Korg Forums

Starting a topic for programming the MOD-7 synth.

I'll kick it off with a question.

I really enjoy the hardware controls on the Tone Adjust control surface when programming a patch. MOD-7 has some really good defaults set up with OSC ratios, mutes, levels, etc. It makes it feel more like programming a hardware synth, which is quite nice.

While you can change the controller assignments, it seems like each program can only have 1 tone adjust definition that is saved along with the program.

It would be really nice if you could have a set of controls for, say EG2, for example. Even if it's just temporary while you're working on that envelope.

I had thought about creating some essentially empty programs that just have the Tone Adjust controls set a certain way - like a program to hold assignments to control EG2. And then just use the copy from function in the menu to import the controller assignments into the current "in progress" patch. Do some work on EG2, write the patch, then import some Tone Adjust controls for something else.

However, the individual EGx controls are not there in the assignment choices. Only the amp/filter, and I think one of those aggregates also controls EG8. So that idea was kind of foiled.

Just wondering if anyone has a best practice or workaround here for adding hardware control to additional elements of the MOD-7 synth.

Let's say you set Filter ADSR to the first 4 sliders. By default an INIT MOD7 uses EG8 for the filter. So now you can shape your envelope using the sliders, then copy EG8 to any of the other envelopes you want.

You also have the Amp EG ADSR you can assign to sliders 5-8. HTH

Yep, good idea!

So change the filter EG to something other than EG8, so you're not adjusting the filter envelope at the same time, assuming the filter is even in use at this point. Then assign EG8 to whatever parameter, tweak it to taste, copy to another envelope and re-assign the pasted-to EG to the parameter once finished?

I will try this. Should work pretty well in conjunction with the tone adjust template preset idea, I'd think.

My current strategy is to take a stock FM-X sound in the MODX, and examine its algorithm, the ratios, the envelopes, and the levels.

Then I mimic the setup in the Kronos MOD-7 engine.

Brass is easy. Using just a simple 2 OSC algorithm, 1 => 2, all you have to do is feedback 1 into itself a bit, and then turn its level up about 3/4 of max. OSC2 then will sound like a brassy sawtooth.

The famous DX7 tine piano really is pretty simple as well. It just needs 3 pairs: 1 => 2, 3 => 4, and 5 => 6. The even numbered OSC create the sound, the odd numbered OSCs make the color. Set OSC1 and OSC3 to a ratio of 1, to get the pure tone part of the tine sound (detune these slightly to get that classic chorus sound). Set OSC5 to a ratio of 12, to get the high metallic part. Balance levels and tailor envelopes to taste.

There is not a direct 1 to 1 relationship between FM-X and MOD-7. MOD-7 displays a sort classic FM "behavior" as you modulate one wave with another and change the ratios. I seem to recall this is how FM on the Fusion sounded, and FM in software behaves as well. There are certain "unsweet" spots as you change the ratios where you get a kind of warbly effect.

Somehow the FM-X engine has smoothed this behavior out. You can still arrive at the more sparkly FM-X sound on MOD-7, it's just not as intuitive to me at the moment.

Lightbringer wrote:So change the filter EG to something other than EG8, so you're not adjusting the filter envelope at the same time, assuming the filter is even in use at this point. Then assign EG8 to whatever parameter, tweak it to taste, copy to another envelope and re-assign the pasted-to EG to the parameter once finished?

The key concept we're using here is buffers. A working space where various parameters are manipulated live, or also a temporary holding area for contents we want to access later.

In other words, do this:

EG8 is assigned to the filter envelope. Shape your filter envelope however you like using the sliders. Then save EG8 to EG9, which is going to serve as our buffer, or holding area. The filter envelope is now duplicated, so we are free to re-use EG8 how we see fit, since we have a backup of our original filter envelope being held in EG9.

Let's say you want to use 1 => 2 to create a bell attack sound. For EG1, you want a quick attack with a quick decay to zero.

Use the sliders (still assigned to the filter ADSR) to quickly shape your bell attack. Now copy EG8 to EG1. Bell attack done.

