[DELETED] [MUL] Celeste: Farewell - "Beyond the Heart" by PlayfulPiano

[PlayfulPiano]

Ok here we go, I might have a lot to say so bear with me.  You've currently got two different files running concurrently which is ... kind of confusing but I'm just going to comment on the most recent one.

-I definitely disagree with the changes to the harmony in bars 7-10 and similar places further on.  I feel like the dissonance in those opening chords is what gives this piece a lot of its tension (along with the ostinato rhythm itself) and just a F chord in bar 7-8 doesn't really do the trick I don't think.  Personally I'd just keep it as you had it F add4 in bars 7-8 and Gsus4 add2 in bar 9-10.
-The left hand in bar 19-20 is F C D E, not F A B C.
-The left hand in bar 29 should be E C D E, not E A B D.
-Check bars 31-44 again for the right hand.  There's lots of wrong rhythms, extra notes, missed notes, flipped layers.  It's probably worth thinking about this section a bit again though as you often have very large intervals between the top and bottom notes in the right hand.  Sometimes it works well (like bar 35) but sometimes a countermelody that sits along with the melody for a lot of the time can be more trouble than it's worth to keep.  (Heterophonic textures do not work well on piano, at least not to my knowledge anyway.)
-The bottom system on page 3 needs respacing as the bass clef is currently obscured behind notes.
-Generally speaking, across the whole sheet the slurs need tinkering with to avoid collisions with other articulations.
-From bar 7 to bar 54, every single right hand note has an accent on it.  This is not really a good use of articulations and could be achieved by using a different dynamic for the right hand (or just by knowing that performers will clearly articulate the melody above the accompaniment without assistance via notation.)  I'm guessing you did it because it makes the playback sound better, but playback should never inform how a sheet looks; you can just hide all of those accents and the playback will still be satisfactory.  (This doesn't just apply to 7-54 by the way, just that was the most egregious example.)
-You could reduce the number of pages by taking a lot of the places where you have two bars per system and swapping them to three bars per system.  At the moment, some bits of the sheet feel very spread out (bars 7-14 for example) whereas others are much denser (bars 15-46).  Spreading everything out more is not really what we want to do since this sheet is already 9 pages long, so it makes more sense to condense the spread out sections.  (Bear in mind that doing this will undoubtedly require some ties to be edited in the bigger chords.)
-On every page after the first, the bottom is awkwardly empty.  I would suggest spreading out the systems better, or trying to go from 5 systems per page to 6 systems per page (which might also help you lose a page in the process).  If you want to try going to 6 systems per page then you'll likely need to mess with the inner space between staves which you currently have set to 1.1 inches
-Bars 55-57 (and also 114-116) is easily the trickiest section of the piece to make work.  The way the piano was presented earlier certainly made this section very awkward and tricky, but now with it gone it's a bit tame.  I'm not 100% sure what the best thing to do is, but if you want to keep it as you have it I'd highly recommend using lots of articulations for those bars.  If it is played plainly, it will be boring and the performer could really do with some help to make it interesting: slurs and staccatos etc.  I think perhaps msf suggested something similar above.
-The G in the right hand of bar 56 should be an F.
-The end of bar 58 sounds a bit fancier to me.  Like this:

Spoiler

You cannot view this attachment.

[close]

-I think the E at the end of 59 should be an octave up.
-Here's my suggestion for bar 60-61.

Spoiler

You cannot view this attachment.

[close]

