greetings. in this blog post i will compare the musical structure of Twinkle Twinkle Little Star against that of Camellia - crystallized, neither of which are entries in the Locus contest

Most of the music in osu! is from outside sources. The Locus (2025) contest changes that. Teams of people assemble to write brand new original music, and make a visual showcase to go along with it, and a full spread of maps to play it to.

This is an incredibly difficult contest. The amount of effort, skill and time it takes to come up with an entry at all, let alone a good one, is formidible. I am impressed by everyone for doing their best, with lots of work to do and a deadline to do it in, without even being able to ask for mods from the outside world.

The contest has concluded, and the winners will be decided by player votes. I've been assigned 5 entries to play and rank. While the contest just requires me to numerically rank how much I liked the entries, I found that I had a lot more feedback than a simple rating, so I decided to collect it here. These are listed in the order I played them.

I've embedded inline videos for you, courteously rendered by o!rdr. However, these embedded videos have a small intro before the song begins, so whenever I reference a timestamp, you'll have to go ~5 seconds forward in the video to find the part I'm talking about... or longer if the song has an intro. For this reason, I recommend downloading the map and checking it in game! My replays are linked at the bottom of each section.

church of ppy - LONG LIVE THE KING [beatmap]

It opens with a very dramatic visual sequence, which stunned me, but it becomes less showy and more disorganised over the course of the song. They stop storyboarding at around 0:30, which is a shame, because I feel there's a lot they could have done with the visuals for the rest of the map. The black and white outline trick could have made a return, or they could have used tinting like the red section in Krewella - Live for the Night (Nightstep Mix), hosted and storyboarded by Gero. They probably just ran out of time. It happens.

I don't really know how to feel about the music. It's odd. After a couple of plays, it grew on me, but coming back to it after a week, damn, this is pretty obnoxious actually. I wish the lyrics were storyboarded, because I can't hear them with my ears. There's a few osu! references in the background, so maybe the music is also telling a story about the game, but I can't understand it.

My preferred diff is Insane, because at 1:25 they forget that they're mapping an Insane and start mapping an aim control Another diff. Fun! In Expert, I feel like the aim velocity and slidershapes don't reflect the intensity or emotion of the music very well, and as the song continues, the map begins to feel less intentional and more spammy, particularly in the section at 2:20. The short sliders leading into streams clutters the map and makes reading a lot harder for no real reason. I don't feel it adds dynamics to the music, I would have preferred more interesting stream shapes or spacing instead. In general, I feel like the different types of music in the music aren't represented differently, except for when they are, and then it feels like I'm playing a different song and map entirely. Unfortunately, they're mapping high contrast as high difficulty instead. The accelerating BPM part of the music is wasted because it's just mapped as 1/1 sliders.

Or maybe I'm not good enough at the Expert to judge it properly? You decide.

• Insane [web/dl] - [A] 687k - 96.32% 531x - 35 ok 2 miss
• Expert [web/dl] - [B] 346k - 85.62% 319x - 110 ok 43 miss
• Visual settings: 70% dim, storyboard on, hitsounds on, Rosula skin

Overall opinion: Fairly impressed! This seems like a fun start to the contest.

hdm enjoyers - unbound [beatmap]

It sounds like a Dance Dance Revolution song. This statement is intended to be derogatory. It's kind of like you took Redside's characteristic stuttering technique from Sidequest and made an industrial dance music song with it, but it's not even industrial, it's like, annoying, lopsided, grating, repetitive, piercing, unyielding.

I commented that the previous song's Expert felt like the mapper was spamming, but here it feels like I'm the one spamming to try to play it. It's very irregular and I don't find it fun. With +HT you can hear how poorly matched to the music it is. The background video is like a procedurally generated something that gives me a headache if I leave it on while playing the map. While it kinda looks cool standalone, it also comes off as a bit lazy to me. Even if they wrote and rendered the algorithm themselves (likely), they didn't try to match it to the music, which is a big opportunity loss for creativity. I would have liked to see a similar technique as Raphlesia & BilliumMoto - My Love, video by Noffy where it kind of moves in time to the song and has distinct shapes for the verse, chorus, and bridge.

I really don't like this one at all. The music doesn't appeal to me, the background hurts my head a bit, and I can't play the dang map. Sorry!

