Intro:

When starting this unit, I was so excited to be able to have greater freedom of what I could create, as I noted heavily in my previous blog post. I was overjoyed to be able to have my musical ideas come to fruition with something other than Garageband. At the same time, however, it was overwhelming finding a place to start. I would spend a lot of time just staring at my computer screen thinking about what I would want to come next, compared to the previous project which had long clips that I could just put together.

I first started my project by thinking about the bassline that I wanted, since that was something that really didn’t go well last time. The bassline from “7 Nation Army” was really stuck in my head, and I wanted something that clean but didn’t want to copy it, so it was sort of unproductive. I also tried putting a step clip with a drum sampler to make the beat but ended up switching back to the MIDI input version because I couldn’t figure out how to get more instruments to control. At the end of my first session, though, I already decided to create a piece that had a lead synth that was lively and melodic.

Early on in my project, I created a melody that I would stick with until the end:

In the end, I ended up building off of my melody, then adding the drum/percussion, adding the bass, and then adding the ambiance/sound effects.

Components:

Melody:

My melody portion consists of a thematic repeated section spaced out with distinct other sections, as shown in the diagram below.

With a tempo of 120bpm, I started off fast with the “thematic” melody, which I was able to make manageable by having slower synth notes (otherwise all fast synth notes would be absolutely painful). I ended finishing the song with a variation of the last few notes of the main theme. The key signatures I tried to stick to were D Major/Minor.

When deciding the instrument that I would use, I decided to use the 4OSC plugin. Initially, I was thinking of using the Piano or Flute instrument because I was more familiar and comfortable with orchestral instruments, but I ended up using the “Solid Lead” one with was cleaner and livelier.

Drums/Percussion:

I created this part using the built-in 808 drumset in the drum sampler after making the melody. I ended up using quite a lot of drum kicks, which I really thought was able to propel the song forward and create a steady rhythm. A particularly interesting thing when creating this part was using the claves in the “Intermission” section of the melody I mentioned above. Traditionally, in my head, I always imagine basic beatboxing when I think of percussion (kick, snare, hi-hat), so experimenting with claves was something exciting and new. I found that claves serve very well as a gentler alternative to the snare, which worked perfectly for me in the “Intermission”, where I still wanted to have a beat but didn’t want something too heavy and disruptive.

I also found the clap a good alternative to the snare, which produces a more dramatic effect, which I used at the end of measures at the ends of sections.

Bassline:

The bassline was not very interesting, as I was already quite satisfied with my drumline and melody. Besides a groovy bassline I used for the thematic melody portion, the rest of the bassline was essentially just supporting the melody, having the lower octave of a note played with a bass that I found in Subtractive. I found that the bass notes coincided with the timing of the kick, but I found that didn’t really affect the overall feel too much, because the note sustained longer than the kick, so the note could still be heard.

Ambiance/Sound FX:

Originally, I planned to have a synth pad sample that would serve as the ambiance, but the key was just way too off, so I had to delete it. Instead, I switched to flowing water as the background for a couple of the sections. The motivation for doing this was because 1) I remember this idea from before and wanted to try it, and 2) I thought that flowing water would serve to make my piece more “liquid” and more casual with the random pitches from flowing water. Of course, I had to use an equalizer to get rid of a lot of the excess white noise and used a fade out to make the noise transition better, but I think it turned out decent. I also decided to add a “sound effects” section to make my piece livelier and not just purely instrumental. I chose to use the “YO” sound effect because I thought it would be a good transition between sections that gave a more human element to my piece. I admit, though, that for both of these, I added them at the end, so I didn’t really see these as necessities, so I gave them a smaller role compared to the MIDI components.

Plugins/Panning

Panning:

In all cases of panning, I used it with automation, to have the music change between the left and right inputs. I used this in two places: the main melody and the “YO” sound effect. My main melody repeated 3 times, the second time an octave lower than the first, and the third time an octave higher than the first. I panned the first centered, the second part left, and the third part to the right. I thought that this would create a cool contrast, but I feel like in the end the reverb/delay may have made the experience less optimal. For the “YO” sound effect, I used automation to make panning occur during the sound effect to create the effect of the “YO” passing by.

Equalizers:

I used an equalizer on each track, using them as more of a lowpass/highpass filter style. For the melody, I removed the higher frequencies that were produced from the synth to get a cleaner and less jarring sound. For the bassline, I also got rid of the higher frequencies, but this was not a very noticeable effect. For the drumline, I increased the higher frequencies, to emphasize the claves more and give a bit more commanding power to the snare. As I said above, for the water sound, I got rid of a lot of the low-frequency noise to get a sound more similar to water dripping. Finally, for the “YO” sound effect, I essentially made a bandpass filter that centered around the human voice range.

