Having been greatly impressed by Ken Elhardt's imitative synthesis efforts, I set out to try some myself. Combining a few sawtooths from VCO's and a pair of Zeroscillators doing their FM thing, a low-pass filter and a Moog 907 fixed filter bank, a couple of VCA's and a few EG's, I created some samples, comparing them all along to reference samples of, in this case, brass instruments - mostly a trumpet ensemble sound from Kontakt. I used Blue Cat's free spectrum analyzer - highly recommended for the price! In any case, I discovered that my samples needed massive EQ work to get close. In addition to the fixed filter bank, I used two sets of parametric eq's within Cubase - the native EQ and a UAD plug-in (the new Helios Type 69 - very nice!). I found the results quite pleasing. So, why is it that (performance) synthesizers have traditionally only had one measly filter? Why essentially no eq at all? Why even within the modular realm has there been only limited eq capability? Honestly, traditional fixed filter banks were quite limited in scope. I think there's a real opportunity here for a clever new module that allows severe (big boost/cut) and detailed (many frequencies) EQ capability. How about with frequencies under voltage control to shift the whole shebang up/down the keyboard? Thoughts for a digital implementation in MOTM 2.0? With memory for storage of different "instrument" settings?
Message
Imitative Synthesis and Implications for Hardware
2007-04-19 by rogerpellegrini
Attachments
- No local attachments were found for this message.