days I prefer to just spell it out in \tag #'midi.) This also allows, for example, writing abbreviated glissandos in the score but actually spelling it out in the midi, 'cos I don't think Lilypond's midi output can actually produce the right notes otherwise.

All of this is a lot of tedious setup, of course, so I eventually wrote a little helper program that generates all of the foregoing automatically, given a simple input file that specifies the orchestral layout I want for a particular piece.  If you're interested I can provide more details, but I don't want to clutter this forum with tedious technical details unless somebody actually finds it helpful.

So in a nutshell, my setup is Lilypond (separate input files for the actual music, conductor's score, midi "score") -> TiMidity++ (multiple midi files, esp. for a large orchestra) -> SOX (combined audio file) -> LAME (mp3 encoder).

Of course, I left out a lot of details, such as which soundfonts to choose, how to balance them with each other, etc.. But if there's a specific area you're interested in, I'll be happy to elaborate.

As for the music part of your comment: I can understand if you're not interested in writing a full-fledged sonata form... or, for that matter, working on an actual, performable orchestration -- it's a lot of tedious work! Well, I'd still recommend the latter, if only for the educational value of knowing how to write instrument parts effectively, even if in the final product you actually work with a generic "string section" patch.  Thinking in terms of 5 independent voices in the string section, I've found, actually greatly improves the quality of my string writing, as opposed to when I just wrote for a generic string section with an arbitrary number of voices.

But as for form -- if you want "all the candy" but don't want to deal with the strictures of a sonata form (or, for that matter, a 3- or 4-movement piano concerto), I'd say it's still doable, but you should still at least stick to a consistent plan.  For such a form, what comes to mind would be a series of sections of more-or-less equal length, with clearly defined boundaries.  Kind of like an opera or ballet condensed into a series of mini movements encapsulating the main themes in the full version, like Tchaikovsky's Nutcracker Suite, which is condensed from the full ballet. The music itself can still have smooth transitions between sections, of course, but it should be clear, upon listening, that you're entering a new section. And again, writing for a full orchestra here would give you a convenient tool for this purpose: change the orchestration between sections so that the transition is obvious.  What you have right now doesn't have a very clear sense of distinct, more-or-less self-contained sections of each idea, so one is led to expect a full-fledged form (classical or otherwise).

I haven't analysed the rest of the piece in detail beyond what I did earlier, so I can't be more specific at this time, but maybe later I might find some time to try to follow along the plan you presented to identify possible areas of improvement that might help you realize your vision more effectively.

H.S.:

Wow! That's something complicated for an audio setup. I hesitate to follow your route... but it would be cool to be able to combine different sound fonts. I might go the easy route at some point and get a dedicated non-Linux computer for doing my audio stuff. But right now I'm on a budget.

I'm looking forward to when or if you get time to analyse the rest of the piece.

I would really like to do a real orchestration, but it will take me some time since I'm a beginner at orchestration at this point.

If you want to learn orchestration, I highly recommend getting Adler's textbook The Study of Orchestration. Of all the orchestration books I have (Piston, Rimsky-Korsakov, and one other IIRC), I found Adler's most helpful in terms of getting me started, and remains the most useful as a go-to reference.

Of course, score study is an invaluable part of it too (i.e., study orchestral scores of familiar works to learn how the composer achieved a particular effect), but you do need a textbook as a guide to get you started first, otherwise you may find yourself staring at a score and not knowing where to even begin, or what to look for. :-D

I made some changes which should hopefully bring out the structure better. Made the piano quiet during the introduction, and other things. The sound quality of the new file may or may not be better.

Attachments:

Listened to the new version.

- mm.12-14 still bother me ... I still thought of having background pizzicato to accompany the piano. But it occurred to me that perhaps if the string B natural had a diminuendo, it would allow the piano to make a dramatic solo entrance. Perhaps that's the solution closest to what you envisioned.

- After hearing mm.18-19 several times by now, I see that they are intended to be a quick diversion... but the sudden key change still sounds awkward to me. Could just be my personal bias... but perhaps there's some way to pull this off without the awkwardness. Have to think on that a bit.

- mm. 35-38: I feel like the piano should be playing some background accompaniment figures here, topped off with an upwards arpeggio in m.38 before the melody reenters on the piano.

- mm.60-68: I feel like the piano arpeggios should become increasingly elaborate towards the end, to intensify the tenseness of the music before cadencing in m.68. Perhaps break them out into 16th note sextuplets in the second half of the passage, or something.

- mm.69-83 (H): I see you're trying to write a modulating passage here... but currently it sounds too repetitious, just repeating the theme in different keys. The music seems to be asking for something more here. You can probably still get away with repeating the theme in different keys, but you should orchestrate each occurrence differently. There are two reasons for this: (1) to differentiate this modulating section from section G, so that a casual listener (who may not consciously notice the key changes) won't mistake it as merely a repetition of what came before; and (2) to add more interest so that when you reach J, the new theme sounds truly novel.  There are many ways of doing this, of course. Simple devices include playing the melody in a different register, accompanying it with different figures, use different instrumentation (even if you have just a string orchestra, you can still do a wide variety of textures: string tremolos, pizzicato, string chords in a different register, counterpoint, etc.). The key is to clearly differentiate the A major section from the previous material, and the D major section from the A major section.

- I found the new theme in section J quite fresh and inspired, actually. I agree that it's quite inspired, and I like it. It adds a new freshness to the themes that have already been heard before that. Something seems odd with the mp3, though.. I can't seem to hear the piano's LH figures?