Compressor:

I used a compressor for each of the tracks to get rid of the jarring loud sounds that sometimes would end up clipping. I feel the most important part was applying the compressor to the drumline, which had the loudest output that was sometimes just way too dominant. The effect of the compressor on other tracks felt very minimal.

Reverb:

I added reverb to make sounds way larger since my composition was basically all MIDI and felt too “artificial”.

Challenges:

I was surprised by how difficult it actually was to compose my own music. I haven’t ever made a whole musical piece before, and basically just sometimes generate melodies in my head, so creating something succinct and blending together was difficult. Also, with so much freedom in MIDI, I didn’t really find much use for recorded samples in my project besides serving as the background or for sound effects. Also, the MIDI interface was a bit frustrating, as I kept switching between the pen and the selection tool using the mouse. But otherwise, it was a fun and enjoyable experience.

Overall:

This project really forced me to think more musically when creating a piece. I learned quite a lot about what things sound good together from getting a hands-on experience of literally making the notes one by one, and how difficult it can be. I now have a greater appreciation for those that are good at making music, since it seems like a really precise form of art still with a good degree of freedom to it. In the future, I hope to try and recreate retro pop songs in Waveform in MIDI and make my own parodies/dubs. Anyways, below is my piece, and don’t be afraid to roast me!

Looking for inspiration, I found this piano thing deep in the cobwebs of my voice memos. It felt very loopable and I liked both the rolling short-short-looong-looong-looong feel and the harmony (on a macro level I think you see a Bb dominant natural 11 chord “resolve” via voice leading to a Bbm9). So here’s the kernel (has some nice authentic background distractions):

Then I inputed the loop into Waveform using the Subtractive oscillator, and I really liked the Africa Horns (like, the Toto song) synth as a basic piano sound. I really liked the punch, but the pitch was a bit too grating and squishy, sorta. Lowering the frequency cutoff of the lowpass filter basically solved almost everything. To make it more piano-like, I also made the attack of the filter envelope agressive and shortened the amplitude release so that the sound didn’t accumulate and get soupy.

Africa Horns before and after (it’s subtle!):

I also added quick 808 drums (drum sampler) and bass (I chopped up and rearranged the Brass Brigade Sub Bass loop). Lastly, I added in some eerie quartal harmony in a pad that I made in Subtractive. I really liked the sound of a lead synth called “sub bass,” so I turned off the monophonic setting and made just a few adjustments (faster attack in general, and less detuning). Reverb and stereo widening especially helped the surreal effect.

At this point I figured out that we’re in 6/4 and we could repeat four cycles of the groove like so:

As you can hear, I also sped up the tempo at this point from 100 bpm to 140 bpm. Instead of feeling just sluggish, the groove suddenly felt both like it was plowing ahead and like it still had a heavy, behemoth quality. Especially once I added some dramatic strings, it became, as my floormates put it, “very sci-fi.”

Let’s talk about the strings. I wasn’t going for cliche or anything– not tryna make generic clip art–but it just needed some film-score dramatic strings. I settled Orchestra Loop #1 because the chords worked, and pitch-shifted it way down. I spent a lot of time *fiddling* with the string sounds in Elastique Pro, first of all, because the chord changes needed to align with the beat (I didn’t realize how hilariously out-of-sync it was until I showed my non-overthinking-musician friends, who noticed immediately). It was also frustratingly weird that the key you transposed the samples to didn’t match. It’s not that I have perfect pitch, which I only ever use as a party trick. It’s that, when you input a key that is higher, the program literally transposes the audio to a lower key. Same goes for bpm, in fact (higher bpm = slower speed). Ah well. Let me know if I’m missing something.

I followed the exact same process with the high strings, splicing up, transposing, and time-stretching the “Nightingale Drive String Line” loop. This is where it became very movie score.

Meanwhile, I ordered the entrances of the players (piano, drum, bass and pad, strings). We started out with just the original piano groove, with some claps that fade in. My non-overthinking-musician suitemate thought that the claps were too confusing, and I agreed. Honestly, I feel like clapping on the 2 and 4 is way cooler, but here everything is just way too fast for it to lock in. Compare the offbeat claps:

With downbeat (what I went with):

For the drums I used a notch filter to reduce the annoying pitchiness of the ride cymbal. I relied on a low pass filter for now to make them tolerable before I found something better, and I really liked automating the frequency cutoff of the lowpass to rise so that the drums intensified until a mini drop, where I said “hey.”  For the “hey” vocal, I phased it, panned it hard-left, and added a wonky “energy” filter. Then I notched out a bunch of extraneous frequencies, creating some nice stalactites:

The drum build sounds like this:

I didn’t know where to go next so I asked myself, what’s the opposite of this? I listed adjectives like suspended, glassy, high-pitch, falsetto, human, hovering, slow, fragile. Subsequently, I had an existential moment during a sorta boring music theory lecture. I saw myself in my Zoom window appearing to be suspended in the sky and tree leaves that were above me. Feeling poetic, I imagined falling from the sky and not knowing who I was. Then I pulled out a rhyme dictionary and got this:

Falling from the sky

Who am I?