I preserve the contour of the melody going into bar 60 (moving up to the D rather than down to it) and I include harmony / voice the chords a little more evenly.  (Missing G in bar 60 beat 3 and missing B on both chords in bar 61).  I also don't drop the bass an octave on beat 3 in bar 61 because I don't hear that in the original.
-Slurs should go to the final note in a tie.  i.e. in bar 67 the slur should go to the semibreve in bar 68 instead.  This applies to 75-80 as well.
-I understand the intent behind the first beat in the left hand in bars like 15, 16 etc.  My understanding is that you're trying to mirror the lower string line and then move to the synth on beat 2, but I'm not convinced entirely on how effective it is.  Firstly, it's not particularly accurate as the what that lower string line does changes a lot and it's never just as simple as what you've written, but more importantly I think it gives the accompaniment a kind of stilted feel compared to the original.  Feel free to disagree with me, but I'd suggest thinking about it again or simply just only taking the synth line like you do in later bars (27 etc.)
-The dynamic placement in 73-74 needs cleaning up a little.  At the moment the dynamics aren't placed close enough to the staff they apply to so they kind of look like they're randomly placed on the sheet.
-Speaking of the dynamics, I think that pages 5-6 need a bit of a rethink.  Splitting dynamics is usually ok in small amounts but you need to be careful because it can quickly become very confusing to the reader/performer.  People aren't as good as the playback at remembering exactly what dynamic each hand is meant to be all the time.  For example, you could place the piano dynamic in the centre at bar 66, then place the mp above the staff on beat 3 of bar 67.  The crescendo in bar 73 shouldn't start at the start of the bar since there's nothing on the top staff place there.  In fact, I wouldn't even put a crescendo at all since it only applies to a single note.  Then I think I'd restate that the left hand should still be piano in bar 74 (put the dynamics in brackets) and I'd do a similar thing for the right hand in bar 82.  It'll just make it a lot easier to follow, without having to constantly refer back to previous bars.
-The chords in the ostinato in bars 90-105 don't sound right to me.  I haven't worked them exactly but it sounds like they alternate between Fadd#4 and Gadd2 although there might be some Em7 later on.  Definitely check these again.
-Once again, check the rhythms in the right hand in 90-113.  Some of them aren't quite right, like the A in bar 105 should be on beat 4, not beat 4.5.
-All my previous comments about 55-61 apply to the 114-120 as well.
-I think the chord at the start of bar 62 should be inverted down once.  It sounds like it's an E on top, not an A.
-I don't think the chords in bars 66 onwards are quite right either.  The bass notes should go something like F -> E -> D -> C -> D -> C -> D -> E - > F - > C -> D -> E (taking you all the way up to bar 90).  I'd have another go at these.  Especially in 82 onwards where we haven't got anything else going on, it's important to get the harmonies right.
-The D.S. al coda at the end looks like it's been written in a different font or something?  I don't quite know what's happened there.
-I'd consider dropping the lower octave in the grace note in bar 110.  It's just quite awkward to play it like that.
-In bars 82-89 you mix the melody in with the ostinato which is not particularly easy to read.  I see you used slurs to some extent to alleviate the problem, but I think the best solution is to continue to use two layers here to distinguish them from one another.
-I'm not sure I quite understand the intent behind dropping the semiquavers in bars 82-89.  It feels like we're building up here (all the way from 62 to the end) but losing that motion seems to work against that.  Zeila suggested earlier doing the exact opposite of this, which was introducing the semiquavers later one and leaving them out at the start, which I think would potentially make more sense.  It's definitely worth having another think about this whole build up section to make sure that we're representing it in the best possible way.
-It might be worth pointing out that the glissando in bar 2-3 is actually an octave higher.  You might not want to have the ostinato at that height though, so I can understand keeping it lower if you want.
-Celeste is still published by Matt Makes Games I believe, so I think that's what should be on the copyright.
-Since there is no game entitled Celeste: Farewell, I really think that this should just be titled under Celeste.  The DLC doesn't have a separate release or anything anyway, it's just free content added to the original game.

If you need any help on particulars or want me to go more in depth with something I've said then please ask.  I hope the feedback helps!


Yes, Static just pointed this out to me as well.  It doesn't look like the pdfs of your sheets currently on site so something funny is going on.  If you need me to I can export a pdf for you at the end so it doesn't do this.

I saw this, I just am currently a bit busy with schoolwork so I'll come address this once i'm free. Thanks for giving all of this feedback though.

[PlayfulPiano]

-I definitely disagree with the changes to the harmony in bars 7-10 and similar places further on.  I feel like the dissonance in those opening chords is what gives this piece a lot of its tension (along with the ostinato rhythm itself) and just a F chord in bar 7-8 doesn't really do the trick I don't think.  Personally I'd just keep it as you had it F add4 in bars 7-8 and Gsus4 add2 in bar 9-10.

Been listening a bit more and yeah, I agree with the F add4 chord at least. I do disagree with making it Gsus4 add 2, just due to how the A chord is Asus2 without any 4th note, which makes the transition (at least to me) a bit busy. I made it Gsus4 instead of Gsus2, do you think this is better?

-The left hand in bar 19-20 is F C D E, not F A B C.
-The left hand in bar 29 should be E C D E, not E A B D.

Will change the former, not sure about the latter, maybe sounds C# ish? Not to mention that fingering wise this feels somewhat unnatural to do. What might be an alternative?

