The Wind and the Lake (classical / symphonic poem)

[Rainmak3r]

Hi all,

I'm very excited to finally share this... thanks to the holidays, I've finally been able to get to a symphonic poem I started working on many months ago. No rock, no contaminations: this is 100% classical music, and more precisely an attempt to write a late-romantic symphonic poem, a kind of work that was quite popular in the second half of the 19th century. I'm particularly excited because this is one of the two pieces that I REALLY wanted to finally write down (the other being the symphony I promised I'd finish by the end of the year, and that I'm only at about 35% now!), when I first registered here more than a year ago

You can listen to the work here:
https://soundcloud.com/lminiero/the-wind-and-the-lake

I uploaded the whole score (that can be played and followed using the MuseScore soundfont) as well:
https://musescore.com/lminiero/the-wind-and-the-lake

This is probably the thing I've shared I'm the proudest of, so far. I've had many of the themes in my head for more than 15 years: at the time I remember writing some of them down in Sibelius (when I was still using Windows, so ages ago), but those files were lost forever unfortunately. At the beginning of the year, having worked on multiple tracks with orchestral elements already, I decided it was time to try and really start working on it, and many new ideas started flowing! I don't think it's perfect: there are a couple of parts that still feel a bit "disjoined" to me, for instance, and that I could have tied better. But I really like how it came out nevertheless, and considering the complexity of the score, I have to say I didn't think I'd be able to pull it off!

As anticipated, the aim was to write a late-romantic symphonic poem: no modern classical influence, as I usually am not a fan at all of classical music from the 20th century, but only late-romantic music, which I've been in love with for decades. It is quite blatantly derivative of Tchaikovsky's work, and that's to be expected since he's my favourite composer of all times. If you're familiar with his works, for instance, it will be hard to miss how similar the beginning is to "The Tempest" starting notes, or how I used the same "tragic transmutation" of the main theme at the end as the Maestro did in "Hamlet", but you'll find his influence all over (like in the woodwind section that precedes the waltzes, or the waltzes themselves). Some very strong influences came from Rimsky-Korsakov as well, who is probably a close second in my "favourite composers" list, especially considering his astonishing orchestration and arrangement skills: the "Antar" Symphony was a strong influence on the first part, but the brief harp interlude may remind you of a more popular work of his, "Scheherazade". Hopefully I didn't unconsciously plagiarize either, and I managed to remain on "derivative"

Those amazing composers usually wrote poems on existing works (e.g., Shakespeare was a frequent inspiration for Tchaikovsky). In my case, the theme of the symphonic poem is an "original" image I've had in my mind for a long time: a deep, dark and sleeping lake is awaken by a gentle breeze, which initially brings life, but then becomes a strong wind and agitates the lake causing a storm; until the wind leaves and the lake, alone once more, goes to sleep again. If you listen to the work, it should hopefully be clear which part is being described (if not, I did something wrong!). This also gave me the opportunity to play a lot with themes, even more than I usually do. You'll find some themes that come back more than once, often "transformed" one way or another: the parallel minor chords that introduce the opera, for instance, take different forms, until they become the frame for the sad ending, where the same notes are played but with a different harmonic background; one of the happy waltzes becomes dark and scary to describe the moment you realize a storm is coming; and the theme of the lake too, which initially is just unsettling, becomes loud once the storm arrives. I tried to use the orchestra to mimic the "sounds" of nature as well, especially during the "storm": I didn't get exactly what I was looking for, as for instance I wanted to use the panning of the instruments to my advantage, e.g., to simulate the wind (in particular when woodwinds play left-to-right), but that didn't really work out as I expected... that said, I think there are still a couple of nice moments there.

Technically, it's been quite a challenge, and I'm not just talking about the pure composition and orchestration aspect. This is going to be long, so feel free to skip if you don't really care! In a nutshell, the process was basically quite "traditional" and apparently straightforward:

• the whole work was composed and written down using MuseScore;
• at the end, the MIDI parts were imported in Ardour, and rendered with Virtual Playing Orchestra's better sounds.