Gonna die, say goodbye

Who am I?

Time goes by.

Not exactly Robert Frost, but extremely laconic and morbid, which I liked because it felt so fragile and open-ended. I set the text to a melody and took like seven takes trying to ensure that my falsetto was 90% rather than 100% unbearable to listen to. Afterwards I automated pitch shift on the vox to fix some wonky notes (so much for perfect pitch lol).

I also made a pad sound to go along with the piano. Going for a warm but not too heavy sine sound, I made a sound in 4OSC with four sine waves with a steep (24dB slope) lowpass filter. I tuned one oscillator to an octave above pitch and another to 2 octaves above, to make sure the sound wasn’t too flat/low/heavy.

One of my favorite discoveries was adding in a mellow wineglass hit (with emphasized partials in the mids and reverb) when the first words enter, as a subtle bell tolling effect:

To transition to the second theme, I knew I wanted to drop out all the instruments in the initial groove until the high strings were suspended… suspensefully. Showing this to my non-overthinking-musician (are you sensing a theme??) suitemate, he came up an incredible transition into the second section. “The lyrics are ‘falling from the sky.’ What if you have the high strings fall down in pitch?” Genius. I added some beautiful wind chimes and wind from Freesound and got this:

After the second theme, I simply tried doing the reverse of this transition (the chimes, chords, and wind pitch-shifted upwards by a whole tone). This gave me the idea that it would be really cool to have the original theme enter back, in full force, in a new, brighter key (2 degrees sharper on the circle of fifths – think C modulating to D). I sent this reprised first theme to a bus with an automated Low Pass filter where the frequency cutoff gets higher (mirroring the original entrance, but this time with everybody involved).

At the end, I wanted the vocals from the second theme to join in with the first theme, with so much cathartic distortion that they felt almost electric-guitar like. I got a bit more of a “third grader trying to do mouth trumpet” sound, because I didn’t really understand what the distortion plugins were doing, but at least I tried.

Finally, I really enjoyed the end part, where I brought some of the sound effects (wine glass, chimes, nice chords from the 4OSC track) together on a V-I cadence that was both jumbled (lots of elements) and resolved (harmonically pleasing). It was actually really hard to keep them all in tune with each other, and I automated the pitch shifter at tiny microtonal levels on many many tracks throughout the piece!! I hope that that felt appropriately in spirit with a piece that basically puts two contrasting themes together and attempts an ambitious crossover of the two with a sorta gross-sounding screamy vocal.

Here’s the final product. Hope you enjoy!

Concept
I started by using the keyboard and trying to come up with chords that I liked. I came up with a chord progression and a melody, and I put those into MIDI (which took a long time!). I experimented with a few different sounds on Subtractive — I wanted something relatively clean/natural-sounding (because I liked how the piece sounded on piano), so I clicked on “Clean” under the character category and found Classic PWM JH (which I used for the chords — though I did end up later adding some filters and it no longer sounded as natural). For the melody MIDI, I used “Turn Me Loose Gliss Brass” – it had a more pointed feel, which I liked for the melody.

2 MIDIs – Melody & Chords

• I wanted the piece to start out relatively quieter, and I thought that the chord MIDI was too brassy, so I applied a low-pass filter (which was very effective!). I additionally applied the filter to the melody MIDI, which was also a little too brassy.
  • I wanted the second 4-measure phrase of the melody to be brassier, so I increased the cutoff frequency
• In the chords MIDI, I wanted the 3rd and 4th 4-measure phrases (counting from when the melody comes in) to be lighter, so I used a high-pass filter to reduce the lower notes; however, I wanted the last D major chord before the next section to stand out, so there I increased the low-shelf gain
• I also played around with panning in the beginning, having it go from subtly left to subtly right a few times (using “Automation Write Mode” in the bottom right corner)
  

  

• I randomized the velocities for the MIDI clips (but couldn’t completely hear the difference)
• There were specific notes in the chords (around middle C) that I wanted to emphasize, so I used an EQ filter and changed some velocities
• I experimented a little with “Nonlinear Space” (because I wanted to experiment with the space of the sound), but wasn’t a huge fan of how it made the MIDIs sound, so I turned it off

Percussion/Sounds

• I used “CM 7 Kicks” from the SSLIB because I liked the sound — I used the Time Warp filter to change the rhythm to fit my project
  • When I added reverb, there was a high pitched snare-type noise, so I used EQ to emphasize the lowest frequencies
• I also used “Ambition Beat” from Waveform itself – I wanted to emphasize the higher-frequency noises, so I used EQ (see picture below)
  