-Check bars 31-44 again for the right hand.  There's lots of wrong rhythms, extra notes, missed notes, flipped layers.  It's probably worth thinking about this section a bit again though as you often have very large intervals between the top and bottom notes in the right hand.  Sometimes it works well (like bar 35) but sometimes a countermelody that sits along with the melody for a lot of the time can be more trouble than it's worth to keep.  (Heterophonic textures do not work well on piano, at least not to my knowledge anyway.)

I spotted the wrong F in m39 (now an E), I took out the extra D in m43, and made the Bb a whole note in m44. I don't recognize any other mistakes besides those, so let me know if there are others. After those corrections let me know if it still needs modifying.

-The bottom system on page 3 needs respacing as the bass clef is currently obscured behind notes.

Corrected I think.

-Generally speaking, across the whole sheet the slurs need tinkering with to avoid collisions with other articulations.

Slurs were added mainly as an earlier suggestion but I didn't want to do custom edits since I'm not exactly sure what the rules are regarding over / under for it. Also, do you think they should be required in the first place? (Removed the slurs for now)

-From bar 7 to bar 54, every single right hand note has an accent on it.  This is not really a good use of articulations and could be achieved by using a different dynamic for the right hand (or just by knowing that performers will clearly articulate the melody above the accompaniment without assistance via notation.)  I'm guessing you did it because it makes the playback sound better, but playback should never inform how a sheet looks; you can just hide all of those accents and the playback will still be satisfactory.  (This doesn't just apply to 7-54 by the way, just that was the most egregious example.)

I don't believe I can hide accents via Allegro 2007, but I can remove them for now. Question though: what would be an articulation for the opposite of an accent / muting a note? Need to know this so I can apply it to the softer notes in the mid/late section of the arrangement. (Unmodified until I know about the articulation)

-You could reduce the number of pages by taking a lot of the places where you have two bars per system and swapping them to three bars per system.  At the moment, some bits of the sheet feel very spread out (bars 7-14 for example) whereas others are much denser (bars 15-46).  Spreading everything out more is not really what we want to do since this sheet is already 9 pages long, so it makes more sense to condense the spread out sections.  (Bear in mind that doing this will undoubtedly require some ties to be edited in the bigger chords.)

I did the two bar per system at points mainly due to me referencing bespinben's solo primal dialga arrangement and making sure notes and their rhythms are legible, but I can reduce them. I also did it based on 2 or 4 division so each cycle can be basically set to each new system. Should be better now.

-On every page after the first, the bottom is awkwardly empty.  I would suggest spreading out the systems better, or trying to go from 5 systems per page to 6 systems per page (which might also help you lose a page in the process).  If you want to try going to 6 systems per page then you'll likely need to mess with the inner space between staves which you currently have set to 1.1 inches

Done, down to 7 pages. Hopefully it looks fine.

-Bars 55-57 (and also 114-116) is easily the trickiest section of the piece to make work.  The way the piano was presented earlier certainly made this section very awkward and tricky, but now with it gone it's a bit tame.  I'm not 100% sure what the best thing to do is, but if you want to keep it as you have it I'd highly recommend using lots of articulations for those bars.  If it is played plainly, it will be boring and the performer could really do with some help to make it interesting: slurs and staccatos etc.  I think perhaps msf suggested something similar above.

Added some staccatos to the right hand, besides that not sure what else to do. Slurs not doing as per the previous quote.

-The G in the right hand of bar 56 should be an F.
-The end of bar 58 sounds a bit fancier to me.  Like this:

Spoiler

You cannot view this attachment.

[close]

-I think the E at the end of 59 should be an octave up.

Fixed.

-Here's my suggestion for bar 60-61.

Spoiler

You cannot view this attachment.

[close]

I preserve the contour of the melody going into bar 60 (moving up to the D rather than down to it) and I include harmony / voice the chords a little more evenly.  (Missing G in bar 60 beat 3 and missing B on both chords in bar 61).  I also don't drop the bass an octave on beat 3 in bar 61 because I don't hear that in the original.

Some parts I do agree others I disagree. I do think putting it up an octave is a good idea, but instead of keeping the midpoint notes at the same position, I moved them down so note wise they're the same octave. Due to this, I didn't change the actual chords of the right hand though. Also added a missed note in beat 4 (another low G#). Left hand I moved up an octave (accuracy) and kept it at the same octave for the second e.

-Slurs should go to the final note in a tie.  i.e. in bar 67 the slur should go to the semibreve in bar 68 instead.  This applies to 75-80 as well.