That said, it was quite complex, at least for me. First of all, I used MuseScore exclusively for the composition, which means I initially didn't use it thinking of the virtual orchestration I would have had to deal with sooner or later. This was done on purpose, as I wanted to approach the work as a regular "composer", and I wanted the score to stand on its own feet: I don't delude myself thinking this will ever be performed by a real orchestra (or even that the way I wrote it is indeed performable), but I still wanted to write a score that could be read without being tied to virtual orchestration details. This obviously means that, as soon as I considered the work on the score done (days ago), it clearly was a mess when VPO came into the picture, which I expected and had prepared for.

I had already decided to use the excellent VPO template for Ardour by @Michael Willis as a basis: I had already played with it a bit back when I worked on my Daydreaming Waltz, and loved how lightweight it was despite the amount of instruments: besides, it already pans all the instruments depending on where they sit in the orchestra, and the template comes with three different Dragonfly Reverb instances depending on how far instruments are, which is amazing! That said, Michael uses the PERF versions of the SFZ files there, which wouldn't work for me and, since I needed more dynamic SFZ files (e.g., to be able to go from modwheel to staccato on the same track), neither did the existing VPO SFZ files (while there are keyswitch versions, they use notes out of the instrument range, so would require messy score changes). For this reason, a few months ago I started working on some alternative SFZ files, that would use a specific MIDI CC controller to trigger sample changes instead: it took a few revisions, and some ugly workarounds, but I eventually made it and you can find it here if you're interested.

Now that I had dynamic SFZ files, I created a copy of the MuseScore project and started editing that. Specifically, I added support for my custom CC events, and added them wherever needed: e.g., if I needed a staccato, I'd remove the "traditional" way of rendering it in MuseScore, and would add my own CC instead, so that the SFZ player would know it would have to switch to the staccato samples of the instrument. This took quite a bit of work, since all the events must be put on the note/rest before (my repo explains why), and besides there could be A LOT of articulation changes in a segment... for a work that is more than 10 minutes long and has two dozens instruments it took a while Eventually, though, I got something that seemed to work (MuseScore supports SFZ files through Zerberus), even though of course still not the end result since volumes were completely unbalanced due to the differences between the MuseScore soundfont I had worked with so far and Zerberus' rendering of the VPO files (which seems to be buggy too, with samples that stop playing way too soon for instance). Anyway, I created MIDI files out of this modified version of the score, so that I could use them in Ardour.

Again, I used Michael's template as a basis, but I made a few changes:

• First of all, I duplicated most of the tracks: in fact, Michael's template has a single track for each instrument, while I needed, e.g., 2 for flutes, oboes, and so on.
• At the same time, I also got rid of the duplicate strings tracks: Michael's template has separate tracks for bowed/pizzicato/tremolo strings, which makes sense in most cases but didn't for me. In fact, with MuseScore I'd always write a single track per section anyway, and as anticipated I had created my own SFZ files to change samples depending on the articulation, so a single track would suffice.
• Finally, I replaced all the PERF SFZ files the template defaulted to with my own CC-switched SFZ files, which is was what required the most work. Not the SFZ replacement per-se (this is made very easy by sfizz), but the volumes adjustements. In fact, as Michael already told me in a different thread, all the volumes for each instrument had been carefully tuned for those specific SFZ files, and changing those meant the instruments were now completely unbalanced. So I basically had to slowly try and balance them again according to my own SFZ files, and make sure they'd remain balanced when switching articulation too. Not sure if I did it right, but I think it sounds nice enough now, and in theory I should be able to use this project to create my own template for upcoming classical/orchestral works.

Once all that ws taken care of, the last part eventually was mostly just fine tuning: e.g., some swell effects where needed (especially for strings, but not only), or mod-wheel automation to give different attacks to the instruments depending on what they were playing. This is also where I noticed I failed to script all the articulation changes in MuseScore, which meant manually applying them in Ardour, which was simpler than I expected though.

That's all! As usual, I ended up writing way too much, but I'm really looking forward to hear what you think of it: I'm especially curious about whether you'll like the themes or not...