• I added a “whoosh” noise (SWEEP003 from SSLIB) to end the first section, and then added an “A” played on the harp (also from SSLIB) to add some texture/variation to the second section

Challenges/Issues/Errors

• With the melody MIDI (Turn Me Loose Gliss Brass), at first when I had two notes at the same time it slid from one to the other (ex. I had a B and then added an A halfway through, and it slid from B to A) — to fix this, I changed “Glide” in Subtractive to 0
• Errors
  • I had a weird error: with the notes [c bflat c], the b flat wasn’t playing; to fix, I quit and re-opened Waveform
  • I had a strange error when I was quitting Waveform — no idea why it happened (see picture below)
• For a while I was writing automation but wasn’t able to hear it — I realized that I had turned “Automation Read Mode” off
  • in the picture to the right, the top left is “Read Mode”, and the second button in the first row is “Write Mode”

Overall

Overall, I enjoyed the freedom of using MIDI to be able to write my own melodies/chord progressions. However, it was definitely a lot of work to input each note individually! Hopefully, in the future, I’ll be able to use a keyboard or other instrument to input notes. With more time, I would have added strings and other instruments to the project and experimented more with the settings within Subtractive itself.

The Final Project

The track:

PROJECT OVERVIEW:

This project I made was meant to give more of an eerie, off-putting, sad feeling (what’s a better way to encapsulate midterm season, am I right?). I was very hesitant to put in my vocals because I think my voice is annoying, but I decided to in order to give that kind of breathy vocal sound that I was able to exploit to my liking using Reverb. I really went along with this eerie idea once I found the Cloud 1 Subtractive synth sound, which was a part of the Ambience category when looking for different subtractive MIDI sounds, and it really worked to my liking to create the odd feeling.

MIDI:

For MIDI, I chose to make four tracks – one I called “scintillating,” which was the arpeggio subtractive synth that is constant throughout the sound except for when it drops out during my vocals. The next MIDI is the cloud ambience, which makes you feel like you are in outer space. The third and fourth MIDIs were combined, and influenced heavily by our drum sampler MIDI tutorial lecture in class which I rewatched ~many times~ to get this right. It was the drum sampler and STEP track (I really liked using a STEP to add “flavor” to the drums).

For each of the MIDI tracks, I experimented with the 4-band equalizer, as hinted at in the prompt for this homework. Using the EQ, I was able to scrape the higher frequencies and not allow for anymore clipping like in the last project I submitted (yay for progress!). This was especially important, in tandem with the compressor, for the drum sampler MIDI recordings, since the hi-hat’s decibel level would go way too high when looking at the bottom left corner of Waveform PRO that shows the sound levels. With the compressor and EQ, I was able to handle this problem. Below is a screenshot of the step and drum sampler pattern I created in the beginning of the track, where the step pattern does not continue at the end.

The other MIDI I used was the Cloud 1 subtractive synth, which I used a lot of reverb on through the aux bus, to make the “room bigger.” The other MIDI connected to the aux bus was the drums and step. I also used the nonlinear space plugin on the Cloud 1 subtractive synth, but wasn’t really sure if I heard a difference. A really good compressor plug-in was the AUMultipleBandCompressor. The MIDI was hard to navigate at first because for a whole first two days, I thought there was something wrong with my virtual MIDI input and Waveform itself, because I couldn’t input values and hear the sounds. When I went back to the lecture, I still was confused as to how to make those sounds. Finally, I realized I needed to choose the output to be the MIDI and then play on the piano that pops up virtually on the left of the track.

For the “scintillator” MIDI track, which was actually called “Comb Harp JH,” I was able to create an arrangement/pattern that repeated throughout going through notes. It was interesting because this MIDI would not really stop playing even when pausing the track. I also used the automation on all/most of the MIDIs, using the volume and pan automation. The volume automation was used for the Comb Harp JH in order to quiet it down in the middle then bring it back up. I also used the volume, pan, and reverb automation on the vocal track. The most important plug-in for the aesthetic of the piece, I’d say, was reverb tweaking the “room space” because that was the important heady/airy quality that was given to the Cloud 1 synth and the vocals, making them sound much more spanning/full. Below is the automation curve and pattern for the Comb Harp JH synth.

 

For the vocal track, I made an interesting panning automation so that it could sound like I’m singing in different ears, where it goes from left to right as the vocals progress.

You can also see in this image that there is a overlap in vocal tracks that I recorded. This is because at one point, I didn’t really like the final note vocal thing I did, so I wanted to delete it. But, because Waveform doesn’t let you delete in between the tempo lines, I had an awkward gap in between vocal words that I kept for about 5 days because I didn’t know what to do with that gap and thought I could just use Echo or Reverb to carry over it. But then, I realized I could create a cool sound by overlapping the vocals and it did just that! I was able to create like a sense of call and answer very quickly like a cry of help in the vocals, giving that kind of lonely feeling.

