I was talking with faust last night about note placement--how you can tell notes and patterns that "fit" the song. I am completely incapable of this task, though there are a lot of very good mappers out there who really know what they're doing. However, I can't seem to grasp the concept due to my heavy emphasis on logic and rhythm.

What I want is a collection of ideas on note placement. Obviously rhythm plays a part, but not entirely. What is the thought process between placing a note (creating the rhythm) and then putting it in a particular spot on the map? Why did you do "x"? I'm hoping this will be similar to my rhythm guide I posted earlier in that I explained how to avoid awkward rhythms, but I'm afraid it might just be way too vague.

My current method doesn't really amount to much: I set a particular distance snap and try to match some sort of a pattern to the rhythm I've made. I make sure that any part that I think a jump would go, I put a jump there. I avoid things I hate (like anti-jumps and hiding notes).

A map with examples would be awesome too.

The main problem is that there is not one correct way, but there are many ways that are accepted, and many ways that aren't generally accepted. For example, regular pentagons and stars are much more attractive than poorly spaced ones.

Also, as another rule of thumb: if you're going to start a pattern (of circles, sliders, kickbacks, etc.) based on the music, continue that pattern, or don't start using that pattern at all. One thing I dislike seeing is the discontinuation of patterns even though the music doesn't suggest otherwise.

On that note, the list in your collection of ideas on note placement will never be exhausted, but if you get enough opinions, you may get a lengthy list.

Art =/= Science.

That's about it pretty much, if you try to apply formulas to your mapping it wont end up good, just feel the music and map it.

Here's a trick that generally worked for me.

Just close your eyes and listen to the song for a while, and imagine as if you're not EDITING it, but playing it, like someone else's map. Maybe an official EBA/Ouendan map. Once you get an idea in your dream what it might look like, start putting it down on paper (or in this case, the editor). TEST CONSTANTLY (use f5). This is the trick to mapping. If a part doesn't feel right, experiment until you are happy with it.

My general rule is this: An edgy or dramatic song should have a lot of edgy and dramatic patterns, with lots of directional changes. The Jarby way of making patterns is to use straight lines for emphasis, while otherwise making patterns curve/shake. I find that this works quite well. Interlocking/honeycomb patterns are useful for intense or intricate sections of a song.

Ideally, patterns should move more or less in opposite directions, as this prevents the boring sensation of going around in circles. Making patterns move up and down with the melody is also a sure way of inflecting the music, while mapping sudden changes to low voices/percussion further down the screen creates a heavy feel.

The only way to assure that your patterns aren't boring is to decide whether or not you enjoy playing them. If they're dull, then change them into different shapes and play around with the occasional honeycomb. Alternatively, play EBA and Ouendan, because their patterns are more or less perfect. The thing to remember is that those games are based on dancing, so there is usually a definite side to side motion.

Really, if you use long sliders well enough, then you can worry a lot less about circle patterns. A lot of the time, placing a few nice sliders properly means that patterns can be worked around them, thus making the map a lot more pretty.

Also, high slider velocities and 1x spacing makes it much easier to make patterns feel like they're going somewhere, simply because they cover more of the screen. I encourage everybody to use this anyway, because it makes transitions between sliders and circles a lot more fluent.

I like to put stacks on unisons.

So other dream patterns like me .

Anyway, a more general way is about how easily you can feel it while playing: it work only if you can play it, and is useful to test even if some undermapping/overmappiung is intuitive. (if you feel something is strange...than it's strange).

Differently, you can watch your map movements with your mind: imaging how patterns go and so. Useful when you can't play a difficulty, or you are to map something harder than your possibilities.

I can't give an utter definition, I kinda agree with Sakura's words

Mapping is somehow abstract, not like SB stuff or custom skin dimension that can be quantified under no matter what situation

It's more rely on sense, or even inspiration.

My opinion does not help practically, but I believe in it :3

It's all about translating sound into visual art and make it the best playable map as possible.