Will do slurs later.

-I understand the intent behind the first beat in the left hand in bars like 15, 16 etc.  My understanding is that you're trying to mirror the lower string line and then move to the synth on beat 2, but I'm not convinced entirely on how effective it is.  Firstly, it's not particularly accurate as the what that lower string line does changes a lot and it's never just as simple as what you've written, but more importantly I think it gives the accompaniment a kind of stilted feel compared to the original.  Feel free to disagree with me, but I'd suggest thinking about it again or simply just only taking the synth line like you do in later bars (27 etc.)

I've been feeling around and I disagree on removing it, but I do think adding staccatos to the strings is better. Let me know if that seems better.


Have to do some work so this is as far as I've done change-wise. I have an idea that should probably be better in the mid-late section, just haven't gotten to it yet. But for the above changes, the portal is updated.

[PlayfulPiano]

Alright, going through the rest of it now.

-The dynamic placement in 73-74 needs cleaning up a little.  At the moment the dynamics aren't placed close enough to the staff they apply to so they kind of look like they're randomly placed on the sheet.
-Speaking of the dynamics, I think that pages 5-6 need a bit of a rethink.  Splitting dynamics is usually ok in small amounts but you need to be careful because it can quickly become very confusing to the reader/performer.  People aren't as good as the playback at remembering exactly what dynamic each hand is meant to be all the time.  For example, you could place the piano dynamic in the centre at bar 66, then place the mp above the staff on beat 3 of bar 67.  The crescendo in bar 73 shouldn't start at the start of the bar since there's nothing on the top staff place there.  In fact, I wouldn't even put a crescendo at all since it only applies to a single note.  Then I think I'd restate that the left hand should still be piano in bar 74 (put the dynamics in brackets) and I'd do a similar thing for the right hand in bar 82.  It'll just make it a lot easier to follow, without having to constantly refer back to previous bars.

Honestly, the whole dynamic split is due to me using playback heavily. Due to what you said before with not basing things on playback, i'll likely just not have the split at all and redo the dynamics towards the end of the arrangement process. (unchanged)

-The chords in the ostinato in bars 90-105 don't sound right to me.  I haven't worked them exactly but it sounds like they alternate between Fadd#4 and Gadd2 although there might be some Em7 later on.  Definitely check these again.

I think I got it. Fadd#4, Gadd2, Amadd2, Cmaj7 (2nd inv, so GBCE).

-Once again, check the rhythms in the right hand in 90-113.  Some of them aren't quite right, like the A in bar 105 should be on beat 4, not beat 4.5.

Just looked through it again, should be right now (although baring might be off).

-All my previous comments about 55-61 apply to the 114-120 as well.

Applied.

-I think the chord at the start of bar 62 should be inverted down once.  It sounds like it's an E on top, not an A.

Yeah I agree, it's just A-E I think.
And I just painstakingly looked again at m60/61, and turns out I got the notes wrongish for the right hand.
Should be FD E BbF G EA B A EG# G#. Since the FD will overlap with the left hand then, it'll also have the left hand's Bb on top.

-I don't think the chords in bars 66 onwards are quite right either.  The bass notes should go something like F -> E -> D -> C -> D -> C -> D -> E - > F - > C -> D -> E (taking you all the way up to bar 90).  I'd have another go at these.  Especially in 82 onwards where we haven't got anything else going on, it's important to get the harmonies right.

These chords are so freaking hard to figure out. I have something in place now but honestly it's getting really frustrating to figure out what's right. Gonna ask lena on twitter and see if she could possibly tell me what they are or not.

-The D.S. al coda at the end looks like it's been written in a different font or something?  I don't quite know what's happened there.

That's using the default font for repeats on allegro, same with any other repeat text. It doesn't look like I can actually edit the text though, or I can't find where I can edit it.

-I'd consider dropping the lower octave in the grace note in bar 110.  It's just quite awkward to play it like that.

Done.

-In bars 82-89 you mix the melody in with the ostinato which is not particularly easy to read.  I see you used slurs to some extent to alleviate the problem, but I think the best solution is to continue to use two layers here to distinguish them from one another.

Iirc the reasoning was because the ostinato and the melody didn't overlap positioning wise but are in the same range of notes, causing the beaming to be weird. I'll try making it two layers again, though.

