Going For The Obvious: E.ORGAN 1

Mastering a new synthesizer is a lot like mastering a new video game. There are all kinds of twists and turns along the way. The bad guys get tougher, and there are usually a lot more of them as you wind your way through whatever virtual world you happen to be in. There are dangers, pitfalls, and careless mistakes that put you right back where you started.

A divergence between synths and video games, though, is in video games each level of mastery can get increasingly complex and difficult. On a synthesizer, a program might arise that is deceptively simple. It feels like you're going back and playing an earlier, easier level of the game. Video RPGs are not often played by going BACK a level, though this is certainly an option. Perhaps you missed one or more objectives. Perhaps there's a magic gold coin that you MUST have to complete another level farther down the road. Or you just want extra points. But typically you aren't working backwards.

On synthesizers, you're going to write easy patches from time to time. But just as with the video game, the level may appear easy, or you made it through on pure luck ONE TIME. What you have to beware is the unexpected demon lurking in the shadows.

That's the discussion for today: Going for the obvious, or when patches look easy.

Let's take a look at E.Organ 1. This is modeled on the Hammon B3 organ. Before we proceed, what is a B3? At the most basic level, it's a tonewheel organ. Everyone knows that, right? Just a bunch of sine waves, right? That's going for the obvious--but it's a deceptively simple process. Saying the B3 was "just a tonewheel organ" is like saying the 1980 Firebird TransAm was "just a car." There are a lot of keyboards out there, just like there are a lot of cars. The difference isn't just the body, it's what's under the hood.

What were some of the things that made the B3 special? Drawbars. Tube circuitry. Key click. Percussion. Chorus/vibrato. Leslie cab with horn rotation adjustable in realtime. Sure. But that's just what you found on the dashboard!

B3's could be really noisy, especially the older they got. What about tonewheel bleedthrough? Aged tubes? What, you really think 50 year old tubes are going to sound the same as the day the B3 rolled out of the factory? I don't know how else to explain it, but electro-mechanical instruments are BAD quirky. If the B3 isn't bad enough, what about those first analog synthesizers that NEVER stay in tune?

I'm not trying to overthink a simple patch--just trying to give a little perspective.

Let me sit down at my DX and load up FM8 on my iBook. I've got E.Organ 1 right in front of me. Oh, and before I forget: I'm looking at the original DX7 presets, not the mkII. I tried loading the originals on the mkII and it was epic fail. The DX7II was an improvement over its predecessor, not a copy with extra features. FM8 does a better job as an analysis tool, but the sound is WAY clean compared to the DX7. So any analysis using advanced tools has to be done with a grain of salt.

And by the way, that reminds me: FM8 really is fairly advanced for what I'm trying to do. As I've said before, Absynth goes a long way to completing the big picture (I don't have it at the moment). But the best way to get a true DX7 timbre analysis is to sample the waveforms while turning various operators on and off.

So anyway, the E.Organ 1. We're using 6 operators in one of the last algorithms on the synth. This means there are NO modulator waves. Yippee!!!!! OK, let's load up two Thors, no FM amount, and copy the harmonic ratios; adjust EGs to get the right effects.

NO.

WRONG.

We're only going to use ONE Thor.

WHAT??? Are you kidding me?

The obvious isn't so obvious anymore, is it? Let's review: What's the goal, a faithful replication of the sound or a close approximation? We don't have to stay true to the original, we're just using it as a model and getting as close as we can. Let's look at patch in FM8 again.

Op A has fast attack/decay time and a harmonic ratio of 3; in other words, an octave+5th. This is our percussion, which is not the most important part of the patch.

Op B, ratio 1, is part of the core sound. So is Op C (ratio 0.5), Op D (1.5), Op E (1), and Op F (0.5). All values are rounded, of course, and do not take chorus detuning into account.

Notice what we have in common: Ops B and E (1), C and F (0.5). We no longer have need for chorus detuning, so we can condense these into two oscillators. We also have Op D, which is basically an octave+5th above our lowest (0.5) operator. Remember, Thor doesn't do anything less than whole numbers, so we need to program twice the value of these essential operators: 3:2:1. Why is this so interesting? This is the additive basis for a saw wave! So let start with that.

In Thor, make sure oscillator 1 is active. By default, Thor's FM oscillators are set close to a saw wave, or at least the FM equivalent. Bring the FM amount down to 4 or 5.

Now, our organ isn't a perfect analog saw wave; remember the B3 was a tonewheel additive instrument using drawbars to control the timbre. On the DX7 patch, all overtones (or DX7 virtual drawbars) have equal treatment. Our job is nearly finished with just one oscillator, but why stop there? Let's active osc. 2 and boost the third partial.

We have some options here. We can use a plain, boring sine wave and set the carrier ratio to 3. This does work, but it's not a very rich, full sound. We could change it to a saw wave by setting the mod ratio to 3 and kick the FM amount to 4. Fine, but we seem to be missing the second partial. Set the carrier to 2 and leave the rest of the settings as-is. What do you think?

So we just rebuilt the DX7 preset E.Organ 1 with only two FM pairs in Thor! Amazingly efficient, right? But we're still not done. Remember the B3 percussion sound in Op A? Activate Osc. 3. Up to this point, we've been setting our ratios twice as high as they should be. So for accuracy, they will all need to be tuned down an octave. Osc. 3 we can leave alone.

Now, the DX7 uses a sine wave for the percussion effect. It's a very short town, however, so there's no real reason why it HAS to remain a sine wave. I took some creative license here and programmed carrier 3 (as you'd expect) and mod 2. It's just a little brighter than what you'd expect, but it's not bad! FM amount is 2.

We're dealing with very small, subtle settings. A B3 is much more complex than what you'd expect, but it's not brutally strange.

Now for the fine tuning: the B3 is a simple on/off switch, so set the amp EG accordingly. Route the mod EG to Osc 3 level in the mixer, turn the Osc 3 level all the way down, set an appropriate amount in the matrix. Set the mod EG decay time to about 1 sec, release at 45.2 ms. In the matrix, set the amount to 88. We could use a good vibrato effect. Route the LFO to Osc 2 pitch, I like amt about 16. I prefer a triangle wave, but a sine will do fine as well. I like the rate a little on the fast side, about 8.47 Hz.

This is all you need, the GOTTA HAVE IT stuff. Of course, a little chorus effect would be nice, and delay if desired. I'm always scared I'm going to overuse these effects. But think about the DX7--every patch I've copied has in some way used detuning to simulate chorus, so there's certainly no shame in relying on chorus.

Now you have a perfectly good E.Organ 1 sound. Enjoy!