Articles 31-40 in the Musical Creativity Series
Mixing rhythms can produce effective results.
Different time signatures
Think of the stress pattern of a 4/4 bar. It will have the main stress on the first beat of the bar. So that's:
1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4
Now let's look at a 5/4 signature, the stress would again be on the first beat
That's 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5
If we keep them at the same speed, same beats-per-minutes, then the 5/4 bar is 25% longer than the 4/4 bar due to additional crotchet.
Now for a 3/4 signature, it would be 1 2 3 1 2 3 1 2 3 1 2 3 and only be 75% the length of the 4/4 bar due to the missing crotchet.
Adding them together
The stress patterns combine at different points in the song. I'll just use patterns with the accent on 1 for this:
1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4
1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2
1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1
The first we see is where the 1 of 3/4 and the 1 of 4/4 align to result in a heavier accent. Then it happens again with 3/4 and 5/4. And so on. It's like a musical version of the game fizz-buzz. After 15 bars, the process starts again.
Edit - I'm afraid the full repetition wouldn't fit on the page, so I've had to remove it
That's for a very simple process where the accent is on the first beat of the bar for each time signature. Usually, we'd throw in other accents to achieve a rhythm.
I've taken a simple concept of a 3 track piece with very repetitive rhythms to show this. Each channel features a note from the chord of Am (A, C and E) using soft synths triggered by midi pattern.
How did I do this?
Order
Have you thought about how you compose? Not which chords, but the order in which you approach composing?
My experience
I use a few different workflows, depending on what I'm aiming to produce. I change my approach if there's a video to sync to or if I need to incorporate other musicians and so on. I'm not saying that any one of the approaches is the correct approach for everyone, but I do what works for me.
Write first or play first
The main difference is in how I start. For some I
There are advantages and disadvantages to all of them. If I'm being honest, I use a mixture of all of them in almost every composition. It's how I start that counts.
Scenarios
Here are a few of the different scenarios I find myself in:
Something like sync to video, I'll play first while watching the video. Sometimes what I play will make me think further and I'll switch to composing mentally before recording that part.
Why the difference?
I get different results depending on how I set out. With that in mind, I force myself to swap how I start to ensure that I get a variety of results. The sound quality is of an equal level with consistent production values no matter how I start, but it sounds different. Some approaches require an additional step for re-recording, e.g. if I've played first, then want to record it better.
Suggestions
Try starting from a different point.
I'll go into more detail about my normal workflows in a later article
I wrote the bulk of this text a couple of months ago after a day of not producing as much as I'd liked to. There are no real suggestions of how to improve or learn, but I thought I'd describe my thought processes so others may learn from my experiences.
Yesterday
I had one of those days yesterday where inspiration was lacking. Partly it may have been the brief I'd been given since it didn't have the depth of information I'd have liked. Mainly, though I think it was just me, I couldn't really find much to do, not that sounded right anyway. But I still managed to deliver something of sufficient standard.
How?
I tried creating music for the part from scratch. I wrote some interesting ideas. But they didn't really fit the brief as far as I was concerned. They're now exported to mp3 and the relevant project files archived for future reference. You never know what may be useful later on - I revisit unfinished files on a regular basis and often manage to complete them another time.
I then reviewed my collection of part-finished works hoping there was something nearly finished in there. Unfortunately there wasn't anything near finished. So I took two that really needed a lot of work and moved them forwards. One was almost finished from a composition point of view, but required a lot of mixing and processing, the other required more composition and almost no mixing.
Both featured a repeating rhythmic phrase and some background drums; one had a few bits of melody as well. The originals were rough versions of ideas that I thought worth keeping but didn't have the inspiration to finish when I first recorded them. They were a mixture of midi-triggered samples and audio recordings.
First Track
The first track was a sad and slow bluesy-jazz and had the following tracks:
For this one, it was mainly a case of arranging and mixing rather than composing. The original idea was over 3 minutes, the brief was for 1 min 50 seconds. So I worked at taking the best bits and arranging them into something musically attractive.
Before Arranging
Before I could arrange the work, I had to sort out the audio. It just didn't sound right. The original was recorded in one take per track, i.e. place some drums out, record the rhythm guitar in one go, then record the bass, then the lead guitar. Any of the audio processing was done very roughly just to convey the idea.
Rather than tweak and tweak and tweak some more, I removed all the processing and started again. After having listened to the track a few times, I knew where I wanted to go with it.
Clear some space
First of all I had to clear some space, the instruments were overlapping way too much. This involved some careful eq mainly focussed around making room for the bass guitar to be heard. Rolling off the low-end on the guitars and nudging the drums down in the same area helped bring the bass through considerably.