-I'm not sure I quite understand the intent behind dropping the semiquavers in bars 82-89.  It feels like we're building up here (all the way from 62 to the end) but losing that motion seems to work against that.  Zeila suggested earlier doing the exact opposite of this, which was introducing the semiquavers later on and leaving them out at the start, which I think would potentially make more sense.  It's definitely worth having another think about this whole build up section to make sure that we're representing it in the best possible way.

Yeah I've been brainstorming a bit here in terms of what to do. Things I think I want to do:
-Repeat 1 being left hand chords and right hand pad in the correct octave (starting at F4), with the melody replacing the pad. Issue is there will be some layer overlapping unless I shrink the duration of some of the melody's notes. {Alternative is left hand chords & pad octave down, right hand is just melody}
-Repeat 2 being left hand chords and pad (octave down so starting at F3), with the right hand being the A5 ostinato combined with the melody. Keeping both the melody and the ostinato together would mean that you require really fast hand motion due to the gap, not to mention that you have to time two different rhythms together with one hand (although none of the gap notes are played simultaneously). {Alternative is having the ostinato not play during the melody}
-Repeat 3 being left hand chords and pad (octave down so starting at F3), with the right hand being the octave ostinato combined with the melody (melody replaces ostinato). This should be fine, just need to figure out how to get the layers right.
Also, I'm also thinking if I should try having the pad (correct octave, starting at F4) in the right hand when the melody isn't playing for m90-m105.
Let me know which aspects of the above you think is the best call for the sheet.
For now, the current edit will be the following:
R1 - Left hand chords, Right hand pad (F4) & Melody
R2 - Left hand chords & pad (F3), Right hand Ostinato & Melody (Overlapped)
R3 - Left hand chords & pad (F3), Right hand Ostinato & Melody
m90-m105 - Left hand chords & strings, Right hand Melody

-It might be worth pointing out that the glissando in bar 2-3 is actually an octave higher.  You might not want to have the ostinato at that height though, so I can understand keeping it lower if you want.

Was there initially but I brought it down an octave so the strings can be played / had some layer and note changes when I first did this. Bringing it back but modified in a way which I think should work.

-Celeste is still published by Matt Makes Games I believe, so I think that's what should be on the copyright.

Matt Makes Games and Extremely OK Games are the same company just rebranded. I'm fairly sure that in turn means the latter is what published Celeste. As per their twitter bio: "Creators of @celeste_game and #TowerFall. Purveyors of digital amusements!"

-Since there is no game entitled Celeste: Farewell, I really think that this should just be titled under Celeste.  The DLC doesn't have a separate release or anything anyway, it's just free content added to the original game.

While it is true that the DLC wasn't published separately, this track does not come from the official Celeste OST, rather it is part of its own OST specific to the DLC. I was using the OST distinction since this is about the music, not the base game necessarily.
Question: Would you consider the tracks from Hollow Knight: Gods & Nighmares (Separate OST to the original, 15 tracks, over 30 minutes of music) count for the orignal game or count for a separate listing due to it being in a different OST?

Yes, Static just pointed this out to me as well.  It doesn't look like the pdfs of your sheets currently on site so something funny is going on.  If you need me to I can export a pdf for you at the end so it doesn't do this.

PDFs of mine on the site didn't come from my computer, rather an updater made the files using their more up to date version of finale. Usually due to some feature that I didn't have in Allegro 2007.

Also besides the above, did some other general corrections with grace notes and the likes. Did the massedit rebar music & note spacing, so if there's any issues it's due to finale.

Anyways, portal updated again.

[PlayfulPiano]

So since the last portal update, branflakes created his VRC6 cover of beyond the heart and uploaded it to youtube. Finally having another person's arrangement to compare, I made some changes that should now be more reflective of the original.
His cover for reference:

