Making Musician Rehearsal tapes via Simulated Instruments

Through the use of Creative's SoundFont technology, you can get a pretty realistic sounding rehearsal tape for free, depending on the type of music you're working on. Here's a link to tons of different sounfont sites, with the key being the larger ones are usually better, including a few full orchestral sound fonts to let you play out full scores.

How I usually go from score to audio cd is as follows.

First, a midi is needed (I can't play piano fast or well enough to encode this myself for the long pieces I usually do this for). The Silvis Woodshed has been great for many classical pieces, but most of the time a simple google search for "[piece name here] AND MIDI" works.

Second, I use Noteworthy Composer to import the midis, saving them each as NWC files, so I can figure out which midi channel goes with each voice part. If I picked the right midi, each voice part is on a different channel. If not, go back to step 1.

Third, windows explorer, copy the set of midis into four folders, in folders: NWC\S, NWC\A, NWC\T, NWC\B for each of the voice parts needed.

Fourth, go into Noteworthy Composer (NWC), and for each file in each voice part, adjust the volumes. I usually use 60 for orchestral/accompaniment, 75 for the other voice parts, and leave 127 for the current folder's voice part. Be careful here -- most imported midis tie the first channel to the first staff AND the rhythm staff.

Fifth, map the staff's channels to the appropriate soundfont via Creative's SoundBank manager software. Audigy series cards all come with this software to my knowledge.  Be awware that you'll want to share system ram with the sound card so you'll have enough to load in some of the larger soundfonts, and increase the ram size to something large, like 384 or 512.

Sixth, (after quite a lot of manual work), this is the painstaking part.  I open up Adobe Audition (previously Syntrillium's Cool Edit), and start a new document. Open up the first NWC file, press record in Audition, and Play in NWC, repeat ad nauseum for each midi file, for each voice part. Save in WAV\S, etc.

Seventh, open up each document, & normalize all the audio.

Eighth, burn each audio cd master, adding in CD-Text (very cool for display on players that support it). The end result is a tape where the musician can learn their part, with it being the dominant notes played, while still hearing the harmonies behind them.

Ninth, obviously the duplication process where each cd is documented.

It's my hope to create a c# little app to automate steps 4, 6, and 7, but for the 3 times or so a year I do this, it may not be worth my time.