Guitar Tone and Amp Placement
I actually liked the tone I'd originally recorded with but it didn't fit in my vision of the mix. I changed the speaker emulation for a more familiar British 2x12 and turned the gain down to reduce the distortion. I also removed the reverb from the plug-in so I could add a different one later to fit in with the rest of the instruments. I'd considered re-recording with guitar amp and microphones but decided against this since the tone was sufficient for the background rhythm guitar.
Delay for lead
The lead guitar sings and needs to be heard above and beyond its accompaniment. I used a stereo delay on a aux bus for this and set the send so that it was audible. I did think about backing it off as I would do for a reverb, but I liked the extra character the delay gave the guitar.
Compression for drums
The drum loops were from BetaMonkey. They're good quality loops and are usually very dry or, in the case of the early ones, in a good-sounding room. In this particular song, the drums didn't cut through as I'd like. I could hear the snare too loud in the mix compared to the other drums. So I used a multiband compression on the drums with the gain mainly in the lower two bands and stepped down in the higher two respectively. Although setting the compressor up in this way is probably the opposite of what I'm more used to (i.e. heavier treble, lighter bass), this changed the character nicely.
Reverb for drums
To make the snare and cymbals live more in the mix, I used a reverb on an aux channel with a high-end pass. This added a nice, but barely audible quality to the drum track. The compression used in the previous step had brought the other drums up to a good level so this action balanced that out.
Mixbuss
Although nicely balanced, the instruments almost sounded like they were recorded in different rooms and the overall recording was slightly flat. A small eq alteration helped. I added the slightest amount of reverb to bond it together. I'm usually happy with a convolution reverb of a nice-sounding studio to get the feel for a track and then choose from there when I've got an idea of where I want it to be. For this particular track, I used a jazz club reverb for a more live sound. It fitted the bluesy-jazz feel of the whole track.
The final element in the chain was a limiter. I usually have two exports, one with the compressor in the chain and one with out. Then I've got the flexibility to hear what it would sound like compressed like many other tracks but still give clients the more dynamically accurate version without the compression.
The second track
This was a lot simpler in scope, but I spent a lot more time on it due to having to compose more parts and phrases to glue it together. The original idea was 50 seconds long and consisted of a midi-triggered samples of a piano, acoustic double-bass and an acoustic drum-kit. That's not that long.
Timing
I'd originally played all the parts on a midi keyboard. So the first step was to ensure all the notes were on time. This was a combination of quantising and manually adjusting the placement of notes. Some of the notes were also wrong, just not fitting in with the key. I manually adjusted these as well, rather than relying on any logical processing.
Drums
Oddly for me, the drums were samples rather than loops. I tend to find that loops fit in a lot better than any samples ever can. For this song I'd only used 4 drum sounds. I'd envisaged a drummer with an older style kick, snare and hihat, rather than the mammoth rock kits of the 80s. This simplicity, specifically the lack of cymbals, helped make the sound more realistic.
To add further realism, I patched in a light drum room reverb, just light enough to hear.
Piano
The piano sample sounded too artificial when solo'ed. I auditioned a few other piano samples, including some that were sampled from much better pianos and with much better recording. Yet I decided on the original but with two major changes:
Structure
The original was 50 seconds long. I needed an extra minute. By thinking through the structure I wanted to introduce, I was able to create the 1 minute 50 required. It was just on the edge of having to introduce a different chord progression, theme or motif. Any longer and I would have had to. As it is, the different instruments coming in and out make it interesting enough for the short time. I looped the piano, bass and drum parts. Cutting out a few loops to bring attention to other instruments. I then added the guitar over the top and then the jazz organ.
Guitar
The guitar was troublesome. I knew it needed another instrument or two to keep it going to 1 minutes 50 seconds, but every time I played along to the song, I was left short of ideas and just had a tangled mess. I also sat down with the chords and tried working out a melody. Again nothing worked. It was accurate harmonically, just not interesting. I tried changing guitar sounds a few times - I find the inspiration from that can help a lot, even from a totally inappropriate guitar tone. In the end, I switched on the cycle function and hit record. About 45 minutes, later I had enough pieces I could use. These were pasted into the arrangement and unused pieces deleted.
The tone was deliberately kept mellow.
Jazz Organ
Again another instrument to fill in the gap from the guitar and bring the listener out of the emptier phrase in which it starts. I played this, then quantised/shifted notes manually. I was careful to use the correct quantisation for each phrase (mainly 16ths and 24ths, i.e, triplet 16ths).
Finally
There was almost no eq of the individual tracks, they sat well enough together without much alteration. I added a simple jazz reverb on the mix buss and compression similar to the first track. That worked well.
Results
I'm happy that I finished two pieces both of the required length meeting the brief.
Was it a good day though?
Probably not.
I wouldn't call either piece usual for me. That's not necessarily a bad thing. It wasn't so much that it was a different type of music to what I'm used to (they're not, I'm quite happy with that style). Instead it was more than I felt uncomfortable with the end results. They're good enough for background music which is what they're intended to be, but don't have the same melodic focus I'd usually want.
