In this blog post we have the pleasure to show off our first original song written for Velvet Remedy and we elaborate a bit on our adaptive in-game music engine which we are implementing and for which we are composing. Also, don't miss our call for singers!
Velvet Remedy Songs
Inevitably, when the time is right, there will be dubstep. "There is always a DJ Pon3" and we all know how the original DJ Pon3 handled her dishes. However, the wasteland is way too big just for a single subculture.
To be consistent with the style and tone of the original Fallout game series, the choice of Velvet’s genre was clear – the 1940s big band swing music. To pay a tribute to Velvet Remedy herself, all of her songs are and will be original compositions.
Here is Step Around, the first song that we wrote and produced for Velvet Remedy:
It’s only a minute-long sample and the voice you hear obviously is NOT Velvet Remedy. We want to keep the spoilers to a minimum. In our adaptation of the Fallout Equestria story, Step Around is a happy song written before the events of the war. Not far down our production line are songs Mane Squeeze about an imperfect relationship of two ponies and You Saved My Life and It's Your Problem about the infamous zebra logic.
Call for Singers
Attention Ladies! (We’ll get to you later, guys. Just hang tight!) Think you’ve got voice that would make the wasteland a little brighter? Want it to be heard in our game? Great! We want to hear from you. There are two options for you to submit:
(1) Here is the karaoke version of Step Around. Show us what you’ve got. Just use the Wasteland Wailer’s dropbox page on SoundCloud or simply send an e-mail with attachment to firstname.lastname@example.org. You will find the lyrics in the description.
We’ve also included a version a third down for mezzos if that’s your jam.
Important Note: Take your time learning the piece, fine tune your takes, and make us fall in love.
Update: We have filled in the role of Velvet's singing voice, however we will require other singing voices as well, so feel free to send your takes to us anyway!
(2) Submit or link us to a recording, no more than 16 bars, of any piece to show us the extent and skill of your vocal range. We’re looking for a wide variety of styles and voice types, so don’t be afraid to show off.
(3) Point us to your existing online profile via e-mail.
Please keep in mind that as much as we love the multinational diversity of our team (for example, a good part of this post was written by a Czech), we honor the authenticity of the American Standard accent. Unless the script says otherwise, it is a requirement on all our VAs. Thank you for understanding.
Adaptive Music for Ever-changing Wasteland
TL;DR: Our music engine observes what is happening in the game and chooses what parts of the soundtrack will be played next. The segments are produced in a way that will yield a seamless playback.
It is a common practice in game development to associate certain music loops with specific locations. As the player moves from place to place (or level to level), different music tracks, often looped, are being played. Good music must also deliver appropriate emotions, moods, and expectations relevant to the current point in the game. Since all of this constantly changes during gameplay, the music should keep up. We need slow-moving strings and light percussions when the player can afford to relax and explore but we want an instant transition to fast-paced full-orchestra dramatic music when the player is under attack.
Music composed and played back in the way that allows seamless transitions between moods and locations is often referred to as the adaptive music.
We took inspiration from a number of existing games and decided to implement our own adaptive music engine for the in-game background. Our engine is based on a number of techniques how the adaptive music is often implemented: Specific locations in the game correspond to music scenes. Going from one location triggers a switch of the scenes via a transition pattern. The "topology" of our soundtrack for Stable 2 looks a bit like this:
For an uninterrupted flow of the music it is important that the transitions from scene to scene are seamless. We illustrate this on a music from a game called Quido (used with permission). Let's imagine a player who walks from room 1 to room 2 which requires a transition from music scene 1 to music scene 2. Play back the following example and read these comments:
- Stopping a track when the player walks into a different room, and starting another track is a no-go as it disrupts the pace of the game as well as the rhythm of the song.
- A simple cross-fade is lame. It makes the rhythm of the two scenes clash.
- The right way to do this is to synchronize the beat of the two loops and start the second one upon the end of the next measure within the first
- A better way yet is to have a dedicated transition pattern leading the music from one location to another.
Each scene is composed of a number of patterns. A pattern is a few beats or measures long. Every pattern can be linked by an arrow to one or more patterns from the same scene. A link indicates that the music may continue by the pattern to which the arrow points.
We show an example of two scenes in the figure below. Each box or chevron represents one pattern. The chevron-shaped patterns are the transitions from scene to scene. The patterns are chosen randomly or according to defined rules. For example, in a multiplayer strategy game the background music can play the anthem of an army that's currently winning in a battle.
A pattern is represented as a number of synchronous tracks which play together. But they don't have to: Each track can be assigned a condition that must be met in order that its playback is enabled. In the following example we use four tracks: Bass & Beat, Chords, Happy, and Very Happy. Very Happy and Happy are mutually exclusive. They both are melodic tracks but they contain slightly differing melodies and so they can't sound at the same time.
Overall, our design has several advantages.
- Despite the complex structure of the music, the individual tracks can be easily prepared using the standard digital audio workstation software. We don't have to worry about the compatibility of various file formats, VSTs, etc.
- The reverberations won't get cut off at the end of a pattern, which would be necessary otherwise in order to keep the timing correct.
- There is no prescription of the number of parallel tracks or the length of the patterns. Rapidly changing soundscapes require quick responses to events in the game (short patterns) while it is enough for calm scenes to be composed of longer patterns.
- This structure can handle changes of tempo without breaking the flow of the music.
At this point, we have the basic soundtrack for four locations within the Stable 2, and two more for New Appleloosa. Now we are working on the "dynamizing" of these tracks, i.e. we are cutting the tracks into patterns, we're making alternate versions of these patterns for various emotion states, and we're designing the transition patterns for smooth changes of the scenes. We are looking forward to showing you more within some of our next blog posts. Stay tuned!
(P.S.: The music in the examples are using snippets from Quido, an experimental videogame soundtrack. Its author is one person on the music team and he gave his permission to use it. Other than that, this project is not related to our game in any way.)