• Insane [web/dl] - [A] 650k - 93.31% 454x - 58 ok 4 miss
• Expert [web/dl] - [C] 242k - 78.43% 187x - 181 ok 40 miss
• Visual settings: 86% dim, video off, hitsounds on, argon pro skin

:owoyay: - against cacophony [beatmap]

Wow, I really like this song! Although I can't understand the lyrics, I like how they're storyboarded, and I like how the synth sounds. The music is very layered with lots of instruments working together. The only part I don't really like is the hard beats section that begins at 2:07. It feels out of place compared to the rest of the song, and the complimentary hitsounds are so loud that they amplify the dissonance further. The storyboard is good, it feels very traditional rather than bespoke, but that's fine. It provides visual interest while not distracting from gameplay. There's some fun styles and a visualiser section. I like the little sweeping transitions between each section of the storyboard too. Great!

The mapping plays nicely too. The Insane diff is light slider tech and the Expert diff is tough slider tech and finger control, which feels like a natural escalation of difficulty. I don't know the right words to explain this, but the slider tech movement felt fun and clean here whereas in LONG LIVE THE KING it felt overwhelming and disorganised. Both are fun to replay, they involve the music well, and I don't have any real complaints apart from (again) the accelerating BPM section, which this time I feel is overmapped. I think my favourite way for accelerating BPM to be mapped is to gradually lighten on finger strain and reduce bursts while increasing the required aim velocity, usually with fast sliders or big spacing, to make it feel faster while still giving the player enough room to not hit a burst of 5x100s. It's also important not to introduce too many new ideas or rhythms too quickly, because the player's brain is mostly focused on trying to tap to the BPM, and has less capacity available to understand any new ideas being added. I couldn't find a brilliant example of what I was looking for, but you can kind of get the idea from 2:35 in Kurokotei - Galaxy Collapse, mapped by Corne2Plum3 [Not Galactic].

• Insane [web/dl] - [A] 716k - 97.31% 422x - 23 ok 2 miss
• Expert [web/dl] - [A] 434k - 90.08% 289x - 104 ok 19 miss
• Visual settings: 70% dim, storyboard on, hitsounds on, argon skin

The day after writing this, I found myself quietly whistling the song, and after I noticed this, it took me a moment to realise where the song was from. After a moment, I realised that it might have been that one song from Locus, so I opened the game to check, and yep! It's got to be a great song then, if it managed to add itself to my mental library.

we do well with deadlines - Pondering Magic [beatmap]

It doesn't do much bad but it doesn't do much well either. Overall, it feels underproduced and mid. This midness can be seen across all three categories.

Most obviously, this map has no visual storyboard or video, only an image background. Despite having Basically Nothing, I'm still giving this a rating of 3/5 in visuals, because it's a Nice image, and because there's no intense motion that could make me dizzy.

Similarly, the movement in the map feels alright, but I have the strange feeling that it's derivative, like the mapper is copying patterns they've seen without really knowing why they work or what music they're supposed to fit against. It generally fits the music, but I'm not sure about those 13 note streams that they threw into Expert, which don't seem to have any musical backing. The hitsounds are poor.

To me, the most notable quality of this production is the melody line in the music, which I find notably amateurish. I think it might be difficult to understand what I mean by amateurish if you don't have a background in music writing yourself, but I'll do my best to explain.

Common features can be seen between many songs, even across genres, like the verse/chorus structure, question and answer phrases, rise and fall, standard chord progressions, an anchor melody that other parts of the song relate back to, and everything coming in pairs. I'll give a couple of examples.

Twinkle Twinkle Little Star, despite its deceptive simplicity, has many of the above to create an inoffensive, very easy to learn, and memorable song, even when only considering the most widely known first verse in isolation. Let's check it line by line. Yes, really. This is what we're doing now in this blog post.

Twinkle, twinkle, little star,

The very first note in the song is the tonic, the base note of the key that the song is in. It starts at the bottom of the scale. The word "twinkle" is sung as the same pitch twice in a row, once for each syllable. This shows that it's okay, even sensible, to have the same note twice in a row, when it supports the structure of the song. Over the course of this phrase, the pitch rises stepwise. The last word, "star", ends on the dominant (5th) note of the key, a natural sounding note to take the song to. This rising pitch across the phrase reflects the rising pitch of one's voice when asking a question, and ending on the dominant provides an unresolved note to continue on from.