I was nowhere near as creative in the whole day as I was in the first hour of the following day. On that day, I wrote three good ideas in less than an hour. I spent an hour tidying up one of the tracks, bear in mind it was 7 minutes long. Like the above tracks, they'll take longer to finish, but that can wait for another day when I need to progress them further.
Part of a series by Award Sounds offering a selection of creative ideas to kick-start or rejuvenate a composition.
As well as composing and recording, I play guitar in a band, playing a mixture of covers and our own originals. We don't stick too closely to the originals when we cover them, realising that we only have four members (guitar, bass, drums and vocals) so many songs have to be stripped down to work.
Going from a song that has many parts to a cover version with only four instruments can leave you feeling a bit naked when playing live. It doesn't help that as guitarist, I often have to play several parts at the same time.
It's just arranging
You've got 6 strings. So if you're adding in a piano part and a guitar part, then it's time to pick the most important notes. After all, you can only play a maximum of 6 at a time.
Listen
Listen to the music and pick out what makes the tune. A lot of notes are filler. The music sounds different without it, but that's ok.
Listen to other listeners
Not only listen to the track yourself, but ask the other band members what elements of the song stick out to them. Singers pick out different notes to bass players and so on. So there's me as a guitarist saying that some notes are missing from our arrangement, but no-one else has noticed. And it happens the other way around, where one of the other members will ask where a certain part is.
Who's the arranger?
If you are, then don't be afraid to ask the other musicians to try playing parts that they wouldn't have considered. I'm fortunate to be in a band with an accomplished and adventurous bassist. That's good because it allows us to swap when it comes to supporting the melody or the rhythm. The drummer's also open to new ideas. They, in turn, ask me to play other parts I hadn't considered.
The tone of the guitar
For notes played at the same time (e.g. chords), we've a limited tonal range. Each note in the chord has to be nearly the same tone and volume. There's some flexibility if you pluck with your fingers, but it's still pretty restricted.
If you look across the length of the song, then guitarists have access to a massive range of tone since we can vary it as we go along. The trick there is to tie the tone to a musical part, e.g. the piano lead.
Two at once
The guitarist can support the rhythm by striking staccato chords, but then there's an empty space for pads to ring. My way around this is to play the ringing chord and then play the rhythm on the low 6th string, sometimes with the 5th depending on what the chords are and how much I need to reinforce the rhythm. This works well clean or crunch sounds, not quite as well with hi-gain tone, but it can still work. Just depends on the chords being played and the rhythm that it needs to fit into.
Swapping between instruments on the same tone
To differentiate between two instrument parts, I've used upstrokes on one and downstrokes on another. Swapping between primarily plucked and primarily legato/glissando can help differentiate.
Try alternate strokes.
Similar to the ringing chords plus staccato rhythm above, the root notes can be played on the downstroke, then the ringing chords on the upstroke. Even more useful is the downstroke on the root, then several staccato chords using upstrokes. Not exactly special, but can be very useful for syncopated rhythms.
Let the bass play
If the bassist is playing the root notes, then the guitarist doesn't need to. Frees the guitarist up to play other parts.
Too much to do
Guitarists only use one pick at a time, but can use several fingers. So think about whether some finger picking would help. In addition, I use a combination, playing some notes with pick, then treble notes with fingers, often alternating, but sometimes together. Works well at emulating piano parts.
It's also the only way I've found to keeping the strength of the normal guitar notes while playing a melody on the higher strings. It takes practice but is well worth it since it opens up a whole new area of complexity.
Stretch
Bassists usually only play one note at a time. That's an overstatement, but stretch your bass player. For instance, if they play chords or at least octaves, they can free up the guitarist from chordal duties. Also the bass is a great instrument for filling in string parts. It doesn't sound like a string section, but fits in a live mix well enough, especially when played around the 5-9th frets on the A and D strings.
Composing
Why is this in the composition section? The more techniques a player has, then less restricted the music will be. By thinking how to deconstruct a song to a minimum of instruments, then you get to realise how much of a a song is superfluous. Look on youtube and you can watch any number of acoustic covers of songs, not all good mind you. The better versions have been arranged by musicians who have thought about which notes matter most and have fitted them into the composition.
Any other ideas
I'm interested to hear of other ideas or your experiences with this.
Part of a series by Award Sounds offering a selection of creative ideas to kick-start or rejuvenate a composition.
Create a random part and modify it until it becomes musical. I'll describe a few options for taking a random parts and the processes I use for making them more musical.
1. Creating the randomness
I set Logic on a 4 bar cycle and hit keys at random on my keyboard. I chose a clean electric piano sound since they highlight any dissonance. It's not truly random because it's difficult to unlearn how to play, but after cycling through the 4 bars a few times it did become a mess of notes.
