11-25-2023, 11:17 AM
For this guide, headphones == speakers
The four things to consider when mixing/writing your OS song are
Frequency response
Stereo width
Dynamic range
Focal point
Frequency response is the different general gains (gain is a more accurate word to describe volume) of every frequency your ears can hear. This sounds complicated but don't worry, it basically means you can think of your song as high frequencies, mid frewuencies, and low frequencies. Ideally your song, in the most intense parts, should have something playing in all three ranges. That something can be as complicated as an orchestral arrangement or as simple as a single hihat. As long as there is something in each you should have a good intense part.
To take it a step further, lets divide it up more, and add sub frequencies below low frequncies, low mids between mid frequencied and low frequencies, high mids between mids and highs, and air frequencies above high frequencies, like this!
Air - sparkly sounds like the top of a hihat, other cymbals, "Ssss" sounds, etc. These give your song a sense of openness, and make it feel more airy I guess
Highs - sounds like the crack of the snare drum, the noisy part of a distorted guitar, etc. These give your song energy
High mids - This is where the voices are most clear, where you'll hear flutes, koto, stuff like that. This is where the ears are most sensitive so its important
Mids - The body of the song. This is where most instruments will play, like piano, guitar, loeer voices, harp, I could go on.
Low mids - This is where your song gets power from. Think of distorted guitars, the punch of a snare drum, etc.
Lows - this is where your bass and toms will sit. It kind of glues the song together and makes everything else sound like it has more foundation.
Sub - the lowest sounds you can hear. These give weight to the song like the thump of a kick drum, or sine with detune all the way down on the lowest notes of OS
You can change what ranges are playing throughout the song in order to build the structure of your song! Some instruments can take up multiple ranges also. If you think a part you added sounds lacking try adding things in other ranges and most of the time you'll realize one range was missing. Be unafraid to delete stuff if it doesn't work, like if you have a part thats too intense and doesn't flow well with the previous part of your song, try removing it and see.
Good speakers or good headphones are essential to doing this well. Bad ones can't play back all the frequencies at the correct gains so you make adjustments that might sound good only on those speakers but nothing else! Be cautious when boosting sub and low frequencies volume on bad speakers/headphones especially.
Stereo width is how different the sound coming out of each speaker is. Speakers are funny because when they both play the same thing, the music sounds small and right in front of you, but if they play different things, its like you unfolded the music from a tiny strip into an entire poster. Kind of hard to describe actually sorry. Think of thin instruments on OS as Electric Guitar, Acoustic Guitar, Koto. They sound in a way smaller, or thinner, than stuff like grand piano, smooth synth, and distorted sustain guitar. We people typically prefer wide sounding music, music that sounds like a really big thing rather than a small thing. Its really hard to describe, I'm sorry.
Basically you want to maximize this stereo width by moving different things into each speaker. You can use panning to do this on OS! To get started, it doesn't matter what direction you pan stuff, as long as you can balance it out on the other side with something else. Instruments that are thin sounding can usually be panned a lot further than wide instruments while still sounding good. You can also pan one melody played on two similar sounds all the way left and right to make the impression of a wide instrument. Guitarists do this all the time.
You should probably avoid panning the most important parts of your song though. Kick, snare, whatever you are using as a bass, and your main melody should not be panned around or else the song will sound unbalanced. Try it and see what happens, to get an idea of how it sounds bad.
Since those things are not good with panning, you should pan all the background stuff around so that you still get the width to your song while having it sound good too!
Be careful with panning too many things in one direction without balancing it out on the other side with things inside the same frequency ranges or else you may find the song feels uncomfortable to listen to.
Dynamic range is the range of gain levels from quiet to loud basically. Usually you want to design the song so that there is as little dynamic range as possible, or else the song will sound messy and weak rather than strong and powerful.
