The idea behind pulse is the fragile serenity in the rhythm of a heartbeat.
Reflection:
With different circumstances this project could have more time to grow. As it is, i did learn alot in terms of audio effects and have only scratched the surface of whatās possible. Iāve become much more confident in terms of using Ableton, but need more practice and experience producing music. More mini experimentation sessions could help to develop more sounds and techniques.Ā
My mixing skills have vastly improved with a better understanding of audio effects such as Compression, EQ & Filter.Ā
With inspiration low, the next best thing to do is to experiment. Iāve been throwing things at the wall and seeing what sticks.
Iāve gotten more confident with the Auto Filter and Glue Compressor.
Iāve also discovered new effects such as the Saturator, Overdrive, Flanger & Frequency Shifter. I especially like the Frequency Shifter as it can dramatically change a track while remaining in relative harmony. The Saturator is also very useful in making sound warmer and helps bring up the sound on the lower end.
Recent events have kept me distracted and uninspired, and no matter how hard I try nothing seems to be going right. Iām missing the spark that droveĀ āSceneā.Ā
The basic outline and idea is there, but its still sorely lacking character and flavor.
Iāve experimented with all the things Iāve learnt from my research, but they donāt seem to fit together.
Maybe my mind isnāt in the right place.
I did manage to setup my own synth as a preset, and have gotten familiar with the Auto Filter effect. Using Auto FIlter, EQ & Chorus all together allows me to modify sound into something very different. Useful for experimenting with new sounds.
Another tangent, this time into classical. For reasons of which I could write a book about, AirĀ from the second movementĀ in Johann Sebastian Bach's Orchestral Suite No. 3 in D major, BWV 1068, has had a profound impact on my life. Simply put, it is now my earworm that I hear everyday.
This does become a bit of a problem when trying to produce music in Ableton that is of a different style altogether, as my brain will tend to drift into the classical.
So I thought it best to, just this once, submit to temptation and just recreate this piece in Ableton. As an extra challenge, I wanted to test my hearing so Iāve done this almost entirely by ear.
A bit of a technical note: I originally wanted to upload this through youtube but it keeps getting flagged (copyrights yay). Trying to get this through Tumblr proved tricky, at first the file was too large so I had to cut down the quality. Then it went through but got stuck in processing, so I removed audio from the background (I was listening to a performance for reference) and it finally got here. Crappy, but here.
Back on topic.
First off, it was imperative that I found a good sample/sound pack as the default Ableton has is either lacking or terribly synthetic. I found one rather easily from the Ableton website itself and began working. It took at least an hour to get my ear get into the swing of things but once that was sorted, things began to take shape.
Since the piece is slow and has few, easily identifiable, layers it was relatively easy to get notes into the MIDI. The real difficulty is trying to make it sound natural. Even with the best of samples/sound packs, there will be a telltale sign it was assembledĀ rather than performed. The key difference is that humans are never perfectly on time, and instruments donāt always produce a clean and consistent sound. Specifically, violins sound differently depending on how the bow is held and which part of the bow is contacting the strings. Additionally, a violinist has to move the bow up and down which produces a very different sound.
So I have to go in and make it sound more natural, which is easier said then done.
Ableton does have a feature calledĀ āVelocityā, which simulates the behavior of a piano mechanism; a note struck on a piano is louder if the key is struck more forcefully. By changing the velocity, it changes the piece to sound a little more natural. The violins arenāt being played at full force nano-second perfect on time, which makes big difference.
Adding reverb helps a lot too, by creating a sense of space that simulates the performance happening in a real environment.
I think I did a reasonable job on this tangent of a project. Although more time will help make it better, I think I need to be exposed to more orchestral performances in real life to understand better what the real deal feelsĀ like.Ā Technical things like compression, EQ and so on will help create a more controlled and cleaner sound.
As part of my feedback on my formative, I was encouraged to research more into other electronic soundtrack composers. This was a good wake-up call to widen my scope, as I tend to become narrow-minded as I focus on a project.
Clint very helpfully provided a list that compiles the ā10 Best Synth Soundtracks & The Equipment The Composers Useā: https://soundontime.com/best-synth-soundtracks/
Aside from Vangelis, the next thing that caught my eye on this list was Kavinsky: Nightcall. One of my all time favorites.
The song is reminiscent of the 80ā²s electropop but definitely closer to the now modern genre of synth-wave. It is pretty much the quintessential synth-wave track, perfect to drive down a highway late at night. Which is perfect in the movie āDriveā, which on a side note is really good and worth studying. Although the drums keep a steady pace, the unique synths make it feel like the track is slower. I really like the lower synths at the start, I believe the right term is āfatā.