The easiest way for you to learn how maps are made to music, is to play other maps made by other players. So you don't have to waste trying out everything yourself. Which is why most newer maps are better than older ones. They lack experience and inspiration. The more of these things, the better maps you usually make.
If you're visually presenting your map in a appealing way and make sure the excitement and feel in music adapts to the beatmap, you will get good rating and people will like your mapping. However you might get less recognising and feedback if the music (which isn't yours) is presented to them in a way they don't like. English songs usually get bad ratings due to this, beatmapping goes both ways in beatmap ratings. Don't forget you can't always blame it on the song.

While mapping, you should always try out every single thing you've mapped throughout the map. If you're experienced enough to be able to play your difficulty perfectly, you may find it much easier to notice flaws and patterns that doesn't play as good as they should.

You can start by dividing the song in different parts by different slider velocities. The slider velocity will determine the overall speed and feel of the beatmap while the song remains unchanged.
Now the slider velocity can always be different in any way, but you have to adapt it to how the music is going to sound. If you expect a lot of long sliders, try using a slower SV since making long sliders are hard to map while at the same time making sure they play good. And vice versa.

You also need to consider the overall feel of the song, is it powerful or not, does the song give out an illusion of faster BPM, etc. So that has to be adapted with spacing and SVs.

When should you use jumps? Jumps are not common in beatmaps since they should only be placed where they are needed. So if we translate that into music, we get something that is sudden and not very common in the song. A sudden but powerful sound in the music? Yes, this is where you should add a jump in the beatmap. Using more jumps than necessary can be tiring and not fun to play.

Sliders should most definitely be used where the sound is stretched out in any way. This can be quite hard to grasp when you're new to mapping but if you're experienced in both mapping and playing, you should automatically expect some notes to be stretched out than others.

Notes are simply put out where short musical notes are heard in the song. The only time a slider is more acceptable to use is when the current sound is less common to hear than the other notes. For example on claps and other hitsounds that can be added in the beatmap.

Note placement:
You want to try to make it as visually good looking as possible while still keeping it fun to play.
So what is beautiful? Well if you've noticed, everything that has to do with geometry in mapping is beautiful. Together with other patterns and notes, whole parts and sections of a map will be beautiful if it's connected geometrically in some sort of way. It's kind of like drawing, you move around with your cursor (with pen or mouse or whatever you're using) and draw things with z and x. But what you draw is notes you're pressing to match music. So what feels good to draw/play? Geometry.
You can use geometry by making multiple non-geometrically shapes, creating an overall geometrical feel. It's all a matter how you analyse things. Being extremely analytic is how you notice details and flaws. That's what you want to be.

The more sudden sudden changes in a map, the more excitement you get out. Using the same parts that you've used previously in the same map can be less exciting than remapping that part. Even though the music sounds exactly the same. You should keep the rhythm the same (or at least similar) but with different note placement. Same distance with jumps and all, but differently placed around in the beatmap.

Problems:

Players tend to ignore what actually is right for the beatmap and focus on what's fun or not. When they then try to map, the result can be quite overmapped and feel to extreme for the actual song.
This can also go the other way around where it's undermapped. Too long and boring breaks which makes it less exciting and so on. If you ask me, only add breaks when you can really hear a calm or/but ending new part after a quite long previous one.

You have to try to balance the beatmap by making it both fun to play and fit to the song. This should mostly be done by following the basic mapping criteria and the simple way of mapping a beatmap explained by me earlier.

Since you can add hitsounds to osu!, you have to make sure they are good too. Now compared to notes and all that shit, hitsounds are sound, just like the music is. From experience and statistics, if the hitsound is placed so that they matches the music, more positive feedback will be received. Are you or the artist better at making sound patterns? You have clearly less exciting sounds to deal with than the artist and the safest way to go is for you is to make the hitsounds after the music.
After time, hitsounds and everything else in a beatmap have become more complex. This is where we're heading. Beatmaps sound and look more like you would imagine in the song. This is how you go about making a beatmap basically.

How do you know if something fits in a map or not? Look up where it fits, see how it is used. If the song matches yours, you should be free to do the same but if not, try using something that would match your song better. Analyse what play good already and see how it would do in your map.