[j_e_f_f_g]

Surprisingly nice. I say "surprisingly" because this is the first time someone here has posted an original orchestral piece that sounds like it could actually be a real orchestra playing. Previous examples either lacked enough "human articulation" to sound convincingly real (it sounded like a bunch of orchestral samples being rudimentarily sequenced). Or, the arrangement itself just didn't sound like something an experienced orchestrator would write. You've studied your influences well. Even if you hadn't mentioned Tchaikovsky as an influence, my immediate impression was "this guy likes Tchaikovsky, Mussorgsky, Rimsky-Korsakov, and Grieg". The early-to-mid 1800 slavic influences are very prominent. (To be honest, our tastes in classical repertoire are polar opposites. The slavic guys aren't my cup of tea. Tchaikovsky is one of my least favs. I don't get interested in that region until Stravinsky and Prokofiev show up. OTOH, Aaron Copland is one of my favs.)

This is a well-constructed orchestral piece. Even without your description of the work involved, anyone who has orchestrated, particularly with sampled orchestras, can tell that this was a lot of work, just from listening to the result.

There's just one thing that didn't work for me. That solo trumpet toward the end. It's not the part itself. It's the sound. It's a harsh, static sound that doesn't sit well in a romantic period piece. It could perhaps be softened or given more "passion" with more vibrato, and some volume swells/dips. But I think that you should investigate a different solo trumpet sfz. (The one from No Budget Orch is a more resonant one. That's what I think would work -- a more resonant, fuller trumpet with vibrato).

[Rainmak3r]

Surprisingly nice. I say "surprisingly" because this is the first time someone here has posted an original orchestral piece that sounds like it could actually be a real orchestra playing. Previous examples either lacked enough "human articulation" to sound convincingly real (it sounded like a bunch of orchestral samples being rudimentarily sequenced). Or, the arrangement itself just didn't sound like something an experienced orchestrator would write. You've studied your influences well. Even if you hadn't mentioned Tchaikovsky as an influence, my immediate impression was "this guy likes Tchaikovsky, Mussorgsky, Rimsky-Korsakov, and Grieg". The early-to-mid 1800 slavic influences are very prominent. (To be honest, our tastes in classical repertoire are polar opposites. The slavic guys aren't my cup of tea. Tchaikovsky is one of my least favs. I don't get interested in that region until Stravinsky and Prokofiev show up. OTOH, Aaron Copland is one of my favs.)

Thanks for listening and for the nice words! I particularly appreciate the words on the arrangement, as I did spend a lot of time trying to get it right.

On classical music, we are polar opposited indeed While I appreciate Prokofiev, I never managed to like Stravinsky, and the same applies to many 20th century composers. I like the occasional dissonance for effect and where it sounds good (like in the Tristan Chord), but modern classical music is often *just* dissonances, and I really can't stand atonal music, no matter how hard I try... I guess I'm really stuck in the 19th century, with wildly passionate and sometimes over-the-top pieces

This is a well-constructed orchestral piece. Even without your description of the work involved, anyone who has orchestrated, particularly with sampled orchestras, can tell that this was a lot of work, just from listening to the result.

Thanks! Just this morning I've found out MuseScore is revamping their software, and v4 will integrate much better with external VSTs and stuff like that. They explicitly mentioned NotePerformer, which I've been intrigued by for a long time: apparently they manage to get reasonably good sounding results by just working on the score, so taking advantage of articulations as written there, rather than expecting you to do jumps in order to get the outcome you'd like. I'm really looking forward to that, as it would greatly simplify things, the way I write music.

There's just one thing that didn't work for me. That solo trumpet toward the end. It's not the part itself. It's the sound. It's a harsh, static sound that doesn't sit well in a romantic period piece. It could perhaps be softened or given more "passion" with more vibrato, and some volume swells/dips. But I think that you should investigate a different solo trumpet sfz. (The one from No Budget Orch is a more resonant one. That's what I think would work -- a more resonant, fuller trumpet with vibrato).