- Similarly, section K seems to be missing the 16th notes in the audio?

- mm.125-139: this section feels like it should be climactic, but the piano part seems so weak!!! If I were to write this, I'd put crashing chords here, maybe even chordal arpeggios (i.e. fill in entire chords below each of the ascending 8th notes).  Couple that with bass octaves in the LH for added strength.  Then in m.134 I'd break loose and have the piano go wild with virtuosic stunts to go along with the intensified upward leap in the strings melody, ending with a giant arpeggio on the D7 in m.139 that spans the entire keyboard, as a dramatic opening for the cadenza.

- The cadenza seems to want to be written more virtuosically. Here is where the pianist wants to show off his prowess, so throw everything you've got at him.  I'd write this with lots of heavy chords for added emphasis, and the LH arpeggios in m.147 onwards I'd write at least as 16th sextuplets, if not more elaborate.  In mm.155-158 I'd expand those LH arpeggios to cover at least 2-3 more octaves in width, and in 16th notes (or even 32nd notes), perhaps with the RH picking up the upward rushing tails for the extra flair.  At the very least, double your LH part an octave below, otherwise the piano part just sounds wimpy.

- m.159 the A natural in the bass in beat 1 bothers me a lot.  It makes me cringe and want it to sound a Bb instead, then A on beat 3. And I'd transpose the LH in mm.161-162 down an octave for a stronger conclusion to the cadenca.

- mm.162-165: the ending seems to come a bit too early. After the drama of the cadenza, the audience are raving to hear more than just another 4 bars!!!! If I were you, I'd seriously considering writing a coda here to wrap everything up and bring the piece to a strong, firm conclusion. The short ending you wrote here, I feel, doesn't do justice to what came before.

H.S:

Yes, sorry! -- the audio and score went out of sync. The audio gives what is my current version (with only a simple melody in the piano in J, without left hand, and then starting the arpeggios in K in a gradual way with eights first). The updated score can be found in the opening post. A problem with this change, though, is that on the wrong instrument and with a bad conductor or the pianist having a bad day, the melody may be drowned by the orchestra in J. But I wanted simplicity. Overall in my audio, the piano is perhaps unrealistically loud in many passages, more like an enhanced studio recording.

Yes, there should be a diminuendo on the B in bar 11 -- I simply don't know how to create that in Rosegarden, which generates my current audio. I have experimented today though with letting lilypond generate midi from a separate score geared for midi generation, as you suggested. My current problem right now is that the volume goes mute randomly, and it seems to ignore my \fffff marks in the piano part. And I'm not sure the audio is better. But I'm really tired of Rosegarden, so it would be great to make it work.

I wanted to give the pianist a break in m.35-38, and also create different texture. It's after all an introductory passage to the side theme. Maybe the arpeggiated chord in 34 should not be arpeggiated, but just accented, so that the pianist goes out with a bang before his/her 4 measure repose?

I see what you're saying about texture in H. Will try to come up with something. There could be some accelerando here too.  

125-139: Problem with widening the piano chords in 125-139 is that it clashes with the melody and bass line of the orchestra. Live, this would of course be a full tutti passage, with the pianist controling the pace of the accelerando, which could well be more exaggerated. What I have right now in the piano will at least _sound_ OK (if not drowned, though), and also look funny, as the hands jump up and down in order to create the desired articulation.

Having bigger arpeggios in first 4 measures of the cadenza was my first thought, but they sounded tasteless with the jazzy chords. But the cadenza part before the sixteenth trill is possibly far too short, and if I extend that part there may be more room for virtuosity. The trill is in sixteenth notes currently, but it should probably be a real trill.

The following measures of the cadenza are meant to show off cantabile playing, which is also a kind of virtuosity. But I think you're right that 155-158 should increase the tension with bigger arpeggios.

159: I see no A natural here, only an E natural (part of a C7 chord)? I think the bass line wants to go up, not down, to G in 161. I probably want 159-160 to go down in dynamics again from the previous measures. Then I'll write f or ff in 161. As a pianist, I think that will emphasize things enough, without lowering the bass line an octave.   

162-165: I see the last 16 measures of the cadenza as a kind of coda, though (they come after the trill). But something is lacking here as regards normal expectations of a piano concert, that is true.

Thank you for going all way through with your analysis! My comments may be too small-minded -- these are just the first things that popped up in my head as an answer, but there may well be clever solutions that will realize your intentions while not subject to those objections that I may have raised.

The files in the opening post have been updated with today's latest version. Significant changes: Gentle piano participation in measures 35-38. Slightly more virtuosity/emphasis in piano from measures 125 onwards. More virtuosity in the solo cadenza. Cadenza actually sort of ends after the trill, so the final section "O" now has gentle orchestra accompaniment.

The sound file in the opening post has been updated. The low and slow strings are now using another soundfont. It's not production ready audio in any way, but should be somewhat easier on your ears.

Here is a very unprofessionally made "live recording" (acoustic piano + midi strings). It's complete with coughs and a mess up by the pianist (me) during the climax. But should still give some better idea of the desired sound than the midi piano version (which you can still find in the opening post).

Also, there is now a full score for piano, strings and electric bass guitar. Would love comments on how I could improve on that.

Attachments:

Reply to Discussion

RSS

Sign up info

Read before you sign up to find out what the requirements are!

Store

© 2021   Created by Gav Brown.   Powered by

Badges  |  Report an Issue  |  Terms of Service