The Miracle of Performance Modeling

Performance Modeling Software

As we are sometimes faced with some serious obstacles to generating useful sound sources our focus should shift to the believability of our sequences. Even when listened through a really bad sample set sequences tweaked with performance modeling techniques can sound pretty good. When listened over a good sample set diligently modeled performances can sound very convincing. By taking the time to model the message-based portion of our electronic music production, we can dramatically increase the overall experience of the listener/player.

Performance modeling software tools allow for the generation and/or manipulation of large amounts of MIDI data in musically meaningful ways. These tools can do everything from act as a dynamic processor on MIDI note data to simulating the particularities of performing on specific instruments to generating musical material themselves. They are relatively inexpensive and can be easily worked into even the smallest audio budget.

Most sequencing packages come with a suite of built in and plug-in MIDI editing tools. These can be used to model performance data. For instance Sonar, a multitrack audio and sequencing software package, features various plug-ins used to edit MIDI data. Most of these can be used to transmute raw note data into the type of performance you desire. For instance the "Velocity" MIDI Effect has a randomization parameter I use extensively in my productions to eliminate the static sound sequenced music is capable of displaying.

These prepackaged tools are flexible in that they are not restricted to one or two particular types of instruments, making them cost effective to include in the sequencing software. On the other hand they are far from precise and are not intuitive when applied in specific situations. This is where specialized software tools come into play. With tools offering dedicated functions for modeling performances for one instrument type you are given more control over instrument-specific performance parameters.

Rhythm n' Chords from MusicLab is a performance modeling tool used for crafting rhythm guitar parts via MIDI. There are VST and Cakewalk versions of the plug-in available for purchase from MusicLab's site. You can get some really incredible results using this tool and in a significantly shortened amount of time when compared to hand editing every note in your guitar chords. I'll never forget the project where I spent over two weeks just trying to get the guitar strumming sounding right, just to find out for a $100 I could have completed it in hours with stellar results.

Take a listen to what I came up with in mere minutes using Rn'C. [MP3]

A new release from MusicLab is Slicy Drummer. It allows you to cook up drum loops and patterns from stock libraries of individual drum parts. The interface gives you lots of control over how the drum patterns are played. I haven't checked it out yet but if it does what Rhythm n' Chords does for guitar parts, it is undoubtedly invaluable. Definitely worth downloading the lite-version and seeing what it can do you for you in terms of drum/percussion performance modeling.

NTONYX specializes in developing performance modeling software. Their tools include plug-ins for Cakewalk sequencers as well as stand alone apps. Two tools of particular interest to interactive audio pros are Style Enhancer and Style Morpher. Style Enhancer allows sequence producers to add human elements to instrument parts by exposing and combining MIDI parameters in useful ways. Style Enhancer comes with a variety of predefined musical and instrument styles with additional downloadable styles available for free on the NTONYX homepage.

	Style Enhancer from NTONYX.

Style Enhancer from NTONYX provide electronic music producers with time saving methods for modeling music in various styles.

Style Morpher is a tool near and dear to interactive composer's hearts. Using intelligent music-based algorithms, Style Morpher produces a piece of music based on one of your own in a new style. So if you develop a theme for use in a game and you need to produce jazz, rock and orchestrated versions of it, SM can help you 'port' your music into these different styles quickly and efficiently.

KARMA (Kay Algorithmic Real-time Music Architecture) is a software tool developed by Steven Kay of Karma Lab. KARMA was developed in MAX, a powerful graphical programming language for musical applications. You might recognize the name as belonging to a popular keyboard instrument from Korg. That's because KARMA is in the KARMA so too speak. Korg licensed the KARMA software and stuffed it into a keyboard.

KARMA is… well… rather difficult to explain. It's like an arpeggiator on amphetamines. The system employs GEs or generated effects, which are real time controls used for manipulating performance parameters. Once you get your head around the expansive system you can pull off some pretty intense performance modeling. Heck you can even do things that no one is able to do on an acoustic instrument and all in real-time!

KARMA isn't currently available to the public as software just yet. There are plans to release both MAC and PC versions of the software later this year, though like many highly anticipated software releases the date has been pushed back before. So for those of you so enamored with this sweet technology that you just can't wait… buy the keyboard.

Twiddly Bits is a series of performance modeling tools from Keyfax. Unlike many of its counterparts in the performance modeling arena, the Twiddly Bits series isn't an algorithmic modeling tool. They are actually recorded performances of real musicians playing MIDI instruments. When you hear a Twiddly Bit drum pattern you are hearing the result of a professional drummer working the kit (an electric kit of course, but you still get all the feel that can be crammed into an 8-bit file). You can edit them and manipulate them appropriately for use in your project. This method of performance modeling has no learning curve and is very cost effective.

In Closing…

It's probably a safe bet that we will all be living on Mars feeding space pigeons in our well deserved retirements long before sequenced music will be truly indistinguishable from live human performances (if it ever happens at all). In the meantime, music manufacturers will do their best to provide electronic musicians with the best solutions they can to craft impressive synthesized music and we as composers and engineers should do our best to support and take advantage of their research.

Performance modeling is only one of many considerations that goes into creating great music for interactive media. But when you are dealing with a minute amount of sample resources it becomes the primary one. My suggestion is to look into acquiring some of the above mentioned tools and technologies for use in your next project. I am sure you will be pleasantly surprised by the results.

Resources

International Wind Synthesis Homepage: http://windsynth.org/home.html
Sonar from Cakewalk: http://www.cakewalk.com/
Rhythm n' Chords and Slicy Drummer from Music Lab: http://www.musiclab.com/
Style Enhancer and Style Morpher from NTONYX: http://www.ntonyx.com/
KARMA from Karma Lab: http://www.karmalab.com/
Twidly Bits from Keyfax: http://www.keyfax.com/