Questions regarding Creative Music Synth.

On Oct 2, 2:29 am, Richard Dobson <richarddob...@blueyonder.co.uk> wrote 
in http://groups.google.com/group/comp.dsp/msg/11f383f778ac9539 :

 > Radium,
 >
 > Your answers demonstrate why it is almost impossible to help you (any
 > who everyone on these lists is fed up with you),
 > as your knowledge in so
 > many areas is so lacking:

 > Green Xenon [Radium] wrote:

 > >  > Anyway, rather than waste time asking for such things,
 > >  > people have figured out pretty much what those patches are, OPL3
 > >  > emulators and music players are dotted all over the net.

 > > How did they figure out Creative Technology&#4294967295;s
 > > patches unless they were
 > > the ones who designed them?

 > It's called "reverse engineering", you have to be ~really~ smart to do
 > it, and when it comes to things such as FM patches,
 > the procedure is (in
 > the absence of actual path data) to (a) know how FM
 > synthesis works
 > (b) understand the implementation you are using (i.e the OPL3) and
 > (c) change synthesis parameters until the sound matches
 > the one you are
 > trying to copy. So you need a good critical ear too.

 From reading the above paragraph it seems like there is no way to 
figure out the exact mechanisms of the patches unless Creative 
Technology tells you. The 'reverse engineering' seems to involve lots of 
guess work and no exact answers. That's what is frustrating. You want 
something exact but know you can't have it.

 > > Yes but they don&#4294967295;t give info on how Creative Technology
 > > changed the OPL3
 > > synth -- via patches -- prior to soldering it
 > > into their SB16 ISA cards.
 > > I hope this info isn&#4294967295;t necessary yet I fear it is.

 > So you don't even know what a "patch" is? You hardly
 > understand how the
 > things works at all. It is not something they change on the chip prior
 > to soldering (not that sort of chip...), it is FM parameter
 > editing done
 > in software, typically using a "patch editor". Have you never ever
 > created your own patches for your FM synth? Another name would be
 > "preset" or even "voice".  I fear you hardly understand the very
 > technology you are infatuated with! How could you understand the
 > information even if Creative Labs gave it to you?

So the patch is something that is completely software and not in the FM 
chip? I hope so but fear not.

Also...

...I try playing Creative Music Synth [220], through my so called 
'karaoke voice canceller' -- which inverts the phase of one stereo 
channel [right or left] and then combines it the other channel -- which 
results in anything identical in both the left and right channels being 
removed. I get a mono of what was different in the left and right channels.

When I play Creative Music Synth [220] audio through the 
voice-canceller, it sounds more treble, sharper, brighter, warmer, and 
crisper than when I don't use the voice canceller.

I can get the same effect if I use Wavelab [or other audio software] to 
make a stereo recording of the MIDI audio and then invert the phase of 
the one [but not both] of the channels -- left or right &#4294967295; and then I 
convert the stereo file to mono.

Whether I use the voice-canceller or Wavelab to cancel the central 
channel, the effect is the same &#4294967295; the sound is more treble, sharper, 
brighter, warmer, and crisper.

Is it likely the differences I describe [e.g. the differences I hear 
when &#4294967295;voice-canceling&#4294967295; the FM audio] is due to the FM signals 
themselves? I hope not but fear so.

Do these differences occur because that&#4294967295;s how the FM chip was designed? 
Is this because whoever designed the chip, decided for it to work this 
way? I hope not but fear so.

In addition, are these differences I describe due to the patches of 
Creative Technology? I hope not but fear so.

Green Xenon [Radium] wrote:
..
> 
>  From reading the above paragraph it seems like there is no way to 
> figure out the exact mechanisms of the patches unless Creative 
> Technology tells you. The 'reverse engineering' seems to involve lots of 
> guess work and no exact answers. That's what is frustrating. You want 
> something exact but know you can't have it.
> 

The product is supplied with a standard set of preset patches; but these 
can all be modified using the patch editor software they will have 
supplied with the card (nobody would have bought the card unless that 
facility were available). With knowledge and skill, the reverse 
engineering can be virtually exact.

.
> 
> So the patch is something that is completely software and not in the FM 
> chip? I hope so but fear not.
>

I just told you it is software, user-editable.  The operator structure 
in the chip is fixed (just as it is in all other FM synths including the 
DX7 and its descendants), but all the relevant parameters (and there 
will be  a~lot~ of them) are variable from outside. The set of all those 
parameter settings constitutes a "patch".


> Also...
..
> 
> In addition, are these differences I describe due to the patches of 
> Creative Technology? I hope not but fear so.

All these "hopes and fears"  can only be answered by you, doing 
~research~.  For example, by systematically comparing the effects with 
other systems. Which again requires technical knowledge and insights as 
a prerequisite. Nobody else can do that for you. But I suspect the real 
nature of these hopes and fears is that you actually realise you don't 
have the knowledge and skills (yet) to conduct this investigation, or 
understand the equipment you have. Only you can change that. People will 
answer precise technical questions, but will not take on the task of 
educating you in the basics online - not least because there are books, 
websites, Colleges etc that already do all that.


I suggest you learn to use a synthesis language such as Csound, with 
which you can create your own FM synthesiser, and have a go at 
replicating the structure of the OPL3 there. It will be trial and error, 
a lot of the time, but with practice the trials will get more successful 
and the errors more interesting.



Richard Dobson