"But doesn't this remove all the realism of an instrument?" You might ask? Consider this: a vocal is an instrument with a HUGE dynamic range, and in a song, a singer might go from almost whisperimg to yelling as loud as they can. This is a problem because when they whisper it can be hard to hear them over the rest of the song, but when they yell, it can be hard to hear anything else. Whats even worse, is that sometimes the gain changes drastically even within a single word! So the solution is to limit that dynamic range as much as possible, so that everything they sing can be easily heard. This is what every song does. The singer still sounds great, and you probably didn't even realize this was happening! The thing with real instruments is that the change in TIMBRE as they are played harder is much more important than the change in volume, and that is what makes them sound realistic!
The tools you'd typically use for dynamic range also affect the gain of each note so say a bass guitar note is falling in volume rapidly, then you can use a compressor to make it stay at the same volume for longer
So, on OS we don't really have any dynamic range tools, what can we do to minimize the dynamic range? We need to get creative with instrument layering, clones, detune, and distortion! On drums lets say, you can add distortion to a drum instrument to lower the dynamic range, this will also crush the punchy attack of the drums though so maybe layer that with the undistorted drums for a larger drum sound. You may also want to add a reverb to that distorted drum so the volume lasts longer giving you less dynamic rsnge. 8bit instruments already have no dynamic range so they are great for a lot of things. Higher frequency stuff can still sound good with more dynamic range and it isn't as impirtant to minimize that stuff, but for lows, sub, and low mids, even mids, it can make your song feel much more powerful!
The focal points of your song are the only things your average listener (not talking about OS users, just imagine some guy who doesn't think about music) is going to notice when listening to the song. They might only be the main melody or bassline, could also be drums. Its unlikely to be anything else. You want to write and mix your song so that those things are SUPER EASY to focus on, and the rest of the stuff compliments that rather than competes for the listener's attention. Things to avoid are two similar melodies playing at once, especially in the same frewuency range, multiple basslines playing at once, and don't put much into the same frequency range as the main melody. You want each element of your song to be clear also, unclear parts are major attention grabbers even if the listener doesn't know why. If you notice a part of the song has kind of like a muffled noise sound to it, try to minimize that.
Try to hear these concepts being applied in your favorite songs too and it can help you get an idea about how and when to apply this information if you are unsure!
The four things to consider when mixing/writing your OS song are
Frequency response
Stereo width
Dynamic range
Focal point
Frequency response is the different general gains (gain is a more accurate word to describe volume) of every frequency your ears can hear. This sounds complicated but don't worry, it basically means you can think of your song as high frequencies, mid frewuencies, and low frequencies. Ideally your song, in the most intense parts, should have something playing in all three ranges. That something can be as complicated as an orchestral arrangement or as simple as a single hihat. As long as there is something in each you should have a good intense part.
To take it a step further, lets divide it up more, and add sub frequencies below low frequncies, low mids between mid frequencied and low frequencies, high mids between mids and highs, and air frequencies above high frequencies, like this!
Air - sparkly sounds like the top of a hihat, other cymbals, "Ssss" sounds, etc. These give your song a sense of openness, and make it feel more airy I guess
Highs - sounds like the crack of the snare drum, the noisy part of a distorted guitar, etc. These give your song energy
High mids - This is where the voices are most clear, where you'll hear flutes, koto, stuff like that. This is where the ears are most sensitive so its important
Mids - The body of the song. This is where most instruments will play, like piano, guitar, loeer voices, harp, I could go on.
Low mids - This is where your song gets power from. Think of distorted guitars, the punch of a snare drum, etc.
Lows - this is where your bass and toms will sit. It kind of glues the song together and makes everything else sound like it has more foundation.
Sub - the lowest sounds you can hear. These give weight to the song like the thump of a kick drum, or sine with detune all the way down on the lowest notes of OS
You can change what ranges are playing throughout the song in order to build the structure of your song! Some instruments can take up multiple ranges also. If you think a part you added sounds lacking try adding things in other ranges and most of the time you'll realize one range was missing. Be unafraid to delete stuff if it doesn't work, like if you have a part thats too intense and doesn't flow well with the previous part of your song, try removing it and see.
