Designing Cinematic Synthetic Drums

Welcome to the first post of the official Impact Soundworks blog! We hope to give you insight into the process of designing, recording, and editing sample libraries and virtual instruments.

 

logo-300x300-csd
   
The idea for the Cinematic Synthetic Drums (CSD) library was perhaps originally inspired by the soundtrack to Tron Legacy by Daft Punk. As an electronic musician myself, I found the blend of grand orchestral elements with highly electronic synths and drums to be very interesting and effective. What pushed me to actually construct and release this collection was the free, high-quality work of people like Blake Robinson (Blakus) on VI-Control who deserve great respect for their contributions to the TV/game/film music community!

After seeing the great response to CSD, I decided to take the concept and run with it for a full-scale library with hundreds of sounds, more variety, and a sophisticated UI. There are now multiple talented sound designers working on the project instead of just me, while Iain Morland is assisting with a number of editing and programming tasks. So, what goes into making a single sound? Our workflow looks something like this…

Synthesis -> Layering -> Processing -> Editing -> Programming & Integration
The first stage of the process, Synthesis is of crucial importance. This is where most of the magic happens, and it’s also quite challenging given the complexity of synthesizing percussion in particular. To craft a percussive sound, there must be a specific focus on the envelope or shape of the sound – its amplitude, pitch, and any filter and/or modulator.

For example, a standard kick sound may begin as a humble sine wave – easy enough – but it is not a static sound. There is usually a sharp spike in pitch at the beginning of the sound followed by a rapid drop along a carefully-tweaked curve, followed by a rapid decay in volume. Another oscillator (perhaps noise) might add an extra bit of buzz and ‘click’, with a third envelope controlling the cutoff of a resonant lowpass filter. The second oscillator may have its own volume envelope as well, cutting it shorter than the pitch (fundamental) envelope of the first oscillator.

These considerations are only the beginning. Through additive and FM synthesis, one can create much more complex interactions between multiple oscillators and modulators, in which case one must meticulously balance harmonics and overtones. A little too much of a spike can quickly turn an FM sound into pure noise… though that may be useful for something later 🙂 Granular and wavetable synthesis introduce many more possibilities in creating evolving percussive sounds like crashes, or unusual overtone series for exotic snares and thips.

Below: A fairly basic, no-frills kick patch in Zebra 2 (click for full-size).
Layering comes next – but not all the time. In some cases, simplicity and minimalism makes for a more effective sound. In other situations, the use of multiple patches on multiple synthesizers creates a whole that is greater than the sum of its parts. For example, we might use FM synthesis to generate a highpassed series of overtones while an analog synth handles the thunderous low-end with a lowpassed, rezzy triangle. Layering is also particularly useful with some of our complex, dubstep-inspired bass sounds which often have violent and chaotic middle layers, but rock-solid subs.

After the skeleton of the sound has been created through synthesis and layering, we move on to Processing and really get into the nitty-gritty of sound sculpting. Processing encompasses tools like EQ, compression, limiting, saturation, distorting, amping (or re-amping), spatial FX and much more. While not every sound necessarily uses layering, every sound in CSD Pro uses quite a bit of processing to achieve an end result that is immediately usable. I have personally purchased many drum libraries that have fairly ‘flat’ sounds which can be brought to life with plenty of multiband limiting, tape distortion, and re-rendering, but that was definitely not our goal here.

Below: A relatively minimal palette of FX for an aggressive drum sound.

   
From the screenshot you can see that we use quite a variety of tools, from renowned commercial plugins to some very handy free processors that we find to be competitive with even the very best commercial options. However, we’ve also taken processing outside the box with hardware compression / limiting, such as with the finicky and extremely temperamental Valley People Dyna-mite stereo limiter. Though we are keen on making sure every sound is polished to a mirror shine with a custom signal chain, we’re also avoiding a common annoyance among many other drum libraries by generally leaving out spatial FX. Nothing is worse than finding a great impact tone only to find it has baked in cathedral reverb!

At this point, the sound asset has been created (with possibly a slew of variations) and the Editing work begins. Percussive sounds tend to be easier to edit than many other instrument types due to the lack of a human ‘attack’ such as a thumb or bow gently touching the string or the tonguing of a brass instrument. That is not to say that there are no decisions to be made; sloppy editing can truncate a juicy transient sound, while over-editing multiple variations of a sound can make them more sterile and lifeless (ironic for electronic noises, but it really is true!) Meanwhile, sustained sounds such as basses cannot be edited automatically – proper loop points can only be made through very careful work and lots of listening.

Below: Comparing three variations of the same snare. Note the difference in transient timing and timbre.

  
Once sounds have been properly edited, it is time for Programming & Integration. As with CSD, the Pro version will of course include 24-bit WAV files that can be used in any sampler, or dropped into a project as-is. However, we are also creating an advanced, intuitive and useful Kontakt 5 patch that we think will be very useful and time-saving. To do this, our first task is setting up individual samples (WAV files) into groups, and then mapping them across the keyboard in zones. We plan on segmenting the library into two patches: tonal material and percussion. The percussive patch will be playable drum kit style (one key = one drum sound, velocity sensitive) while the tonal patch will have layerable sounds spread across the keyboard with deeper editing options on a per-sound basis.

Kontakt scripting is an arcane art unto itself – though quite powerful, it is also severely lacking in documentation, often creating much consternation on the part of developers. That said, the feeling of compiling and applying thousands of lines of script and not getting an error message is pretty magical and makes it all worthwhile! As far as our script design, our goal is keep the needs of the modern composer in mind (it’s right in our slogan!) and that means creating tools that reduce, rather than add, complexity. We want to simplify and streamline things like post-processing, filter/EQ and modulation, not just by making those controls available on the UI but by creating sub-presets WITHIN a patch that let you create unique combinations of variations instantly, with a minimum of effort.

Below: Under the hood of CSD – download and view/edit the script for yourself!

    
In the coming weeks / months, we hope to show you even more from CSD Pro, from demo songs to individual patches, video and more. Thank you for reading, and please leave your comments below or on our Facebook page!

– Andrew Aversa