Melodies

[azuk]

I've read up on music theory, and I can create chord progressions, but any melody I try to make goes shit after 2 bars, or i get tired of it and scrap it. Anyone?

[dotcurrency]

I've read up on music theory, and I can create chord progressions, but any melody I try to make goes shit after 2 bars, or i get tired of it and scrap it. Anyone?

Well first, define "goes to shit" after 2 bars?

[azuk]

I get down something decent, the nothing i try sounds good after it. Everything sound awkward and stupid.

[ChadDub]

stay in key

[Add9]

Assuming 8-bar melody:

1st two bars: Come up with an interesting melody.

2nd two bars: Repeat it but change a few of the notes or rhythms so it's slightly different.

3rd two bars: Come up with a new melody that's different from the first two, but not so different that it sounds out of place (like, keep the beginning or end of the phrase, or keep the rhythms but change the notes, something like that).

4th two bars: Base the melody off of the first two bars but make it more different than the 2nd two bars were. Changing up the end of the phrase here is nice, especially rhythmic change right at the end of the 8-bar phrase.

If it sounds to repetitive after all this, then change some of the notes or rhythms. Just keep changing notes until it sounds good and has a logical flow to it.

[titchbit]

It's kinda hard to tell someone how to make a melody over the internet. But one thing I can say is that when I took piano lessons, my teacher told me that when you make melodies, think of them as a "musical sentence" - it has a beginning, middle, and an end. And the rhythm of the melody can be thought of as punctuation (ie a comma or something for longer periods of silence). Have your melody start and end with the same note too. This creates a feeling of closure for the listener, as if the musical sentence completed properly (where it started).

[azuk]

Assuming 8-bar melody:

1st two bars: Come up with an interesting melody.

2nd two bars: Repeat it but change a few of the notes or rhythms so it's slightly different.

3rd two bars: Come up with a new melody that's different from the first two, but not so different that it sounds out of place (like, keep the beginning or end of the phrase, or keep the rhythms but change the notes, something like that).

4th two bars: Base the melody off of the first two bars but make it more different than the 2nd two bars were. Changing up the end of the phrase here is nice, especially rhythmic change right at the end of the 8-bar phrase.

If it sounds to repetitive after all this, then change some of the notes or rhythms. Just keep changing notes until it sounds good and has a logical flow to it.

Taking the first 2 bars of the melody then trying to change it seems to be why I can't make good melodies, I'll have a killer 2 bars, but any variation seems awkward, which I think is because to melodies aren't really loopable.

[titchbit]

Assuming 8-bar melody:

1st two bars: Come up with an interesting melody.

2nd two bars: Repeat it but change a few of the notes or rhythms so it's slightly different.

3rd two bars: Come up with a new melody that's different from the first two, but not so different that it sounds out of place (like, keep the beginning or end of the phrase, or keep the rhythms but change the notes, something like that).

4th two bars: Base the melody off of the first two bars but make it more different than the 2nd two bars were. Changing up the end of the phrase here is nice, especially rhythmic change right at the end of the 8-bar phrase.

If it sounds to repetitive after all this, then change some of the notes or rhythms. Just keep changing notes until it sounds good and has a logical flow to it.

Taking the first 2 bars of the melody then trying to change it seems to be why I can't make good melodies, I'll have a killer 2 bars, but any variation seems awkward, which I think is because to melodies aren't really loopable.

Make another melody and try a call & response type thing maybe.

[azuk]

Assuming 8-bar melody:

1st two bars: Come up with an interesting melody.

2nd two bars: Repeat it but change a few of the notes or rhythms so it's slightly different.

3rd two bars: Come up with a new melody that's different from the first two, but not so different that it sounds out of place (like, keep the beginning or end of the phrase, or keep the rhythms but change the notes, something like that).

4th two bars: Base the melody off of the first two bars but make it more different than the 2nd two bars were. Changing up the end of the phrase here is nice, especially rhythmic change right at the end of the 8-bar phrase.

If it sounds to repetitive after all this, then change some of the notes or rhythms. Just keep changing notes until it sounds good and has a logical flow to it.

Taking the first 2 bars of the melody then trying to change it seems to be why I can't make good melodies, I'll have a killer 2 bars, but any variation seems awkward, which I think is because to melodies aren't really loopable.

Make another melody and try a call & response type thing maybe.

I'll try this, call & response seems like it should work.

[orlandobloominonion]

I'm pretty shit at it too. All my melodies are just way too simple. I'm a bassist at heart....

What dubuncked said fits into what Add9 said. Call and response is very similar to a "musical sentence." It's about developing small tension and release throughout the melody.

[firstboyonthemoon]

You can work variation from the other direction, too: change the harmony underneath, but leave the melody the same. For instance, a two-bar phrase over an Am-Dm progression will probably also work over an F chord. A common real-world example is 12-bar blues, where the first phrase is played over the I chord, then repeated over a IV-I progression. (And then finally answered by a phrase over a V-IV-I progression.)

[SkuxxOfficial]

Try make some chords and then some notes using the same keys as the chord, 1or 2 octaves higher.
if your using massive, a squ-swl with the wt-position right in the middle, works for me.

[SunkLo]

Try to extract some motifs from what you already have and then apply them to other ideas. Write a response to your first melody line that uses the same rhythm but different notes or vice versa. Think about the contour of your melodies, where are they stepwise, where do they leap? Where are notes ascending vs descending? Where is tension created and released? If you look at the contour like a function, you can take the derivative and then apply that to other lines. So if you've got a jump of a third between the first and second note of melody 1, you can add in the same jump to your second line, but starting on a different note. This establishes a pattern between individual melody snippets and makes things sound more cohesive.