-Left hand FABC chord (m7 example) reverted again to an Fmaj chord, branflakes also used this and I think the B that we're hearing comes from the pad and not the actual strings. This also makes sense when you think about how the a chords for strings use 3 notes, not 4 (similarly with the GACD {m9 example} or whatever chord, it should be GCD).
-Left hand went all in regarding the beat one fifth strings for section 2 (so m26-m30 and m43-46), and made the last two measures per repeat of this section BE / BCDE and AE /AC#DE as per how bran did it.
-m55 / m114 beat 1 is an E, not F, so this is a correction.
-m60-61 & m119-120 changed again using what branflakes did (so octave right hand and chord left hand), which now sounds extremely accurate to the original.
-m66-105 left hand: Finally got the right chords, or at least the right chord bases. This is a slight derivation of the original due to the note gap / it being piano, but overall the chords that get used are Fmaj7, Emin7, Dmin7, and Cmaj7 (note that Fmaj7/Dmin7 both have the Fmaj chord and that Emin7/Cmaj7 both have the Emin chord).
The order goes F E D C | D C D E | F C D E | (m90) F E D C | D C D E.
For Fmaj7, I'm omitting the 7th (E) since it tends to be complimented in the right hand playing in Amin (so I do an FCA chord).
For Emin7, I'm omitting the 7th (C) since it tends to be complimented in the right hand (so I do an EBG chord).
For Dmin7, I have two variants. One when it's the third of a phrase, the other when it's the first of a phrase. Since the first of the phrase tends to be the F chord, I used FDA. During the third chord in the phrase, I used DAE.
For Cmaj7, I'm omitting the 1st (C) since that is actually not used, rather the C either gets played by the right hand or the C is below this actual chord (that chord being GEB).
-I double checked with what branflakes did for m90-m105, and the notes I have are the same as his +1 base note usually. So for this section i'm keeping the FABC, GBCD, ABCE, and GBCE chords. Might need this changed later, might not, not yet sure.

Large shoutout though to bran for telling me the chords he used, it seriously made my life easier with the 4th section of the track.

[PlayfulPiano]

Hey so it's been a while.

I've been looking over the arrangement a bunch, practicing it for my own performance in a few month's time, and I've made some changes / corrections and overall improvements based off of listening to the original (not other's arrangements) to the point where now I can comfortably say with confidence that the arrangement is finished.

-Dynamics have been overhauled, significantly less accents throughout the arrangement, dynamics are clearer (above the system = right hand, below the system = left hand, between the staves = both hands), notes are accurate and I feel that the arrangement is ready.

You can take also just a direct listen here as I did convert it over on musescore: https://musescore.com/user/6765551/scores/5716526

[mastersuperfan]

Great arrangement of a great track! It's been a long time since I saw this sheet months ago, but having seen it then, I can say that your revisions definitely paid off.

While I think most of the choices you've made have helped, I actually much preferred the way you arranged m55-m57 and m114-116 in the version I saw back in September. I think the constantly moving piano lines really kept the momentum going even though the section was softer. Right now, before m58/117, the cello line alone isn't really doing it for me; I think it feels slower and more empty than it should.

I also recognize that you're trying to capture the strings with the first beat in the LH from m15-46, but I don't really think it's working as desired. The piano can't replicate the "sharp" sound of the strings, which is what helps keep the fast momentum, so instead, I think the eighth notes actually slow the piece down. I personally think it would just be best to go with the sixteenth-note synth the whole way through. As it is now, the LH rhythm almost gives me a Western desert/cowboy vibe instead... (For the record, I didn't even notice the strings on beat 1 until you pointed them out.)

In addition, I think this would be a really cool idea for m53-54 and m112-113:

By bringing back the synth at the end (but quietly), I think this helps keep the piece moving and fills the empty space while also transitioning well to the next part. Obviously you don't have to use this, but I personally think it would be good for both m53-54 and m112-113. (Alternatively, you could only use it in m112-113 to distinguish the two phrases more from each other.)

Finally, here are a lot of smaller fixes and suggestions:

- Was it an intentional stylistic choice to include held some dotted eighth notes in m1-6, instead of making them sixteenth notes/staccato'd eighth note like in the original? It works as is but I would personally prefer the latter.
- Slightly nitpicky but I feel like your dynamic markings are always aligned a little too far to the left, compared to the notes on beat 1. (Is that where Finale places them by default?)
- Perhaps add a grace note to the top RH note in m34 and m43?
- Is there any particular reason why the held RH notes in m7, m47, m49, and m106 aren't held to the next measure? They are prolonged until the next note in the original.
- As far as articulations go, I would do something like this for m55-57:

- I would also put staccatos on two eighth-note octaves in m60 (potentially with slurs starting on previous notes and ending on them as well).
- The way the two layers are set up in the RH of m67, m69, and m71 looks very strange. I can see what you're going for in the MUS with the color-coding, but on the PDF, it looks to me like just a single layer with very weird beam alignment. I would just make those measures into one layer; I think the accents do a good enough job of indicating which notes are part of the melody. (Alternatively, in m71, where ostinato 16th notes beamed to a melody note might obscure the line a bit, you could keep it the way it is but show the 2nd layer eighth rest at the end.)
- The RH note on beat 3 of m67 has no accent—it seems to me that there was supposed to be an accent there, perhaps?
- The accent on the first note of m70 is colliding with the slur in the PDF (although it's fine when I open it in Finale). Same goes for m86 and m110.
- For m73, I would still put a B on beat 3 of Layer 2 just so there's no ambiguity about the seemingly "missing" quarter of a beat in Layer 2 (or, if not this, then put a sixteenth rest in Layer 2). For example:

