I would like to create a forum on realistic sounding artificial music (e.g. by sampling) using Software synthesizers based on samples or Hardware Samplers
I really had trouble to get realistic sounds out of my scores (except for when they were actually played by musicians), until I recently (re)discovered the Swedish software package Reason 2.0 (see under "MP3 Files" header).

I guess I am not the only one who is interested in more realistic sounds.
So I will try to compile information on that subject.

Some time ago I started using Reason 2.0 from Propellerheads (Sweden). All mp3 files on the site are made with this package. Till now the most realistic results I got with sampling synthezisers.
I used the sample libraries that go with reason 2.0. (Orchester Bank and NNT19 Sound Bank)
I imported the MIDI files as they are published on the site into Reason and made a mix whichwas exported as WAV file (44100/16). The resulting WAV file was converted by CDex , a free conversion tool, to MP3 (128 Bit) format. These mp3 files are the ones published here.
There is one major drawback to this procedure, which lies in the fact that Reason does not handle tempo variances coming out of MIDI files. So you will hear these differences. The nwc and midi files will have tempo variances the mp3 file will not have these. But the instruments sound really much, much better!
As reason 2.0 can receive and handle MIDI events (including tempo variances) on the fly it should be possible to get around this problem. The output of Reason must therefore be captured real time by a sampling sound card. At some stage I intend to use my Terratec EWS 24/96 to do this. As soon as I will get reasonable /:) results they will be published on this site.