If you're having problems generating any melodies at all, set up the harmonic structure first. Once you have chords defined you have a hierarchy of notes to be played. Your melody will probably favour the core chord tones more than the extensions, but sneaking in little chromatic guide tones can result in a smoother contour. If you're trying to do this on a grid you're in a rough spot. But if you've got a midi keyboard, you can just play around with different voicings of the chords until you hear some sort of motion that you like. You can then isolate that, prioritize it as a melody (most likely by putting it into an upper voice or transplanting it to a lead instrument), and then develop and embellish it.

Also if you've already got a backing track lined up, take a bandpass filter and sweep it around. If you listen you'll hear little snippets of what could be a proper melody. Or you could put a delay on your keys and fiddle around with delay times while jamming until you get some compelling combination of rhythm and harmony, and use that to build off of. I find setting up an environment where you can play and audition new ideas very quickly is much more useful than trying to pre-meditate everything.

Trying to intellectually brute force your way into something good usually ends up being very frustrating. You need to discover something good. Once you've got some meat to your composition, you can draw some tangents off your existing material to create variation. The seed of of the composition process tends to be quite visceral, whereas the refinement process afterwards is when you can get cerebral about things.



As an aside, this is why it's important to know your software and hardware intuitively. If you're trying to parse your way through a bunch of technical garbage while in the heat of the moment, that creative spark is going to get choked out by the dominance of left brain thinking. You want to relegate that analytical portion of your brain to background tasks while your creative right brain thinking takes precedence.

[titchbit]

Also, as a last resort, if you are 100% stuck, a (cheap) way to make melodies is to take an arpeggiator/step sequencer preset in your synth of choice, and then just play a couple notes. If you just input a single note, the arpeggiator/step formula will spit a melody back at you. Then you would have to record the MIDI somehow, or you could record the audio if you're using Ableton (not sure about other DAWs), use the auido to MIDI feature to bring it back into MIDI, and then take the new MIDI clip and apply it whatever instrument patch you wanted to use in the first place.

I'm not saying this is the way you should make melodies, but it can work. This is a last ditch effort.

[__________]

Just jam out you pussy

[Soulstep]

http://www.8notes.com/resources/notefin ... chords.asp

http://www.scales-chords.com/scalefinder.php

[azuk]

http://www.8notes.com/resources/notefin ... chords.asp

http://www.scales-chords.com/scalefinder.php

Not asking for chords.

[azuk]

Try to extract some motifs from what you already have and then apply them to other ideas. Write a response to your first melody line that uses the same rhythm but different notes or vice versa. Think about the contour of your melodies, where are they stepwise, where do they leap? Where are notes ascending vs descending? Where is tension created and released? If you look at the contour like a function, you can take the derivative and then apply that to other lines. So if you've got a jump of a third between the first and second note of melody 1, you can add in the same jump to your second line, but starting on a different note. This establishes a pattern between individual melody snippets and makes things sound more cohesive.

If you're having problems generating any melodies at all, set up the harmonic structure first. Once you have chords defined you have a hierarchy of notes to be played. Your melody will probably favour the core chord tones more than the extensions, but sneaking in little chromatic guide tones can result in a smoother contour. If you're trying to do this on a grid you're in a rough spot. But if you've got a midi keyboard, you can just play around with different voicings of the chords until you hear some sort of motion that you like. You can then isolate that, prioritize it as a melody (most likely by putting it into an upper voice or transplanting it to a lead instrument), and then develop and embellish it.

Also if you've already got a backing track lined up, take a bandpass filter and sweep it around. If you listen you'll hear little snippets of what could be a proper melody. Or you could put a delay on your keys and fiddle around with delay times while jamming until you get some compelling combination of rhythm and harmony, and use that to build off of. I find setting up an environment where you can play and audition new ideas very quickly is much more useful than trying to pre-meditate everything.

Trying to intellectually brute force your way into something good usually ends up being very frustrating. You need to discover something good. Once you've got some meat to your composition, you can draw some tangents off your existing material to create variation. The seed of of the composition process tends to be quite visceral, whereas the refinement process afterwards is when you can get cerebral about things.



As an aside, this is why it's important to know your software and hardware intuitively. If you're trying to parse your way through a bunch of technical garbage while in the heat of the moment, that creative spark is going to get choked out by the dominance of left brain thinking. You want to relegate that analytical portion of your brain to background tasks while your creative right brain thinking takes precedence.

Damn,

[WolfCryOfficial]

Hey Azuk, i noticed you were having trouble with melodies (setting aside our growl bass differences.)
Here is an example melody of something i've made.
If the melody itself doesn't interest you, but the method does, let me know. I took piano lessons for 15 years and theory lessons for 12 years. Melodies are my strong point of music.
Let me know if i can help ya out.

Specifically when the guitar comes in and the mixdown before the drop.
Soundcloud

[azuk]

Hey Azuk, i noticed you were having trouble with melodies (setting aside our growl bass differences.)
Here is an example melody of something i've made.
If the melody itself doesn't interest you, but the method does, let me know. I took piano lessons for 15 years and theory lessons for 12 years. Melodies are my strong point of music.
Let me know if i can help ya out.

Specifically when the guitar comes in and the mixdown before the drop.
Soundcloud

Rofl self-advertising, but ehh, the melody isn't that interesting,.