Digital Acoustic Modelling.....

Using Native Instruments “Reaktor”.

In this new section I’ll briefly show how some of my early simple analogue acoustic models work in the modern digital world. A selection of short MP3 sound clips, fairly similar in nature to the analogue ones, is included here to show how they actually sound and compare. The actual basic digital replicas are presented as free zip download files containing finished working ensembles (Reaktor *.ens files) with a selection of pre-sets.
The models are developed with Native Instruments “Reaktor 4” This software application contains a comprehensive kit of parts for building all kinds of software based audio engineering projects. It turned out to be a good tool for the job but unfortunately it doesn’t allow its projects to be exported as VST plugins, which is a shame. It does allow its projects to be used as a soft synth within the Reaktor shell. This means they can be used as plugins in any DAW project with each knob and switch controllable from it using envelopes. I have done this often. It works well. The only restricting factor is the power of your computer. Soft synths can be very CPU intensive. Reaktor is no exception.
As for the sound, I was both pleased and relived that they all worked well in Reaktor but in the super-clean mathematical world of digital there are some subtle differences. Over the years I have gained more experience working with models so I had less trouble getting there this time. Also I’ve added a couple of extra clips to show off Reaktor’s polyphonic ability where appropriate.  
With the download files below I’ve merely transferred, as close as is reasonably practical within Reaktor, the original simple analogue models. These are not finely honed polished products but they are fully working and I have labelled all the controls on the front panel (see example picture below). Reaktor has a facility for saving a series of pre-sets so I’ve included a short list of these with some instruments. Various sounds are therefore available at the push of a button with plenty of room for your own to be added. I might revise and tidy these patches up later.
In Reaktor, because you are merely processing numbers, there are naturally some significant differences to working with an old style modular analogue synthesiser. In analogue you have control voltages and signal paths and the divide between the two is often quite blurred. In digital you have separate “audio” and “event” processing of the numbers which are sampled at different rates to save CPU time. In Reaktor you also have a main system of pitch which is strongly related to midi data using the keyboard key numbers. This is again different to simple analogue control voltages. Reaktor has a huge range of ready-made modules that equate fairly closely to standard analogue synthesiser modules, which is a great time saver. You can also play in polyphony which is a revelation! This is a another great time saver. The main downside is there wasn’t a ready-made “Reaktor” delay module similar to my analogue modelling one so I had to create it as a Macro. I created some other useful Macros too. The front panel looks a bit analogue and will have some familiar controls and some new ones. I’ve tried to label them sensibly but bare in mind they are no more than a demo research patches not finished commercial products.   
Reaktor panel: String patch...
Shown above, you can see the 3 main sections: pick, string resonator and simple body resonator. “Transient” provides the kick on releasing the ‘string’. “Noise” is the noise content of the pick. The above controls set how soft or hard thick or thin it is. In the “string resonator” section, the filter is the low-pass string softening filter, “filt mod” tracks it to the note pitch. “Stretch tuning” compensates a little for the added delay of the filter which is fixed. “Transient bend” bends the string with the envelope of the pick controls.  
[SOUND EXAMPLE 15: String development...  MP3 (128kbit/sec)]     [SOUND EXAMPLE 16: Eastern stringed instrument...  MP3 (128kbit/sec)]
[SOUND EXAMPLE 17: Bass guitar 1...  MP3 (128kbit/sec)]                [SOUND EXAMPLE 18: Bass guitar 2...  MP3 (128kbit/sec)]
[SOUND EXAMPLE 19: Bas guitar 3...  MP3 (128kbit/sec)]                  [SOUND EXAMPLE 20: Polyphonic nylon strings...  MP3 (128kbit/sec)]
Opposite is the panel for the simple two delay plate gong. The three delay is similar but with a lot more controls. Its two delay modules are wired in parallel and triggered by the same pick module as the string. There’s an overall feedback control on the far right to control the ring length. Beware it can take off into distortion, which is same the problem I had to overcome with my early analogue models. There’s so many different gong and bell sounds you can make with this so, as with the analogue demo, I’ve created a group of sounds in sound clip 23 to show it off properly.
(Roll mouse over for bigger view).
[SOUND EXAMPLE 24: Trumpet development...  MP3 (128kbit/sec)]       [SOUND EXAMPLE 25: Trumpet glissando... MP3 (128kbit/sec)]
[SOUND EXAMPLE 23: Big group of gongs and bells ...  MP3 (128kbit/sec)]   
As with the analogue model a simple sine wave oscillator drives the lips module and the tube resonator which has to be in tune or a harmonic relation for it to work. In Reaktor tuning is in midi notation so to change by an octave that’s = or - 12. “Severity” is how much the lips affects the harmonic generation round the tube circuit. “Tube res” is how live or echoey the tube is. “Satlevel” and “Satfeed” control how the the resonance is stopped from taking off and overloading everything. Far right the button switches in the glissandi demo. Press a keyboard key and turn the knob.
[SOUND EXAMPLE 26: Tuba...  MP3 (128kbit/sec)]                                  [SOUND EXAMPLE 27: Polyphonic brass group... MP3 (128kbit/sec)]
[SOUND EXAMPLE 21: Plate gong 1...  MP3 (128kbit/sec)]              [SOUND EXAMPLE 22: Plate gong 2... MP3 (128kbit/sec)]
|  Home  |   Previous Page  |  Next Page  |