That's a fair point. That whole part is indeed a bit weird, actually: I liked the idea Ihad come up to, but found it hard to orchestrate, so decided to just keep brass (with more instruments joining in measure after measure), with strings playing legato runs in the background. Instruments left on their own are bound to pop up more, and I agree that the trumpet sound from VPO isn't the best.

I do know VPO borrowed from many libraries, including No Budger Orchestra, though: maybe for trumpets they chose a different one instead of the one you mentioned? I'll probably have to do some more studying there, to see if some alternatives that were "discarded" may actually be better in some cases: I just gave for granted that VPO had taken the best of all worlds, and so I went with that.

Thanks for the feedback!

[Michael Willis]

I have to admit, the first composer that came to my mind was Bedřich Smetana, rather than Tchaikovsky or Grieg or anybody else mentioned on this thread. This was a pleasant listen, the different motifs make me think of scenes in a story, each with their own part to tell. I'll certainly come back for more, as I feel like this kind of music is best appreciated by developing some familiarity with how it flows and the tale it expresses.

I'm glad to hear that my Ardour template was helpful. You are right in treating it as more of a starting point than an end-all-be-all template. I was hoping that it could serve more of the latter purpose, but each time I try doing an orchestral arrangement I find that I want to customize it. For example, you added more woodwind tracks to accomodate your duet parts, which is a common thing to do. As I learn more about virtual orchestration, I am more and more convinced that there are no absolutes on how to make a template. Some arrangements call for a larger-than-life wall-of-sound kind of orchestration while others need a smaller, more intimate sound with less reverb and a lot of attention to the nuances of each instrument.

Anyway, great work on this, it is one of my favorite pieces that you have shared here.

[Rainmak3r]

I have to admit, the first composer that came to my mind was Bedřich Smetana, rather than Tchaikovsky or Grieg or anybody else mentioned on this thread. This was a pleasant listen, the different motifs make me think of scenes in a story, each with their own part to tell. I'll certainly come back for more, as I feel like this kind of music is best appreciated by developing some familiarity with how it flows and the tale it expresses.

I haven't heard much of Smetana, apart from the ultrafamous Moldau that I love, but I do know he's from about the same period, as was his fellowman Dvorak (whose music I really like too). Glad to hear you enjoyed the motifs and the "storytelling"! I've always liked playing with themes, and this time even more so.

I'm glad to hear that my Ardour template was helpful. You are right in treating it as more of a starting point than an end-all-be-all template. I was hoping that it could serve more of the latter purpose, but each time I try doing an orchestral arrangement I find that I want to customize it. For example, you added more woodwind tracks to accomodate your duet parts, which is a common thing to do. As I learn more about virtual orchestration, I am more and more convinced that there are no absolutes on how to make a template. Some arrangements call for a larger-than-life wall-of-sound kind of orchestration while others need a smaller, more intimate sound with less reverb and a lot of attention to the nuances of each instrument.

Very true! I'm pretty sure that even if I'm happy with where I got now, I'll have to make more tweaks for whatever piece I'll work on next, exactly because of what you said: different pieces have different requirements and challenges, and so need different treatment. But your template was one hell of a starting point! As I said, the panning and reverb alone is worth gold! I never would have been able to come up with something so detailed on my own.

Anyway, great work on this, it is one of my favorite pieces that you have shared here.

Thanks, this is really appreciated!

[folderol]

A very pleasant listen, and indeed surprisingly 'authentic'

[Rainmak3r]

A very pleasant listen, and indeed surprisingly 'authentic'

Thanks!

[psyocean]

Masterpiece! Competent and beautiful work, pleasant to listen!

[Rainmak3r]

Masterpiece! Competent and beautiful work, pleasant to listen!

Thanks!

[turbidh20]

Very nice indeed. You are a talented guy

I enjoy listening to come classical occasionally but have no appreciation of composers etc. or of the intricacies involved, so the only thing I can comment on is how much I enjoyed it!

[Rainmak3r]

Very nice indeed. You are a talented guy

I enjoy listening to come classical occasionally but have no appreciation of composers etc. or of the intricacies involved, so the only thing I can comment on is how much I enjoyed it!