Good speakers or good headphones are essential to doing this well. Bad ones can't play back all the frequencies at the correct gains so you make adjustments that might sound good only on those speakers but nothing else! Be cautious when boosting sub and low frequencies volume on bad speakers/headphones especially.
Stereo width is how different the sound coming out of each speaker is. Speakers are funny because when they both play the same thing, the music sounds small and right in front of you, but if they play different things, its like you unfolded the music from a tiny strip into an entire poster. Kind of hard to describe actually sorry. Think of thin instruments on OS as Electric Guitar, Acoustic Guitar, Koto. They sound in a way smaller, or thinner, than stuff like grand piano, smooth synth, and distorted sustain guitar. We people typically prefer wide sounding music, music that sounds like a really big thing rather than a small thing. Its really hard to describe, I'm sorry.
Basically you want to maximize this stereo width by moving different things into each speaker. You can use panning to do this on OS! To get started, it doesn't matter what direction you pan stuff, as long as you can balance it out on the other side with something else. Instruments that are thin sounding can usually be panned a lot further than wide instruments while still sounding good. You can also pan one melody played on two similar sounds all the way left and right to make the impression of a wide instrument. Guitarists do this all the time.
You should probably avoid panning the most important parts of your song though. Kick, snare, whatever you are using as a bass, and your main melody should not be panned around or else the song will sound unbalanced. Try it and see what happens, to get an idea of how it sounds bad.
Since those things are not good with panning, you should pan all the background stuff around so that you still get the width to your song while having it sound good too!
Be careful with panning too many things in one direction without balancing it out on the other side with things inside the same frequency ranges or else you may find the song feels uncomfortable to listen to.
Dynamic range is the range of gain levels from quiet to loud basically. Usually you want to design the song so that there is as little dynamic range as possible, or else the song will sound messy and weak rather than strong and powerful.
"But doesn't this remove all the realism of an instrument?" You might ask? Consider this: a vocal is an instrument with a HUGE dynamic range, and in a song, a singer might go from almost whisperimg to yelling as loud as they can. This is a problem because when they whisper it can be hard to hear them over the rest of the song, but when they yell, it can be hard to hear anything else. Whats even worse, is that sometimes the gain changes drastically even within a single word! So the solution is to limit that dynamic range as much as possible, so that everything they sing can be easily heard. This is what every song does. The singer still sounds great, and you probably didn't even realize this was happening! The thing with real instruments is that the change in TIMBRE as they are played harder is much more important than the change in volume, and that is what makes them sound realistic!
The tools you'd typically use for dynamic range also affect the gain of each note so say a bass guitar note is falling in volume rapidly, then you can use a compressor to make it stay at the same volume for longer
So, on OS we don't really have any dynamic range tools, what can we do to minimize the dynamic range? We need to get creative with instrument layering, clones, detune, and distortion! On drums lets say, you can add distortion to a drum instrument to lower the dynamic range, this will also crush the punchy attack of the drums though so maybe layer that with the undistorted drums for a larger drum sound. You may also want to add a reverb to that distorted drum so the volume lasts longer giving you less dynamic rsnge. 8bit instruments already have no dynamic range so they are great for a lot of things. Higher frequency stuff can still sound good with more dynamic range and it isn't as impirtant to minimize that stuff, but for lows, sub, and low mids, even mids, it can make your song feel much more powerful!
The focal points of your song are the only things your average listener (not talking about OS users, just imagine some guy who doesn't think about music) is going to notice when listening to the song. They might only be the main melody or bassline, could also be drums. Its unlikely to be anything else. You want to write and mix your song so that those things are SUPER EASY to focus on, and the rest of the stuff compliments that rather than competes for the listener's attention. Things to avoid are two similar melodies playing at once, especially in the same frewuency range, multiple basslines playing at once, and don't put much into the same frequency range as the main melody. You want each element of your song to be clear also, unclear parts are major attention grabbers even if the listener doesn't know why. If you notice a part of the song has kind of like a muffled noise sound to it, try to minimize that.
Try to hear these concepts being applied in your favorite songs too and it can help you get an idea about how and when to apply this information if you are unsure!
