Emax

Hi,
The E3 has sample transposition limits in the OS to ensure samples can not
be moved up or down too far. Its a wider range than on the E2 and E1, but
its still a limitation. For example, at 44.1k you can go down about 1.5
octaves and up just over 3 octaves. With an Emax II you don't have these
limits as the G chip can handle 10 octaves of pitch shift, which is why it
made it into the Proteus module first - which saved E-mu Systems from going
under in 1989. The Protues uses short and low sample rate samples, but the G
chip enables them to be transposed all over the place - thereby the Proteus
doesn't require multi-samples and it can pack lots of sounds into a small
RAM = cheap!

The G chip took around 2 years to design and fabricate, and Dave initially
came up with the ideas back in the early 1980's on the E1 and Drumulator
projects. However custom chip fabrication and costs had to catch up with
Dave, before he could go and build one. Creative bough E-mu Systems in 1993
to get hold of Dave Rossum and the unique G chip "wide transposition"
technology. This technology still figures in the Audigy Sound Cards - albeit
rather more sophisticated! The F chip in the E3 was a very simple chip to
provide over 2x sampling. Of course CD players of the 1980's used the same
technique and went for much higher levels of over sampling. The E chip is
not an exact copy of an EII sample engine, it merely sounds similar. It
works in a different way.

As the E3 has transposition limits, you'll need to push the sample up or
down to these limits and listen for pitch shift distortion. Obviously the
limits stop you hearing anything substantial.....Within these limits the E3
works very well, and certainly Dav'es designs are much better than Ensoniq.

Hope this helps...
Regards
Rob

  >Dave Rossum was able to put a much better sample
  transposition algorithm into the Emax II,...

  Wow, I have never noticed transposition artifacts on the EIII.
  What could I do that would make the artifacts audible?

  I mean, I find the sound of transposed ASR samples much worse.

  Is there more info on this around?


[Non-text portions of this message have been removed]