Hammond organ clones are really specialized musical synthesizers, not all that internally different from a Minimoog or a Prophet 5. In the clones, however, the power of the electronics is used to recreate sounds with extreme accuracy, rather than to produce a wide variety of timbres. The sonic quality of clone instruments has benefited from the advancements made in "classic" synthesizers, and so clone designs have roughly followed the progress of their synthesizer counterparts.

Analog Synthesis

Early electronic organs, including later-generation Hammonds, as well as all Farfisa, Vox, Wurlitzer, Lowery and Yamaha organss all used analog circuitry to create their sounds. The basic approach to analog synthesis is, "Get me something electronic that sounds a lot like the thing I'm hearing. If it's not a perfect copy, we'll tweak the circuit to get it closer." Analog synthesis is often called "subtractive synthesis," because the available analog waveforms tend to be simple shapes like sines, pulses, and ramps, so the new circuit is rarely a perfect clone of the original, and unwanted sonic artifacts must be subtracted from the sound using filtering. The same approach was used in electronic synthesizers like the Jupiter 8, Minimoog, ARP Oddessey, and Prophet 5.

In the Hammond clone world, examples of organs which use analog synthesis are the Hammond X-2/X-5, the original KORG BX-3/CX-3, and the Viscount D9/Oberheim OB-3. For the most part, Hammond clones designed before 1990 always used analog methods.

NOTE: The Voce and early KORG products use specialized ICs to generate their raw tones, rather than the earlier top-octave synthesis (TOS) and subsequent filtering designs used by Vox and others, but the approach is still clearly in the analog domain, and requires additional circuits for key click, vibrato/chorus, etc.

Analog synthesis has the advantage of sounding very lifelike, when compared to some digital designs, with a sound that is often described as "fat" or "warm." It's main disadvantage is that because available analog waveforms tend to be simple, while the sounds they are simulating tend to be complex, it takes a lot of analog circuitry to accurately mimic sounds, especially during transients like key strikes.

Sampling or Wavetable Synthesis

The sampling approach says, "Let's record a copy of a 'perfect' example of the original sound we want, and whenever we want to hear it, we'll just play back the recording." This approach was first used on instruments with complex waveforms like violin and piano, and for percussion instruments, since both instrument families are difficult to simulate with analog circuits. Examples of sampled/wavetable synthesizers are the Mirage, the KORG M3, and the Quadrasynth family.

This method also adapts well to the Hammond organ sound, because it allows for the capture of subtle by-products of the sound, like leakage and key click, that are part of the "signature" sound of the organ. One design difficulty inherent in sampling is that shifting the pitch of a recorded sound alters the timing of timbre and volume changes in that sound, sometimes causing the result to seem unnatural. The solution to this problem is to sample every note, or as many notes as memory limitations will allow. High-quality organ clones use multiple samples recorded at different pitches across the instrument's range, so that the organ doesn't have to shift the pitch of any one sample too much during playback.

Examples of Hammond clones which use sampling are the E-mu B-3, the Hammond XB- and XK-series, and the Peavey Spectrum Organ.

Sampling is capable of creating very realistic reproductions of instruments, provided enough sample memory (usually, ROM) is available. It main disadvantage is an inability to provide random-ness and note-to-note interaction in the sound.

Modeling Synthesis

Modeling takes a different approach: "Look, there's no such thing as a 'perfect' sample of the sound, and even if there was, it would only happen once in a performance. Instead of copying the sound, let's find out how the instrument works, then use the power of computing to simulate the sound at any given time. This method has only recently been feasible, thanks to breakthroughs in the processing power of programmable integrated circuits, mostly for the personal computer industry. Modeling works well in synthesis because most musical instruments are mechanical (or analog electronic) in nature, and so must obey certain laws of physics, whose workings can be mathematically described.

Does modeling really work? Actually, rather well. All digital audio and video methods use a similar approach: There is no "audio" or "video" on the media; just digital information for reproducing the original analog material at any particular point in the performance. Yet to the listener (or viewer), it appears to be completely natural. Modeling just takes this approach one step further, and creates the digital information on the fly when needed.

Examples of Hammond clones which use modeling are the new Korg BX-3/CX-3, the Native Instruments B-4, the Nord Electro, the Roland VK-7/VK-77/VK-8, and the Voce V3/V5.