[linux-audio-user] Hammond organ?

John Check j4strngs at bitless.net
Mon Mar 21 18:23:43 EST 2005

On Monday 21 March 2005 06:21 am, tim hall wrote:
> Last Friday 18 March 2005 12:25, Randy Kramer was like:
> > Sorry, I know I'm replying to the wrong post, and I might even be in the
> > wrong thread--somebody posted about possibly not being able to achieve
> > the desired sound (for the Hammond organ) without ~"heated elements
> > moving massive amounts of air" (or something similar.
> That was probably me.
> > I'm curious (and trying to do a boundary check)--has that person (or
> > anyone else) heard a satisfactory reproduction of the desired sound on
> > any CD? (If so, it would seem to me the desired sound can be achieved
> > digitally, it's just ;-) a matter of finding the right waveform(s).)

I have a set of Joey DeFrancesco EPS16+ sampler disks someplace that are very 
sweet. These samplers maxed out at IIRC 2Mb RAM.

> One of the problems with the hammond is that you would need a separate
> sample for each key, to get the proper vibrato of each tone-wheel and a
> good leslie is hard to fake. I'm not saying it's utterly impossible, but it
> would take a sample bank as finely engineered as the original instrument.

That's why it's as important to spec what a patch should do before one starts
as it is to spec a piece of software.
The physical aspects of the instrument one is producing need to be accounted 
for. Am "I" trying to reproduce certain idiosyncracies (features) of a source 
or eliminate them (bugs)? 
 Do I want the pedal thuds in my piano?
 Do I want the motor noise in my leslie?

IMO the "right" way to do things is to break down the instrument one is trying 
to reproduce and then use the available synth architecture to reproduce as 
many of the dynamic qualities of a timbre as possible. The samples themselves 
should be as close to a static nature as possible, i.e. the tone wheels of a 
hammond always output the same waveform. It's the drawbars that dictate the 
waveform that gets sent to the output and there is an infinite combination of 
positions, so trying to sample every possible combination is brain dead.

If one just records a particular hammond & leslie sound and stuffs it in one 
layer of sampler, it won't be anything more than that one patch. If one 
reproduces the tone generators and emulates the construction of the 
instrument, and it's signal chain (should one be necessary for the sound) 
then it's possible to approach the real deal.

> And no, I've never heard the hammond transfered to CD without losing some
> of its power, although the tone of a real instrument generally does shine
> through.

Well, recorded music cannot sound like live music unless perhaps somebody 
discovered the ideal of microphone technique. Consider that in this 
particular case, the output transducer has both a dynamic shape and 
dispersion pattern.

> I love digital for editing and general cleanness etc. But I think there are
> certain analogue sounds that need tape, valves (tubes), strings, skins
> ambient spaces and human interference to get just right. The beauty of a
> digital system is that you only need to insert the analogue where it is
> needed and it's easier to cut down on unwanted noise. I do understand the
> desire to have a B3 on your laptop and if one existed I would almost

horgand is pretty good. 

> certainly use it myself. You could always route it though a valve
> compressor just to warm it up. :-) In terms of useability, Horgand and
> ZynAddSubFX already have some nice organ patches, which does me fine for
> now.

No arguments there

> Anyway, this really is just my uneducated opinion.
> cheers,
> tim hall
> http://glastonburymusic.org.uk

More information about the Linux-audio-user mailing list