I could have used the humaniser function of Logic with the notes set at random (oddly, to de-humanise the part) . Didn't think of this until afterwards.
The glue part was because I was recording twice. I had to glue the parts together. There was no need to hae done this, it was just because of where I'd started and stopped recording and what I'd wanted to include.
2. Lengthen the notes
I changed the notes to half-time, doubling the length of each note. I expanded the region and the cycle region to account for this. So now I've 8 bars instead of 4.
3. Deduplicate overlapping repeating notes
If an event has a closely following event of the same pitch overlapping, this will turn them into two events, but the first will finish just before the second starts instead of overlapping.
4. Quantize
The placement of the events was a mess, they were all over the place with no real rhythm. I played with the grid, visually assessing which grid seemed to fit best. This was purely a quick estimation. I also changed the quantize value a few times and I settled on 24ths (triplet semi quavers). I also liked 12ths (triplet quavers) but it seemed to rigid, especially for having come from random notes.
5. Process length for overlapping notes
I did this, then decided to undo it. I liked both, but decided to keep the longer notes ringing. This option means that Logic will identify chords and ensure that the events are the same length. Without it, some notes in a chord may ring more than others.
6. Humanise Velocity.
I processed the midi, running it through the humanise function focussing solely on velocity. There wasn't a major change here, just slight enough to add some odd stress patterns. Maybe a groove template or similar would have been a better feature. Not sure.
7. Effects
These are in the order I added them (they're in a different order in the signal chain)
7a) It was crying out for a rotary cabinet emulator. Didn't need much, just something to take the edge of the louder, higher pitch notes.
7b) Guitar Amp emulator. I like using guitar amp emulators on non-guitar instruments. It's rare they need the full-on, high-gain distortion models, but mostly, just a clean or slightly overdriven blues amp can make a massive difference to a sound. For this one, I went with a clean amp.
7c) Overdrive - to have more control over the tone of the overdrive than provided with the guitar amp emulator, I went with a separate overdrive effect. Again most of the effect was dialled out to avoid losing the music in the distortion.
7d) Compressor - again to take the edge off and make the sound shine through more.
Some other ideas?
To be honest, add any effect and see what the result is. I tried a few that didn't work out in this case.
Gallery of Screenshots for each part of the process.
Want to do better?
I've attached the midi file so you can try for yourself. Let me know how you get on.
Part of a series by Award Sounds offering a selection of creative ideas to kick-start or rejuvenate a composition.
I guess most drummers will know this, but if you program your own drums or use loops, think about using a different groove for your song.
The basic pattern
Standard drum pattern that most of us learn is a 4 beat combination of bass, drum and hihat. Hihat every beat, bass on the first, snare on the 3rd. And repeat ad nauseum.
Then we get to add more bass, use 8th notes and so on.
What's the focus?
I was collaborating on a track a couple of months ago and as we we browsing the library of drum loops, we just couldn't find anything that suited. We ended up creating one ourselves and it was interesting to see the differences in how the two of us approached it.
I approached it very much from the basic, rock drumming angle, laying down the main beats first. My collaborator took a different approach and focussed on the choked hi-hat pattern. I then had to place the other beats around that.
Which was better?
It all fairness, neither. But here's the crunch, they drastically changed the feel of the music. It changed the feel so much that when I recorded a new part (say a guitar melody over the top), I played that differently.
So we recorded both sets.
One became the music for the opening credits, the other became the closing credits. We were also able to switch between the two main rhythms in the middle of track to change the feeling yet again.
Live Drummer
It's a common trick. Drummers will regularly change the rhythm pattern, if not you may as well use a drum machine! Using a live drummer means they can respond to the way the music's going. Moreover, they can change the rhythm, turn that around to drive it in a different direction. The difference for us was that we'd applied it more excessively than usual, but it still worked.
Ahead or behind?
Is this the same as playing ahead or behind the beat? No, it's neither. It's just syncopated, we were still playing at a regular syncopated interval. The choked hi-hats were on the "ands" of "1 and 2 and 3 and 4 and" for most bars/measures instead of on the beats of "1 2 3 4".
Part of a series by Award Sounds offering a selection of creative ideas to kick-start or rejuvenate a composition.
Take an old track and change the style of it.
I'm amazed how many tracks can be given more life by changing the style of music. I pick tracks that I've never fully finished to my satisfaction and apply a different musical style to them.
Identify the original style
What's wrong with the original track?
I usually pick a style that will help fill the gaps. For instance, if it's just plodding, I'll add something with a groove (e.g. motown). A few ideas:
A good way to approach this is as a remixer or as a band about to cover your music. Ask the following question:
What could another band do if they had to cover your music?
Part of a series by Award Sounds offering a selection of creative ideas to kick-start or rejuvenate a composition.