How I wonder what you are.

This phrase connects to the previous one by having its pitch fall, which contrasts with the rising pitch of the first phrase. It goes step by step across every single note back to the tonic, which makes it easy to sing, because it's much easier to sing a gradual change in notes rather than jumping around the scale. Ending on the tonic (1st note of the scale) brings it back to where it started. It's a concluding pitch, very firm, very final, and so it musically "answers" the "question" that the first phrase asked.

That's been two phrases, and everything comes in pairs, so now it's time to do something different.

Up above the world so high,
Like a diamond in the sky,

This phrase has the opportunity to start something new, and it does so. It reaches a higher note and slowly goes downwards, but at the end of the phase, it only reaches the supertonic (2nd note of the scale), again leaving an unresolved conclusion. You want to step down to the tonic (1st note) to resolve the phrase, but you can't, not yet. It keeps the tension, leaving you hanging.

This phrase repeats in the exact same melody, because everything comes in pairs.

Twinkle, twinkle, little star,
How I wonder what you are.

These phrases are sung identically to the first phrases. The first note is the tonic, which resolves the tension from the previous phrase. And it ends on the tonic, providing a natural resolution to the whole song. It goes up and it goes down. Wonderful.

There are patterns to see throughout the whole song, too. Each note is sung twice in a row. Every single one. The rhythm is also the same in every phrase: short short short short short short pause. Repeat.

There's actually a huge amount of structure in this seemingly simple song that you might not have thought about before! While this song is, yes, for babies, the structure is what supports its simplicity yet also what makes the song good.

Okay, okay, I get it. You're a rhythm gamer. You're not here to read a musical analysis of Twinkle Twinkle Little Star, guest diff by blissfulyoshi. You want something higher concept. I understand.

How about we look at Camellia - crystallized. I'd like you to listen in particular to the section from 1:26 to 2:10, which I will be examining through the lens of breaking things in half. This will be difficult to explain in text, because it's difficult to describe the sound of music with words, but whatever. You can listen along in the linked video and try to figure it out.

There are 32 bars of music in this selection, and they can be cleanly divided down the middle into 16 bars of one distinct section and 16 bars of another distinct section.

The first 16 bars can again by divided into 4 sets of 4 bars. Each of these sets of 4 bars sounds very similar and predictable, though there are small differences in each (mostly in the base pitch). The repetition and consistency drives the song and makes it predictable and comfortable, while the small differences add interest to stop it from actually feeling repetitive and stale.

Each of these 4 bars can again be split in half, with each half being different, and now they're small enough to actually describe. For the first 2 bars, the most prominent melody line is very very straightforward: it rocks up and down by a single note, with an inbalanced rhythm. The second 2 bars let this melody fall away into an "ooh" vocal, dubstep growl, and lower pitched pattern.

Words aren't doing this justice. crystallized looks like this when mapped out:

So in the first 16 bars, there's a 2 bar question and a 2 bar answer phrase. This repeats 4 times with slight variations, but really, it repeats 2 lots of 2 times. The 1st and 3rd variations are more similar to each other, while the other 2nd and 4th variations are more similar to each other. This means that the larger difference between the 1st variation and the 2nd variation kind of forms a question and answer of its own. The largest variation is bars 15-16: because these need to connect through to the next larger section, which is substantially different, they need to depart from the previous pattern in order to build more of a connecting phrase.

This structure makes the song. These techniques, and many others that I haven't touched on here, are the sorts of things you can learn from music theory, whether formalised in textbooks or whether figuring it out on your own by examining songs you like. And if you don't have any of this theory, any of this structure...? Well, you get Pondering Magic. I think the composer lacks music theory background.

I can't find any pairs, any structure, just a meandering melody line that can't decide if it's going up or going down, whether it's ending on the dominant or the tonic or some other random note of the scale. It's undirected.

There's something else that also gives it that amateurish quality. The song's notes follow a constant rhythm, residing on basically every 1/8 beat. This too-consistent even rhythm is a bad idea combined with the wandering aimless melody, because it draws attention to the contrast between them. Same with the beat. Ironically, the drum beat does divide itself into repetitive phrases, 8 bars long. The beat picks up in bars 7-8, as if it's going to lead into something, and the crash cymbal on bar 9 resolves the tension that the drummer built up. So they have picked up on some drum theory, but not much melody theory, which again makes the contrast even more stark as the drum is trying to support a change in the melody and there's nothing there, no structure to match with it.