That got me curious as to how music producers achieve this, and my research led me to āSausage Fattenerā by Dada Life.
I first saw it when I played around with Fruity Loops a few years ago. It has quickly become very popular among music producers due to its simplicity and effectiveness. There is no definitive on what it is actually doing (trade secret!), but it is effectively a compressor, saturation & limiter in one. Its most commonly used on drum tracks.
Sadly it does cost over $40, and with my limited understanding I donāt think itās worth the price just to play around with.
A pattern I found on this list was that at least 3 of the composers used a Yamaha CS-80.
The beast definitely looks the part. I really like the physical sliders and buttons. It was sold at upwards of USD $7000, and coupled with a very complex system with a tough learning curve it wasnāt a commercial success. The CS-80ā²s production ran from 1976 to 1980, quickly replaced by cheaper more accessible options. Nonetheless, it became famous due to its prevalence in pop culture. With the likes of Michael Jackson, Vangelis, Cat Stevens, Toto, Electric Light Orchestra and Daft Punk.
Coming back to the list, the next one to catch my attention is āChariots of Fireā, which I was surprised to find out was composed by Vangelis with a Yamaha CS-80.
The eponymous song itself is interesting, as I only now realized it uses a combination of synthetic and more contemporary instruments. The contrast of the synth and the piano is very apparent, and creates an inspiring atmosphere that is simultaneously calming. I do find it quite amusing that no matter how far I try to go I still end up stumbling upon Vangelis again.
There are others on the list as well, which Iām still busy trying to process yet. However, I understand some basic principles:
1) Experimentation: I need to experiment more with my synths and other instruments.
2) Dirty: I shouldnāt be afraid to create ādirtyā noise, my work so far has been rather clean and sterile.
After much testing and googling, I am confident I have found the issue.
This bad boy here is the real villain!
I had the Echo on one of my Synths set way too high and the sound was echoing into infinity to the point it became white noise. It echoed so hard into another plane of existence that even when the track is not activated/armed, the sound still exists. And the sound persists as long as Ableton is running, regardless of what project you have open.
Bizzare.
Anyway it is a relief that the problem has been resolved.
Itās like machines and software just know when deadlines are coming up, and they love to throw in bizarre problems to keep you awake at night.
As I was overseas for some time, Iāve been working with low quality earphones, and most times none at all. I thought it was an interesting test, seeing just how far I can go without hearing my work.
Anyways, skip ahead to a week before submission. I finally got back to my better headphones. I listened to my work and it sounded pretty good. I then stopped the playback and immediately noticed the sound of air rushing.
I checked and closed every background program and narrowed it down to Ableton. Checked other projects and tried a new project and still same result. The constant sound of air, white noise. And ever since Iāve heard it, I can hear it in playback as well. Itās volume follows the playback.
As you can see in the video above, it sounds like I recorded this with a crappy mic in 2009, but itās actually audio solely from my Ableton. None of the tracks have been activated and still that sound comes through.
A bit of foray into beat-orientated sound. Experimenting with structure and different synths. I especially wanted to attempt theĀ ābass lineā common in similar works.
Ā Iām still contemplating whether to incorporate it with my current project, but Iām unsure how to combine the two cleanly.
In my experience in manning a mixing desk, Iāve been limited to a handful of tools at my disposal. Thatās not to say the mixing desk is lacking, but I guess I now know what I donāt know by learning a DAW.Ā
The tools I usually use are the Compressor, EQ, Gain & HPF, and Gate. There are others too but they are not easily accessible, especially for a live performance. To control the volume of a single channel or the master, the Faders are the primary controls. However in a live situation with varying level of vocal ability and equipment, a lot of unexpected things can happen. The usual is the dreadful ear splitting feedback and unwanted peaks.
This happens almost always with vocals, as they usually start off quite low due to warming up. Then 10 minutes later unexpectedly peak right to the red zone. Either youāre fast with the Faders, or youāre prepared with a Compressor and/or EQ. But either its just me, or these donāt always work either. The EQ is either causing an imbalance, or the Compressor is squeezing the sound.
In Ableton, the Limiter is exactly what I need. It simply a safety net that ensures the output never exceeds a specified level. Allowing the sound to remain relatively untouched until it hits theĀ āceilingā. I have one applied to the Master channel at all times, ensuring the sound is full but not peaking badly.
