As a complete MIDI newcomer in September 1996, I was theoretically aware from the beginning of the variability between soundcards in their treatment of MIDI files, but for economic reasons was limited to a cheap Audio Excel soundcard with only FM synthesis. Not until 6 April 1997 did I have the opportunity of briefly hearing the files through either a wavetable or other FM-only card.
The experience left me in a state of shock, which was the reason for this (originally) hastily-created page! Although my Audio Excel card had all the limitations of FM-only synthesis, I then realised that it was at least better than some in handling polyphony - 32 notes rather than just 16.
For example "So Hard To Stay Close" is quite a busy arrangement, and I had already had to edit out a few relatively unimportant notes to cope with "clipped" and "sticking" notes. However, playing it through a Pine "Schubert" card led to whole chunks of the arrangement simply disappearing. For the time being I cannot suggest any quick way of dealing with that problem, which I am sure will also occur with "Lord, I Believe" and other busier arrangements.
Hearing several arrangements through an AWE32, however, highlighted quite
I had not appreciated a) quite how different the instruments were between soundcards (particularly the treatment of "pad" sounds), and b) how different the relative volumes could be.
For example, on "So Hard To Stay Close" I originally mixed Tenor Sax and Polysynth to get the lead vocal sound I wanted and used Synthstrings 1 for backing vocals. Neither worked at all well through the AWE32 wavetable. Furthermore, the relative volumes between drums, bright acoustic piano and hammond organ were totally different from my intentions. Listening quickly to a few other tracks also demonstrated the wide variation in bass sounds.
So - my suggestions for wavetable (and other) users who are as horrified at the sound through their soundcard as I was, are as follows:
I finally bought an AWE32 in June 1997, and found comparatively little variation from my previous card on the FM synthesis. I couldn't do a detailed check against the wavetable at the time, as my computer wouldn't accept it. In November 1997 I discovered that the problem lay in my Cyrix chip, and changing it has now at last given me access to the wavetable.
Although the wavetable sound is far superior to FM synthesis (particularly for wind instruments and strings), on balance I have decided to continue mixing for FM synthesis. My objective is not to produce a few excellent MIDIs for those with the best hardware, but to provide the maximum music not available elsewhere for the average Internet user.