SAMPLES:

For the samples, I used a synth sample called “Afloat Pad” which I used the phaser plug-in (it’s kind of like 40SC but easier to use in my opinion) and AUDynamicsCompressor for. I think the 4-band equalizer is much better than the AUDynamicsCompressor to be honest. I didn’t really hear a difference with this plug-in. For the vocals samples, I recorded some onto my computer using the MacBook Pro mic (these were the quieter samples that I had to use the EQ to bring the volume up on) and the one on my phone (that I had to use EQ/Compressor to bring the loudness down on). It was interesting how I could apply the EQ/Compressor to different segments within the same track to produce a different result.

MAIN TAKEAWAYS:

Now I know how to modulate volume and ensure no clipping occurs! And, I don’t really like the Distortion plug-in. I tried it once on the Cloud 1 synth, and it sounded so horrible I immediately needed to take a break from the project. Doing this project also made me realize how things can blend together in music if you spend a long time on the same thing, and I really needed to take frequent breaks to prevent “ear blindness.” It was fun messing around in Waveform again, however, and I’m so glad I’m learning so much from this class!

Below is the Final MP3 result. Thank you so much!

 

The Inspiration

I used Daft Punk’s Short Circuit as the inspiration for this song. I liked the chords that the second half of the song had so I did my best to recreate it in Waveform. This gave me a chance to experiment with scales in Waveform and learn to sequence notes to form chords. I used the G#min scale and Waveform is pretty helpful with working with scales because when you set the key, the piano roll is highlighted with notes on and off the scale as seen on the right.

I wanted to have a similar structure to the song so I decided to create two sections. The first is more crunchy while the second section is more atmospheric. The song is in common time.

The Skeleton

First Section – Crunchy

 

MIDI Chords

I used the Crystal 2 preset from the 4OSC plugin. I split the chords into two by duplicating them and applying slightly different effects to each track. For example, the first track uses a phaser.

I automated both of the tracks containing the main chords to change the pulse width of the 4OSC plugin, giving it a more distant sound after the 8th bar.

Drums

The drums follow a simple pattern, not really sure how to describe it. I’m not sure if it’s considered syncopated but here’s how the sequencer looks like for the rhythm. I used the Drum Sampler with the 909 preset for this one and adjusted their sensitivity to get the sound I wanted.

I set the velocity pretty low because I did not want the drums to overpower the other instruments.

Grungy Lows

I used the Subtractive plugin and found a preset called “Kill the Woofer JH” and it sounds exactly like what you think it would sound like. I inserted a step clip to see what it would be like if I used one of those not for the drums. It still worked and I found a pattern that I liked.

Second Section – Atmospheric

Airy Chords

I used the same chords as the first section but only with one track this time. I used the same 4OSC plugin and found and used the NC-17 to increase the sound level.

Bass

I downloaded one of the additional plugins from the Tracktion Download Manager and used the RetroMod 106 plugin for this track. I used it as a lead out to the end of the song and it blended pretty well with the atmospheric

Strings and String Break

Originally, I wanted to add a string solo to the second section. When I rendered a quick export of the piece in MP3, it cut out a measure of the string and produced what sounded like a string breaking. You can hear it at around 2:20 in this first render. (Turn down your volume, it’s normalized)

Instead of a string solo, I wanted to recreate that effect to create a feeling of unrest. I couldn’t find an exact way to recreate the sound so I just decided to add a note from a chaotic-sounding MIDI. I found a preset called “Waveform Percussions” in the Subtractive plugin. It’s an interesting one because it seems to be randomized on every playback and it does the job in different ways.

Using Plugins with MIDIS

A lot of plugins do not work with MIDIs, but there are a few that do. Here are some of the ones I experimented with:

• Reverb – does not work with MIDIs
• Low Pass Filter/High Pass Filters – do not work with MIDIs
• Phaser – works with MIDI. I used it with the 4OSC plugin.

Mixing

After finishing up the main components of my piece, I decided to listen for any changes that should be made to make the sounds complement each other better.

Drums

At first I kept the drums constant throughout the whole piece but I changed that when I felt that it didn’t really fit the second section. Instead, I replaced the second section’s percussion with claps, which I had to turn from a sample into a MIDI to apply Reverb to it. When I applied reverb, it had a lot of highs that was just really unpleasant to hear, so I used the 4-Band Equaliser plugin to get rid of those annoying frequencies.

Bass

During my mixing stage, I felt the bass was too intrusive in the outro so I decided to change the instrument used from the plugin. The bendy bass was called Alpha Juno Bass and I used Juno-60 Bass 02 instead. It mixed better with the atmospheric outro and did not stand out too much from the other instruments.

Main Chords

I panned the two tracks separately; the first is slightly more to the left channel and the second slightly more to the right.