Working with a DAW in some ways is quite cathartic. I have a lot more control over the sound and can apply more order. Working with Abletonās automation also means that I can literally set and forget, stepping back from the work and letting it run itself.Ā
A Shepard tone, named after Roger Shepard, is a sound consisting of a superposition of sine waves separated by octaves. When played with the bass pitch of the tone moving upward or downward, it is referred to as the Shepard scale. This creates the auditory illusion of a tone that continually ascends or descends in pitch, yet which ultimately seems to get no higher or lower.
This auditory illusion which has been the basis of many great soundtracks, was something that Iāve been searching for an explanation for a long time. The problem was it was hard to explain when I myself never understood it before. How do you google something you donāt even know the name to?Ā
All I knew was that the effect creates a illusion that theĀ āsoundā is rising, but never actually rises any further. My simple understanding of music told me this must be impossible.
It wasnāt until the movie Dunkirk was released in 2017, directed by Christopher Nolan and music composed by Hans Zimmer, did I stumble upon a (now dead) reddit thread explaining this phenomenon. In the soundtrackĀ āThe Moleā, a clear example of the Shepard Scale is used to create this feeling of being overwhelmed with tension, fear and anxiety. But it never reaches a crescendo and very effectively leaves you hanging precariously at the edge of your seat.
There are many other examples of this auditory illusion in the movie, Hans Zimmerās works and beyond. But I aim to capture the same effect, evoking this atmosphere of constant rising tension.
To achieve this, you need at least 3 layers of octaves. All 3 octaves play the same note/tune at the same time, rising similarly. The highest octave starts loud but fades out towards its highest point. The middle octave maintains a constant loud volume. The lowest octave slowly fades in and becomes audible at its highest point. Because you can hear at least 2 layers at the same time, it creates an illusion that it is rising.
To create a virtual environment that is immersive and realistic to ground players before we challenge them, there is nothing better than using something familiar. We have decided to base our virtual environment around the studio space, specifically the meeting room in the Year 1 space. We will use this meeting room specifically as anĀ āanchorā that will act as a familiar space that the player will always start and return to. It will also mostly act as a consistent space amidst whatever chaos we throw at the player.Ā
As a comparison, it would be something similar to Tron: Legacy, when Sam first enters The Grid. He finds the entrance to this virtual world in his fatherās arcade, and when he enters he is transported to a virtual recreation of the same arcade. It isn't until he runs out of the building and sees the vastly different environment does he realize he is not in the same place.
Photos courtesy of James Pane.
The meeting is a relatively small space which is perfect, and has a limited view of the outside which is great too.
An early stage recreation of the room.
The limited outside view helps to create a sense a space and a break from feeling enclosed while requiring little modelling and setup. We only need to model and arrange the cityscape on the outside where the player can actually see. This not only helps to focus our work, but optimizes performance. I aim to place strategic views of the outside world around the level, to maximize space without compromising workload.
In terms of styling, we have decided that each specificĀ āpathā has its own distinct style and theme.
Moodboards arranged by Stuart Hoang and Max Kooji.
To ease players into our virtual environment, the early stages of the game will be visually close to reality but in a utopia style. The path of least resistance will be visually inspired by institute futurism, featuring cold white surfaces and minimal use of color. The path of resistance will be visually inspired by cyberpunk, featuring a clutter of various technology in a chaotic fashion.
Proof of concept: Using the new portal system to create a right turn infinite loop, as part of our Non-Euclidean level design.
Thereās still a noticeable seam that identifies where the portal is, but Stuart believes this could be hidden by using a darker material for the walls.
The performance cost can be seen in the framerate display, even on a relatively high-end desktop Iām only reaching 40-50fps on a relatively simple level. I have a suspicion it could be the number of portals in the same level, but in the final product we might have an even higher number.
New improved portal system, with proper tracking. Weāve opted to use pre-existing code from the web to skip ahead and focus on whatās important. We are now using a flattened hemisphere rather than a flat plane, as you can see in this video it allows for a slight illusion of depth and players can stick their head into it to look downwards or upwards. Thereās a slight stutter that occurs during teleportation that we are still ironing out.
Otherwise, the preservation of momentum feature works very well to further the seamless effect.
This new portal system also plays with scale, if the exit portal is smaller than the entry portal the player scaled down. Itās interesting feature but we probably wonāt use it since it might be too much to take in for the first time.
Well, by being an efficient game designer and performance savvy programmer, I thought implementing level streaming would be the next best thing since slice bread.