Melody aside, the music production is alright. The synth is cute and the instrument choice is fine for most of the song. However, in the break at 0:55, when it gets rid of the bass line, it keeps the heavy drum and the high pitched melody. I think this is a bad move. There either needs to be a middle layer to fill out the frequency spectrum, or the drum needs to go which would remove the gap. (The section at 1:35 is fine.)

Despite this, the talent of the composer is bleeding through the struggle. I can tell they have the ability to produce something amazing. They just didn't have the necessary toolkit of music theory when they composed this song. Keep an eye on this artist - after they do a bit of study, I'm sure we'll get something great from them in the future.

• Insane [web/dl] - [A] 662k - 96.06% 226x - 39 ok 3 miss
• Expert [web/dl] - [A] 469k - 92.41% 136x - 70 ok 25 miss
• Visual settings: 70% dim, hitsounds on, argon skin

Ocskai - Almom [beatmap]

Ooh, interesting! I'm not sure exactly what genre this is, but it's certainly an underrepresented one in rhythm games, which makes it especially cool not just that it's been mapped but that it was created specifically for this game. I really like hearing things like this which are totally unlike what you usually see in rhythm games.

I like this song a lot. All of the instruments sound really cool and they go well together, and a non-vocaloid male voice is unusual from an independent production. The drum track is varied and interesting, I like the little fills between the different sections, and the hitsounding on these fills is also really beautiful, definitely case study material. The shape of mapping is overall unremarkable. It feels like the kind of thing I'd download from the ranked section and play without really noticing it. This is good, though! I think the other maps maybe tried really hard to be adventurous and stand out, and went too far, which led to their downfall. See LONG LIVE THE KING's 2:20 section and the whole of unbound. They're so much, and they didn't need to be. Overall, Almom is nice.

The elements of the storyboard really are beautiful. I just can't deal with the way they move, it makes me dizzy like nothing else. I would be totally fine if it had maybe used a gentle swaying like against cacophony and REDALiCE - Pekorap Tropical House Remix, hosted by Yuuma, but instead the intro flashes and the first part scrolls horizontally at quite a fast rate. The part in the middle with the windows is more friendly, which I appreciate.

• Insane [web/dl] - [S] 995k - 99.84% 505x - 1 ok 0 miss
• Expert [web/dl] - [A] 760k - 95.94% 385x - 22 ok 2 miss
• Visual settings: 70% dim, storyboard off, hitsounds on, argon skin

Overall thoughts

Weirdly, even though I voted on the Expert diffs, I do prefer how the Insanes were mapped. This applies to pretty much every entry. I'm not sure why this is, but I can think of a hypothesis: the greater freedom and less formulaic process that comes with mapping a higher diff greatly increases the possibility space for making something weird and poor. Less use of distance snap (above Hard) makes it possible to choose bad distances. Permitted streams (above Insane) makes it possible to put streams in an ill-suited place. Things like that.

I would consider the majority of the Expert diffs I played unrankable as-is. You might be surprised to hear this - maybe you expected the map quality to be much better than the typical graveyarded map? Weren't these maps created by talented people who self-selected for this contest?

Maybe, but that's not the whole story. The ranking process isn't just about how good the mapper's initial submission is. It's a constructive, iterative process, carried out in the open, where anybody can give the mapper feedback on parts they liked and parts they didn't. Lots of people work together to make the map the best it can be, and then you have 2 experienced Beatmap Nominators come in to give it a thorough review and especially precise feedback before it can be ranked.

So although many of my statements here have been critical, this isn't really the fault of the creators. I don't want to point a negative outlook towards the people who made these specific entries. They had to work under tough time constraints with no modders. These same difficulties would be encountered by any team. In any team where this did work, it's more akin to a miracle.

Yes, I'd say the majority of these Expert diffs aren't rankable, but the bigger picture is that nothing gets ranked without modders. Contest maps are no exception. These are fixable problems. With a little outside feedback and a few iterations, all of these could bloom into excellent productions.

Now that their maps are out in the open, I hope that every team I reviewed uses this opportunity to strive for improvement, and I hope that I'll one day get to enjoy each of these after their creators consider them complete.

Thank you!

C.