To illustrate the accuracy of Autoscore's recognition engine we are now offering a series of audio demonstrations of the software in action. To experience these demos you must have the ability to download and play MP3 audio files. (You can download free MP3 players from www.winamp.com or www.quicktime.com.)

Each of the demonstration audio files featured below contain 3 different sections marked by a period of brief silence. Each audio file will contain:

I. An original instrumental or vocal performance
II. The resulting real-time MIDI file created by Autoscore
III. The original performance and the Autoscore-created MIDI file played in unison. This will undoubtedly provide the best opportunity to evaluate Autoscore's pitch recognition.

Additionally, with each example we will provide a visual notated example of Autoscore's performance transcription. A word about the notation...

About the demonstrations: These demos are un-edited MIDI recordings made with Autoscore by the Wildcat Canyon Software team. For this reason you can be sure that you are hearing attainable, real-world results of this unique technology. Please listen closely as you should detect minor idiosyncrasies in several of the performances. You will also have the opportunity to hear how Autoscore translates them.

Demonstration 1- Violin
HandelSonata3.mp3

I. The first part of this demonstration features Technical Support Manager Todd Hodges playing a short excerpt from Handel's Sonata No.3 on his violin.
II. Now, listen to the MIDI file generated by Autoscore. We are using a violin MIDI patch from a Soundblaster AWE64 Gold soundcard and, although the attack of these MIDI sounds are quite different from a real violin, you will see that Autoscore tracked the pitch and timing of the performance very well.
III. Here is the performance of part I and the MIDI file from part II played in unison. For this example you will hear the actual violin played out of your left speaker and the MIDI file played from the right speaker. The timing of the parts are also intentionally slightly offset so you can better distinguish between them.

Demonstration 2- Voice
OverTheRainbow.mp3

I. This is a short version of " Over The Rainbow" sung by our National Sales Manager Bill Rhyne.
II. Here is the Autoscore MIDI file that was created in real-time as Bill sang. In this instance we are playing the MIDI file as an acoustic bass patch transposed down an octave.
III. In the final section we will demonstrate how you can use the audio-to-MIDI features of Autoscore in your own music. Again, here is the original audio performance and the MIDI file you just heard played in unison.

Demonstration 3- Guitar
SoulPower.mp3

I. Here is a brisk version of James Brown's "Soul Power" played on guitar by Marketing Manager Sam Thorpe.
II. Now, here is the Autoscore-created MIDI file as a clean guitar patch. It clearly captures the timing nuances of Sam's performance, good and bad.
III. In an effort to demonstrate some of the ways you can utilize Autoscore in your own music, here is Sam's original guitar line doubled by the Autoscore MIDI file played as a percussive organ patch. Some musicians spend a lot of time getting doubled parts this tight. Autoscore can be used for both transcription and music making purposes. Add some drums and this one is funky.

Demonstration 4- Guitar 2
BeerBarrelPolka.mp3

I. In this guitar example, Bill Ryhne plays the melody of an old tune called "Beer Barrel Polka."
II. Here is the resulting MIDI file played as a jazz guitar patch. Notice how well Autoscore handles this melody's quirky rhythms.
III. Here is the original guitar performance played in unison with the MIDI file. Here we set the MIDI file to play as a music box patch. This makes for an interesting sound.

Outtakes

Here are a few extra outtakes that have significant performance issues, but continue to demonstrate Autoscore's accuracy.

Outtake 1 - Voice
Improvmelody.mp3

I. In this example you will hear Todd Hodges improvising a vocal melody. It was late in the day and Todd starts laughing for no apparent reason, which then proceeds to make others laugh. After Todd regains his composure, he finishes improvising the melody.
II. With the exception of the laugh attack, Autoscore does a nice job tracking Todd's melody. The MIDI file is set to play as a tenor saxophone patch.
III. In this section, we convert the MIDI patch to "fantasia" and double it with Todd's original performance. This is another great example of how you can use Autoscore in your recordings and compositions. The MIDI file is set slightly ahead of the original audio performance.

Outtake 2 - Voice
SpontaneousOpera.mp3

It is unclear how to present this one. Sam Thorpe approached the microphone to perform a vocal demo but, after a notion of self-amusement clearly takes over, he ends up sounding a lot like a wounded animal. The performance, which he calls "Spontaneous Opera," ends when he explodes laughing. Amazingly enough, Autoscore tracked his yelping almost exactly. Pay special attention to the laughter as Autoscore extracts its faint musical pitch in perfect timing. In order to convey the incredibly dramatic nature of this performance, we have set his singing to be doubled by an orchestra hit patch. Instead of giving you all three sections, we're sure that this one example is all you'll need. Note: Autoscore's Silence Between Notes feature was used during this demonstration. This means that Autoscore will not output a new note until it can detect an instance of silence. On a couple of instances in this performance you will hear Autoscore only output the first note of Sam's patented atonal legato slides.

For more information about Autoscore, click here.