Yipes, you are going to force me to write a small book here! OK, I'll try to keep it relatively brief. I used a combination of techniques. I tend to first play each track in from the keyboard to get the basic "feel" close. I never copy and paste from track to track - it's important that each instrument track is unique just as each real player in a real band would be unique. After the basic tracks are there I start editing the note data and adding controllers. I edit the note data from the piano roll view. I listen to each track solo and then start adding them together to see how they work as I build the section. The notes are edited for timing and articulations (tongue/slur.) I use my experience as a player to judge articulations. Then I start drawing in the controller data. I usually start with the basic cc#1 data and then do the cc#64 data. Up to this point things go fairly quickly. From here on it gets tedious. I start to refine the cc#1 data, adding much more detailed inflection. Next is the pitchbend data (usually as applies to attacks) and the vibrato data. Both are situational and stylistic by nature. If release effects are needed I add them with cc#15. Once all of the track data is in place then I move on to the mixing stage to see if there are any problem areas to deal with. Any phasing problems (most common during unison lines) can be helped by stage position, volume and pitch refinements to differentiate as much as possible between instruments, and the judicious use of VAR 1 and VAR2. Finally, more advanced mixing techniques can be used (automated EQ, separate section convolution, etc.) to put the finishing touches on the mix. Use excellent recordings of real big bands as a point of comparison to help judge instrumental balances, room simulation contributions, use of articulations, and overall tone quality of instruments.Originally Posted by ronboykin
If all of this sounds time-consuming and difficult . . . well, it can be. Much depends on how far you wish to take it and how much experience you have as a musician to bring to bear on the editing decisions. As with most things, editing tends to get faster the more you do it. Patience is definitely a virtue in the Land of MIDI! These demos took me several days each working on them here and there between my other responsibilities. I have been known to lavish considerable amounts of time on my own musical projects. I once spent over 4 months editing one MIDI harp part for a duet I composed for sampled harp and (real) flute. That was 8 hours of work a day, 7 days a week and the finished piece was just over 10 minutes of music. The upshot: That was over a decade ago and when I hear that mix today I'm still satisfied that I accomplished what I wished to accomplish.
One other piece of advice that has nothing to do with the (all MIDI) demos I've posted but can really help for real music productions: If you have access to excellent real players, it is always worthwhile using them for solos and even mixed into the section work. I know, I know, this is all supposed to be about sampling and the magic of mimicry, and quite good results can be had, given the necessary skills, as in the demos. But, . . . this actually *should* be about creating the best music you can affordto create with all the tools at your disposal. A good tenor saxophonist can knock off a half dozen fantastic solo takes to choose from in less than an hour. Trying to edit a comparable MIDI solo will eat up many, many hours and will (I guarantee) never sound as good. A few "ringers" in the mix can make a big difference. Anyway, just a suggestion if you want to actually use it for a finished product to be released. On the other hand, if you are using it primarily as a compositional feedback tool or to mockup arrangements for promotion then go ahead and edit that MIDI data to your heart's content! Besides, you will learn a great deal as you are forced to analyze the minutia of instrumental inflection in an attempt to create something reasonably convincing.
Tom
Reply With Quote
Bookmarks