Thanks, I couldn't ask for a better review!

[merlyn]

I thought this was good. Well done. The rhythmic figure at around 4:44 jumped out at me -- I recognised it from a Beethoven quartet number 7 2nd movement. . Was that where you got the idea? There was also The Flight of The Bumble Wave at 8:45, which was good.

[Rainmak3r]

I thought this was good. Well done. The rhythmic figure at around 4:44 jumped out at me -- I recognised it from a Beethoven quartet number 7 2nd movement. . Was that where you got the idea? There was also The Flight of The Bumble Wave at 8:45, which was good.

Thanks, I'm glad you liked it!

I never heard that quartet by Beethoven (I'm more familiar with his symphonic work), so the idea for that part didn't come from there: not sure how it came to my mind, but I wouldn't rule out some metal influence considering it might sound like a 'staccato' riff The Flight of the Bumble Wave might have been a more subconscious influence, considering I love Rimsky-Korsakov, but in that part I explicitly planned a chromatic run (like the alternating woodwinds after that), so it may have been a similarity by accident.

[milo]

I think you've outdone yourself! This is a carefully planned and well executed project. It was very ambitious from the start, and I think you've made something to be really proud of.

Regarding the composition, I am not an expert. But I am a fan of this genre, and Tchaikovsky is certainly in my top 10 favorites. I like your themes, and I think they work well to tell the story. The end of the storm at 10:30 seems a little abrupt, but all of the other transitions worked well for me.

It is interesting to read about your workflow and the tools you used and modified. This is truly a Linux audio project, as evidenced by the fact that you edited the sfz files by hand, presumably in a text editor. You geek. (Emacs or Vi? )

I also looked at the MuseScore file. I'm not accustomed to looking at scores for such large ensembles, and that was interesting. The score is the source code for music, especially in classical music, so I was glad to see that you made the source available.

[Rainmak3r]

I think you've outdone yourself! This is a carefully planned and well executed project. It was very ambitious from the start, and I think you've made something to be really proud of.

Thanks a lot for your kind words!

Regarding the composition, I am not an expert. But I am a fan of this genre, and Tchaikovsky is certainly in my top 10 favorites. I like your themes, and I think they work well to tell the story. The end of the storm at 10:30 seems a little abrupt, but all of the other transitions worked well for me.

It is indeed a bit abrupt, but I wanted to experiment with going from a "fortissimo" to a "pianissimo" quickly, as there are different ways of doing that with an orchestra, and I liked how it came out, with the clarinet making the transition smoother, as it starts playing with the whole orchestra and then remains the only one "standing". I don't remember exactly how the technique was called, but I remember reading about it in Samuel Adler's book on orchestration.

It is interesting to read about your workflow and the tools you used and modified. This is truly a Linux audio project, as evidenced by the fact that you edited the sfz files by hand, presumably in a text editor. You geek. (Emacs or Vi? )

Hahaha neither! I think I used Geany the whole time (which is the editor I use for code too, I'm not a fan of IDEs), even though I might have occasionally tweaked a few things with nano

I also looked at the MuseScore file. I'm not accustomed to looking at scores for such large ensembles, and that was interesting. The score is the source code for music, especially in classical music, so I was glad to see that you made the source available.

Yes, I wanted to share it exactly for that, mostly because I was interested in getting feedback on how well it was written in the first place. Unfortunately it looks like my score sucks quite badly: I posted this on a classical forum as well, and a guy pretty much bashed me for my super-amateurish skills. I managed to get some tips on what was particularly wrong, but I don't think I'll ever get to a point where it actually is a believable classical score: the fact I never got any formal training is probably the main reason, and no matter how many books I read or tutorials I follow, I guess it's reasonable to assume I'll never be as good as those who dedicated years of their life to study exactly that.

Anyway, I don't plan that to stop me: I've already started collecting ideas for another symphonic poem (which will sound more like a soundtrack probably, though), and I still have the symphony to finish. I doubt the scores will look any better than this, but hopefully the music will.