Until we opened the project today to find literal chunks of the level missing, lost into the void. You could hear our hearts (and grades) drop. Upon further investigation by the more sensible Stuart, we found that the items in question stillĀ āexistedā, as in they were still referenced in memory. The only things missing were from the level streaming, not the persistent level.
But every single value to them is at 0.
Like theyāve been sucked into a singularity and now exist a null point space in some other pocket dimension. Changing the values yields nothing, so rollback it is. Fortunately Unreal keeps backups of each individual file so we could rollback by replacing files manually.
But I kid you not an hour later it struck again.
So we have the workaround, but the fact the problem persists is worrying.
Work in progress footage of level streaming and portals. The level streaming is not obvious, which is the point, but essentially the 3 rooms are clones and only load in when in view of the player.Ā
As you can see the portals still require a lot of work, I canāt quite figure out the math to match the view correctly. Performance wise its quite consuming, its constantly capturing a high resolution image and then displaying it onto a plane at a very high rate.
To achieve Non-Euclidean level design without having to break open the game engine itself, we have to use portals to connect spaces together where they cannot exist in theirĀ āactualā location in the virtual space.Ā
Basically if there were two adjacent rooms, in the actual virtual environment they are of a distance apart but the portals create a bridge between the two.
This is where the technical hurdles begin.
Right, firstly, portals basically transport the player from one location to the other. This is easy. The difficulty is projecting the environment onto the portal so that the player can see theĀ āother sideā. Although I already know how to capture an image of an environment and project it onto a plane, I need to figure out how to match the imageās perspective with the players view angle. This is complicated, and really doing my head in. Essentially, your view beyond a doorframe shifts depending on where you stand. I have to recreate this where the doorframe is the portal and the view beyond has to shift depending on where the player is looking from.
Another problem is performance cost, constantly capturing an image and having it read onto a plane constantly is taxing. Even more so when the image has to be of the highest quality to appear seamless, resulting in larger file size.
Another option, which is half breaking the engine itself, is to bring the other space to the player by punching a hole out of the current space. This would save on performance, but would require careful planning and a lot of coding to pull off.
In mathematics, non-Euclidean geometry consists of two geometries based on axioms closely related to those specifying Euclidean geometry. As Euclidean geometry lies at the intersection of metric geometry and affine geometry, non-Euclidean geometry arises when either the metric requirement is relaxed, or the parallel postulate is replaced with an alternative one. In the latter case one obtains hyperbolic geometry and elliptic geometry, the traditional non-Euclidean geometries. When the metric requirement is relaxed, then there are affine planes associated with the planar algebras which give rise to kinematic geometries that have also been called non-Euclidean geometry.
Iām going to pretend I understood all that. Fortunately, when it comes to non-euclidean level design in games it is far more practical and tangible. Here are some videos to better explain it in action:
The first two videos utilize the source engine, Portal 2 and Stanley Parable respectively. The third video is the most relevant to us as it utilizes the Unreal Engine. An older generation engine compared to the one we are using (Unreal Engine 4), but nonetheless still relevant.
Non-Euclidean level design essentially is creating impossible spaces only possible in a virtual environment. If I had two doorways that open into a straight hallway that leads out to a courtyard, one would naturally assume the two hallways would be of same length. But with Non-Euclidean level design, I can make one hallway a meter long and the other hallway a kilometer long.
A simpler example that is commonly used in games is the 4 right/left turns. Given a hallway where a player is forced to make right/left turns, players have a innate ability to track distance and build a sense of space. For this example, if they walk a meter before making a right turn, they would expect after 3 turns to hit a wall, the space beyond being the hallway they started in. But with Non-Euclidean design, there is no restriction and there could be a room or even a 4th right turn that loops into infinity.
A more classic example is from the famous M. C. Escher, a Dutch graphic artistwho made mathematically-inspired woodcuts, lithographs, and mezzotints. A famous work of his which demonstrates the chaos of such a reality:
Another classic but more simpler example is the Penrose Stairs, created by Lionel Penrose and his son Roger Penrose:
I got my hands on Antichamber (featured in the third video), and was surprised just how much more relevant it was to our artefact. Not only does it demonstrate the use of Non-Euclidean level design in the Unreal Engine, it also has a style of narrative to guide the player along. Itās not quite on our level ofĀ ābreak the simulationā, but it strongly encourages the player to think radically outside of the box. I personally enjoyed playing through Antichamber as I had to constantlyĀ āunlearnā a lot of knowledge of video games which kept the whole experience fresh and challenging.Ā