Audio Bus

I noticed that two of my tracks both use phaser so I rerouted them to my final track which I called the Phaser FX track. I applied it to both the grungy lows and primary chords from the first section.

Here’s the result. I had a few troubles getting the volumes right because it sounded like they were at the right levels during playback on Waveform, but exporting them completely changed the levels. I definitely could use some improvement because the second section turned out louder than it should have been.

Phase 1: Troubleshooting Technical Issues

I had trouble finding the subtractive synth, and I realized it came with the pro version. For some reason I had the free version and when I tried signing in with Petersen’s account the registration process could not complete. I removed all the plists from my libraries and preferences to solve this. However, I still had no luck. Over the weekend I decided to get a free trial of the Pro version with a personal account, but I was eventually successful accessing pro with Petersen’s account.

Phase 2: MIDI Fiddling and Recording

The primary synth was a 40SC preset to the stab synth. I fiddled with EQ and compressor. This way, the synth wouldn’t “stab” my ears. For this lead I just used the pen tool to write in my notes. At first, I tried playing the melodies on my laptop keyboard, but my hands kept fumbling and hitting too many wrong notes.

The bass synth was a subtractive synth with a preset to the Bernie’s Mini Bs. I activated the legato button on the bottom left of the virtual instrument. I felt the attack on the notes was too sharp at default, so I set legato on which takes much less time than automating pitch to different notes. Most of the notes for this bass were recorded using my keyboard, but I had to quantize and sometimes manually adjust the notes not only to stay in time, but also to stay in time the way I wanted them.

I wanted to fiddle with velocity for the all synth notes to give a sense that they were being played by a person and not a robot. However, tinkering with velocity made the DAW crash, so I had to go on without velocity manipulation for this project.

Both the 40SC and Subtractive synths were looped for the first half because I wanted to start with a clarinet solo. I sidechained the stab synth with the clarinet so the two wouldn’t compete.

My percussion was the Micro 808 Drum sampler. I decided to use step clips since I wouldn’t have to manually adjust or quantize notes. I wanted to create a “swing” beat for this track. I disliked the overtones the sample sounds made, especially the snare. I did not use a notch filter and instead just used a pitch shifter to remove the ugly “different key” overtones. This fixed my problem and was left here for the rest of the project, though next time I am aware the notch filter might be better at the job. Throughout the song, more percussion notes were added in addition to the continuous swing beat brought by the kick and the snare. Notes were added in off beats as the song progresses. A side stick, a kick beat, an open hi-hat, all added to fill in the blank spaces from the start. Of course the hi-hat’s release lingered for too long, so I was happy I could shorten the duration of the sound in the drum sampler.

For recording samples I used my clarinet again. I followed the same format to improvise, playing whatever came into my head with multiple takes. The clarinet was positioned the same way, above my microphone. Anjali asked me previously what were the added benefits of my Blue Snowball microphone. One pro I found was that the FFT was much more flat than my iPhone 11, so the frequency pickup would improve the sound of my clarinet. However, I realized how airy my clarinet sound was in my previous project. This airiness was not as noticeable in the few iPhone recordings I did, which shows how sensitive my Snowball mic is. Part of it was because of my bad embouchure, but switching reeds made for some notable improvement. The sound of my clarinet was richer here. Like Ethan, I usually have too much fun with reverb, but I also took a step back to try other plugins. Instead of using reverb, I used an EQ plugin and raised the treble and mid boosts to thicken my instrument’s timbre and improve its space in the soundscape. Last time a lot of tracks sounded narrow and isolated on different areas of the channels when I panned them, but with EQ I was able to widen the space for each synth and instrument. Reverb also made the airy sounds from my clarinet audible. The reverb plugin is only used in the song’s final clarinet clip.

For the second half I also recorded myself whistling faintly to create the sound of wind. The wind sounded very pleasant but also very unsettling to me with the whistling, and I found this juxtaposition fascinating. The air hitting the mic directly created some wind noise, creating the faint sound of thunder. 

I followed through with the call/answer technique from last time. Instead of using my clarinet, I used 40SC stab and string synth, panning them to opposite levels. The strings would play for two measures and the stab would answer. I played with various release parameters on the string synth which added a nice flicker effect.

Above is one of many takes of my faint whistling. I automated the panning to oscillate slowly between left and right levels later on in the final product.

Another track is an AUMidi synth. I was fiddling with the line tool earlier and I liked the sound the written patterns made. I used these patterns to transition my piece. The addition gave some variation. 

The ending string synth was a preset on the subtractive synth. I was having trouble figuring out how to end my piece, and I thought it would be interesting to finish on the sadder side. I played a few notes using the keyboard, adjusted them to stay on time, and added a tremolo, since I thought it would better accompany my wind whistling. I also automated the tempo from the starting 133 BPM to a slower 88.08 BPM to signify a conclusion. The tempo was again slowed to 64.34 BPM so the clarinet had time to shine and end the piece.

