michael diemer
Orchestration is the one thing that nobody is born knowing how to do. You can study books and internalize rules, and above all, listen to the great masters; but only hearing what you have done, with as much realism as possible, is going to give to give you the necessary feedback.
My work flow is from the other end (beginning in the notation product), but we converge on the point that high-quality playback is really important for today's composer. I have two colleagues who are both good arrangers and both work with Finale a lot. In the span of 2 weeks, one of them advised me that it was best to play back in Finale's simply MIDI playback because that was the only way you could really hear the harmonies well. The other one was even more extreme. He said that he was most successful playing everything back as piano parts only.
I would never dream of doing that. That may make is easy to head bad notes (notes that are simply mistakes. But voicings must be carefully crafted to get the most musical effect. Those simple playback modes don't have a chance of helping us hear the nuances of how instruments blend and how voices work together (or not). Using the simply MIDI playback, one risks producing a poor quality orchestration that may require multiple readings by the ensemble and multiple revision before the arrangement is suitable for public presentation.
Indeed, you actually can get very realistic playback directly from Finale by using good VST libraries, and I do this every time. This allows you to hear the timbres and the interaction of overtones. Often I can produce a score that requires no revision at all -- or only minor revisions after the first reading.
In other words, it is about time and quality. Isn't everything?
The problem with the notation programs is not that they are unable to use VSTs. They do use the VSTi products if you are lucky, and that can do a good job on the viocings. But there are many other nuances that are not so successful directly within the notation programs. For example, swing feel or breathing/bowing just may not sound quite right without some detailed adjustments to the MIDI. And of course, some additional DAW-style processing can help tremendously. For example, it would often be desirable to use a ducking compression technique if a piece features a solo instrument or voice. And I have often wished I could have automation of track volume/velocity within the notation program in order to bring out the important lines a little more than is indicated by the printed dynamic markings.
Basically, I see this as two sides of the same coin. One world (DAW and notation), not two separate domains. That's the kind of flexibility we need regardless of which platform we use as our base for composition.