- Sixteenth rest in m81 RH is colliding vertically with the quarter note in beat 3. The easy fix is to move the rest up a little bit.
- As in m67/69/71, I would just turn m83, m85, m87, and m89 into a single layer. As two layers, I'm finding it harder to read than it should be, and as before, I think the accents alone work well enough.
- Is the eighth rest in m90 RH necessary? I'm not sure I'm really hearing a break there. Same goes for m102.
- Is there any particular reason not to hold RH notes like in m96, m98, m100 to the next measure? They are still faintly ringing through the next measure in the original track, and I think it'd be nice to keep them held down even when the pedal changes.

I don't know if I'm hearing things, but I think there's a really faint G on beat 4.5 in the RH of m97.

- I'm still hearing this—although I think it's actually on beat 4.25, not 4.5—so I checked the 8-bit arrangement and sure enough, branflakes included it too. So I'm pretty sure it's there.
- There's also a very faint C in the RH at the end of m102 (i.e. B and C sixteenth notes, instead a single B eighth note). It's in branflakes' version too.
- There's a sixteenth-note delay (where the previous notes are still tied) at the start of m108 RH.

EDIT: Also, for the record, I think the best way to list the sheet on-site would be to put it under the same Celeste section, but keep the subtitle on the sheet as Celeste: Farewell. I think most visitors to the site would look for this under Celeste before searching for a separate section for a DLC expansion that isn't even its own game.

[PlayfulPiano]

Alright, just saw this so I'll go respond to the points. Will make edits later (just got back into my dorm tonight).

Great arrangement of a great track! It's been a long time since I saw this sheet months ago, but having seen it then, I can say that your revisions definitely paid off.

While I think most of the choices you've made have helped, I actually much preferred the way you arranged m55-m57 and m114-116 in the version I saw back in September. I think the constantly moving piano lines really kept the momentum going even though the section was softer. Right now, before m58/117, the cello line alone isn't really doing it for me; I think it feels slower and more empty than it should.

I decided against using it mainly due to the sudden difficulty of the section vs. the rest of the measures nearby it. It didn't feel like it flowed as well performing, well, if I could perform that section. Luckily if going the route of using the accurate piano for that section is preferred, Lena did post on her twitter the actual notation of these measures which will make transcribing it a breeze compared to before. I would *think* using the composer's own transcription of a section of their theme is fine to do.

I also recognize that you're trying to capture the strings with the first beat in the LH from m15-46, but I don't really think it's working as desired. The piano can't replicate the "sharp" sound of the strings, which is what helps keep the fast momentum, so instead, I think the eighth notes actually slow the piece down. I personally think it would just be best to go with the sixteenth-note synth the whole way through. As it is now, the LH rhythm almost gives me a Western desert/cowboy vibe instead... (For the record, I didn't even notice the strings on beat 1 until you pointed them out.)

Pace wise you're probably right, but I just can't shake off the need to include the strings in some form. It really feels like it moves the section forward if that makes any sense. Feels sort of boring / stale without some aspect added to the first beat.
Maybe there are some other alternative ways to notate it in a better form so it doesn't ruin the pace but still keeps some focus of that first beat?

In addition, I think this would be a really cool idea for m53-54 and m112-113:

By bringing back the synth at the end (but quietly), I think this helps keep the piece moving and fills the empty space while also transitioning well to the next part. Obviously you don't have to use this, but I personally think it would be good for both m53-54 and m112-113. (Alternatively, you could only use it in m112-113 to distinguish the two phrases more from each other.)

Great idea, will include.

- Was it an intentional stylistic choice to include held some dotted eighth notes in m1-6, instead of making them sixteenth notes/staccato'd eighth note like in the original? It works as is but I would personally prefer the latter.
- Slightly nitpicky but I feel like your dynamic markings are always aligned a little too far to the left, compared to the notes on beat 1. (Is that where Finale places them by default?)
- Perhaps add a grace note to the top RH note in m34 and m43?
- Is there any particular reason why the held RH notes in m7, m47, m49, and m106 aren't held to the next measure? They are prolonged until the next note in the original.