I had a lot of difficulty adjusting attack and release for the subtractive synths. The strings have a very gradual attack and the bass has too quick of a release. During playback, random notes of the subtractive synth would sustain even if I did not write them in. Sometimes I would pause and I would still hear those notes. In my limited experience playing with MIDI, I have dealt with this problem in the past. At the moment this is a mere annoyance but I am worried the problem might escalate. What if I am doing a project and the random note just sustains nonstop, even if I relaunch the software? 

Phase 3: A Conglomeration of Plugins

Matters got very messy here. There were so many different combinations of plugins to achieve different sounds. At the start of the project, many of my plugins were adjusted by ear. Even being aware of what each one did, I still kept trying to fiddle with different combinations. For example, I thought to myself, “What does a treble level of 6dB combined with a compression ratio of 3:1 do?”  Another time I wondered, “Should I insert distortion with a phaser?” Too many questions went through my head.

I ended up having two types of EQ plugins on my percussion track. The AUparametric EQ had automations while the regular EQ improved the timbre of the drums. At some points the drums were overpowering everything else. Removing the AUParametric EQ plugin made the drums faint but evenly balanced.

The same problem happened with my stab synth, but with compressors. One (Compressor/Limiter) played the role of side-chaining with the clarinet while the other (AU Multi-band Compressor) supported the EQ easing the sharpness of the sound. I kept both dynamics filters because of these important filters each one played.

On the bright side, I realize that the EQ plugin is another great tool to improve the soundscape of my music.

Many of the dynamics filters on my tracks were compressors. I used presets for respective MIDI tracks, for example, setting my percussion to the Tracktion Compressor: Drum. I did want my clarinet to lead, but I was worried that it would overpower everything else. The clarinet sounded like it was trying to compete, so I added a compressor and adjusted threshold and ratio by ear. The visual parameter adjustment was extremely helpful for me.

Several songs I know create a buildup by automating a lowpass filter, increasing its cutoff frequency and resonance. I tried doing this with my bass synth track, also automating the volume and pitch of a single, sustained string synth note. My clarinet also holds a sustained note with automated volume crescendo. I added a phaser plugin on the clarinet clip since I liked the oscillating push the plugin added for the buildup.

All these clips mixed to make a rich build in my piece. Following the buildup, I automated the bandpass on the stab synth back to default to restore the established “sharp” noise it originally came with. 

However, after the buildup, everything was significantly louder than before due to earlier plugin automations. I had to automate the master volume to keep the volume levels stable. I did not trust normalizing the audio to do that job. I did not normalize when exporting this time, and it sounds much more three-dimensional. The volume, however, is very imbalanced in some tracks so I had to go back and manually adjust as much as I could.

In addition, I realized that my audio reached the 0dB level several times. I was worried about clipping so I reduced the master volume, but then there was some static in playback. I was unable to fix this with whatever volume I adjusted, so I will have to find a solution in the future.

The mess got even messier with the aux bus. I just used an FX Chorus plugin with an aux output from the lead, string, and bass synths to reinforce the conclusion of the song, and to vary the MIDI tracks. However, after the class on October 13th, I should have known I could just create a new track as the primary output for the other tracks and automate my plugins there.

Phase 4: Mixing 

After installing the Voxengo Span plugin, I monitored the frequency spectrum and it seemed that there was a great absence between 10kHz to 20kHz. There was a spike around 200Hz to 300Hz with my clarinet, so I adjusted EQ and volume as best I could. Halfway through the piece when the 40SC string synth came in, those frequencies began to fill, but they became absent again following the buildup, despite having the string synths playing. I added AUMidi synth samples with a high pass filter to fill in the missing frequencies for the transition into the conclusion. 

Final Thoughts

The piece achieved the goal I set out, but it was much more experimental than my last. 

In experimenting with plugins, it was fun to play around with them but it is hard to keep them organized. Having two different EQ plugins on one track is not ideal, but this experiment helped me appreciate the power of EQ plugins. While I got some great experience with MIDI, I obviously still skimmed the surface with this piece. I did check out as many 40SC and subtractive synth presets as I could, but for this piece most of them just did not match the idea I was looking for. In particular, subtractive synths were difficult to use, since there were so many adjustable parameters on the synth itself. I agree with Petersen’s description that the Joker vomited on the interface. In addition, I think I will need to vary the kinds of plugins I use. This piece mostly had EQ and compression, and I took a step back on FX plugins, but next time I will be more adventuresome.

I DEFINITELY need to work on mixing. After the extension I wanted to improve upon this, but a big barrier was my disorganized plugins. Next time I should use the single-output send method that Petersen shared instead of individually trying to adjust my tracks.  That is not to say I will eliminate individual plugin adjustment methods, but the single-output will be something new to learn, and give my next piece some coherence. Now that I have the Voxengo Span monitor by my side, in Voxengo’s own words, I can “mix not just with my ears, but with my eyes.”

I call this piece “Standing up the Summit”

Summary of Features 

Design Idea: Using percussion to create a swing beat, simultaneously adding a bleak atmosphere.

MIDI: 40SC Stab Synth WMF, Subtractive Synth Bernie Mini Bs, Micro 808 Drum Sampler, 40SC String Synth, Subtractive Synth Cinematic Strings, AUMidi Synth

Plugins: EQ for most tracks, with the lowpass parameter automated. Compressors were also used for several tracks, each one set to a different preset or customized. Tremolo was used on subtractive strings. FX plugins include distortion, phaser, and chorus (aux bus).

Automation: These automation tracks were created separately unlike before, when automation curves were layered on top of my tracks which was very inefficient. Included this time was pitch shift, volume (both individual tracks and master bus), lowpass filter (from EQ), Aux bus automation with Chorus, and Tempo.

Going into this project, I decided I aimed to finally make a song that sounds good without an obscene amount of reverb. Reverb makes sounds that are not well-recorded or well-chosen sound less bad when put together. Also, I particularly love songs with extremely open, reverb-heavy sounds. These have both led me to drown my projects in reverb in the past, and for this project, I decided to make sure it sounded okay without reverb too.

The song is about two people I was close with in high school, and my feelings about leaving each of them behind when I came to college. My theme is leaving the past behind – because I am also leaving bad production techniques in the past!

The song started as just synth hits on the 40SC, a synth bass from the 40SC, and a kick sample and a snare sample that I acquired two years ago. I was feeling great about it and it was the best four-track song seed I’d ever made. I showed it to some friends who said it reminded them of Charlie Puth. He’s one of my favorite pop producers so this was a great compliment, and I decided to roll with this for the rest of the song. Below is the song after one day of working on it (plus the effects I later added to these tracks).

 

I made extensive use of lowpass filters on this song. I wanted it to have a dark, sultry, and reminiscent mood, and putting lowpass filters on many tracks helped me achieve this. I put one on both of the drum tracks and multiple other miscellaneous tracks. I also think that automating a lowpass filter sounds really good in intros and outros (for a fade-in or fade-out) although I’ll admit it is a little bit gimmicky and I don’t normally hear it in real songs.

I had a great relationship with the 40SC synth in this song. The presets were enlightening in how the 40SC synth works. I spent a lot of time isolating one of the waves in a multi-wave preset to see how they interacted with one another. I ended up using only modifications of presets in this song, but I feel like I could make a decent synth sound on my own with the 40SC.

Subtractive was much more difficult to use. I wasn’t skilled enough to create a synth sound in subtractive that sounded like it fit with the rest of my song. No matter what I tried it sounded too grating. By this point I also had plenty of synths from the 40SC, so I decided to go for white noise in Subtractive. I made a sweep with it that I used a few times in my song by automating the cutoff frequency on a lowpass filter.

I wrote the lyrics and recorded in the practice rooms in the basement of Pierson College. I recorded close-mic style so that I would not have to bring down the levels of all of my other tracks by an extreme amount. Here’s a sample of the vocals, with WAY less reverb than they originally had. Looking back on it, I think I still could have backed off on the reverb a little bit. Even without pitch correction, I actually feel okay about how these vocals turned out. Here is a section below to hear my vocals with their final processing.

 

There’s also a vocal harmony that I recorded and two vocal effects – both “ooh”s. One of them you can hear at 0:19 on the audio at the bottom of the post. The pitch is atrocious because it’s not in my vocal range and I probably should have gotten a friend to record it. I still think it adds a nice touch and I drowned it in reverb, which I think was appropriate. It acts as sort of a whistle or a chime and it is supposed to sound really far away, like the whistle at the beginning of Magic by Coldplay.

In the future, I would like to find a better dynamic balance between the vocals and all of the other tracks. I think the vocals ended up too loud on the verses and too quiet on the choruses.

The drums are sparse, and I realize that there is no hi-hat in my song. I started with the kick and clap and moved on to other tracks, and I developed a dark tone that I decided sounded better with no hats at all. The drum tracks both have lowpass filters on them for this effect.

The tempo of the song is 90 BPM which is fairly slow for a song of this genre, but I have a quick melody and a steady sixteenth note synth bass, so the song does not feel slow.

Okay, below is the full song. My main takeaways:

• I still need to rely less on reverb, and maybe experiment with more dynamic effects – compression, chorus, delay, etc.
• I learned how to use and automate a lowpass filter to sound super good (hopefully you agree!)
• Gained lots of experience with the 40SC, still need some practice with subtractive
• Learned how to make a bus