-What finale gave me as the default rebarring iirc. Helps separate individual beats too imo.
-Honestly finale allegro messes up dynamics a bunch for me since I rely a lot on playback & it doesn't work well with staff specific vs. system specific dynamics. I usually just default their positions to "beginning of the measure" option to keep it consistent. I don't move it left/right at all, only up/down.
-What are the grace note pitches for these parts here? I hear what you're talking about, thanks for the catch.

- As far as articulations go, I would do something like this for m55-57:

- I would also put staccatos on two eighth-note octaves in m60 (potentially with slurs starting on previous notes and ending on them as well).
- The way the two layers are set up in the RH of m67, m69, and m71 looks very strange. I can see what you're going for in the MUS with the color-coding, but on the PDF, it looks to me like just a single layer with very weird beam alignment. I would just make those measures into one layer; I think the accents do a good enough job of indicating which notes are part of the melody. (Alternatively, in m71, where ostinato 16th notes beamed to a melody note might obscure the line a bit, you could keep it the way it is but show the 2nd layer eighth rest at the end.)
- The RH note on beat 3 of m67 has no accent—it seems to me that there was supposed to be an accent there, perhaps?
- The accent on the first note of m70 is colliding with the slur in the PDF (although it's fine when I open it in Finale). Same goes for m86 and m110.

-Will do
-Will do
-I think I might be able to keep the differentiating layers on my end but just flip the beaming so it looks better on pdf. Let me know the spots where I should specify a rest though and not have it hidden.
-From what I've heard on playback, it always sounded like the first note of the melody was softer than the rest intentionally. Hence why it lacks an accent.

- For m73, I would still put a B on beat 3 of Layer 2 just so there's no ambiguity about the seemingly "missing" quarter of a beat in Layer 2 (or, if not this, then put a sixteenth rest in Layer 2). For example:

- Sixteenth rest in m81 RH is colliding vertically with the quarter note in beat 3. The easy fix is to move the rest up a little bit.
- As in m67/69/71, I would just turn m83, m85, m87, and m89 into a single layer. As two layers, I'm finding it harder to read than it should be, and as before, I think the accents alone work well enough.
- Is the eighth rest in m90 RH necessary? I'm not sure I'm really hearing a break there. Same goes for m102.
- Is there any particular reason not to hold RH notes like in m96, m98, m100 to the next measure? They are still faintly ringing through the next measure in the original track, and I think it'd be nice to keep them held down even when the pedal changes.
- I'm still hearing this—although I think it's actually on beat 4.25, not 4.5—so I checked the 8-bit arrangement and sure enough, branflakes included it too. So I'm pretty sure it's there.
- There's also a very faint C in the RH at the end of m102 (i.e. B and C sixteenth notes, instead a single B eighth note). It's in branflakes' version too.
- There's a sixteenth-note delay (where the previous notes are still tied) at the start of m108 RH.

-I'll do the rest b/c I always had layering / spacing issues whenever two notes overlapped on the same pitch. Looks better imo as well.
-Will fix.
-Same as before. Just let me know where I should put rests.
-Again since I base it a lot on playback, with the idea that con pedale goes throughout the piece more or less, the need to notate beyond a measure sometimes is lost to me. Also, I usually am fairly heavy on the pedal when I practice this piece myself, so it usually just rings on its own thanks to sustain. If needed I can change this.
-Same as above.
-I am having a lot of trouble hearing this. Sometimes I can, sometimes I can't hear any trace. I don't know if it is there or not.
-Ok, that I can hear. I'll fix this.
-Nice catch, will fix.

EDIT: Also, for the record, I think the best way to list the sheet on-site would be to put it under the same Celeste section, but keep the subtitle on the sheet as Celeste: Farewell. I think most visitors to the site would look for this under Celeste before searching for a separate section for a DLC expansion that isn't even its own game.

-Alright, that's a good compromise imo. Will fix it upon doing the update.


Thanks for all the feedback btw, glad to get some now before I actually go perform this in a few weeks for my college, potentially.

[mastersuperfan]

Bump for arranger.

[PlayfulPiano]

Bump for arranger.

Hey there, sorry about not working on this. College started as I was about to work on this and I haven't gotten time to do any arrangements until fairly recently, and City of Tears was much closer to approval than this one.

Would it be alright if this can get archived so I can resubmit it with your suggestions a different day?

[mastersuperfan]

No problem. Archiving now.