Now let's say you want to use 3 => 4 for a brassy sustain sound. Again, use sliders 1-4 to shape the brass ADSR (med atk, med decay, mid-level sustain, short release). Now copy EG8 to EG3. Brass envelope done.

Okay, done with envelopes. Let's restore the original filter envelope. Copy EG9 back to EG8. Three envelopes done in less than 30 seconds.

EG9 acted as our buffer/temporary holding area this whole time.

psionic311 wrote: There is not a direct 1 to 1 relationship between FM-X and MOD-7. MOD-7 displays a sort classic FM "behavior" as you modulate one wave with another and change the ratios. I seem to recall this is how FM on the Fusion sounded, and FM in software behaves as well. There are certain "unsweet" spots as you change the ratios where you get a kind of warbly effect.

Somehow the FM-X engine has smoothed this behavior out. You can still arrive at the more sparkly FM-X sound on MOD-7, it's just not as intuitive to me at the moment.

Not sure if I understand exactly what you're saying here. Are you saying with a 2=>1 configuration, and 2's ratio being set to some integer ratio X, that for certain fixed values of X that FM-X sounds sweeter?

Or are you saying that if you apply pitch modulation, such as an envelope to OSC2 that the change of timbre over time somehow sounds smoother?

You would think the math would be the same, and so the sounds the same, but I guess there's no telling what Korg or Yamaha actually does under the hood. Are these observations at really high ratios or using non-sinusoidal waveforms? Could you be running into aliasing and possibly FM-X has higher oversampling? Just thinking out loud. "Warbly" actually makes me think there might be some non-key-tracking frequency offset applied, but if you started from an init patch there shouldn't be any, and guessing you've already checked that.

psionic311 wrote: Use the sliders (still assigned to the filter ADSR) to quickly shape your bell attack. Now copy EG8 to EG1. Bell attack done.

So here you're adjusting an EG that will ultimately be applied to OSC1, but still listening to the EG applied to the filter as you make adjustments?

I guess that might be OK for coarse adjustments. I'm not sure my imagination is good enough to really dial something in exactly that way though. But the concept is still good I think. I could temporarily apply EG8 to the true destination, shape it, then copy EG8 to EGX and change the AMS source to EGX when finished. Extra step, but probably still better than having to continually tap the screen and move different EG segments onto the value slider.

Lightbringer wrote:Not sure if I understand exactly what you're saying here. Are you saying with a 2=>1 configuration, and 2's ratio being set to some integer ratio X, that for certain fixed values of X that FM-X sounds sweeter?


You would think the math would be the same, and so the sounds the same, but I guess there's no telling what Korg or Yamaha actually does under the hood. Are these observations at really high ratios or using non-sinusoidal waveforms? Could you be running into aliasing and possibly FM-X has higher oversampling? Just thinking out loud. "Warbly" actually makes me think there might be some non-key-tracking frequency offset applied, but if you started from an init patch there shouldn't be any, and guessing you've already checked that.

Here it is more exactly by example, after further fooling around.

Create: a basic 2 => 1 algorithm
1) Create an INIT MOD7 patch.
2) On the patch panel, connect Osc2 Out to Osc1 In
3) By default, Osc1 (carrier) Out goes direct to Main Mix In1

Now we have Osc2 "coloring" the carrier.

Experiment: go to Control Surface display, Tone Adjust
slider1 = Osc1 volume/level.... leave at 100 for max volume
slider2 = Osc2 volume
slider7 = macro briteness

Play and hold a perfect fifth. On the Kronos, as you raise slider2, you start hearing harmonics being added.

Same for the MODX7. Raising the modulator's level adds harmonics.

The difference between the two engines becomes evident as you hear the harmonics being added.

On the Kronos, as you raise slider2, the sound is smoother, holding the main notes while sweetly adding more harmonic brightness, with a bit of warble as you increase. It's mostly smooth until between 90-100, where it adds a metallic overtone.

On the MODX7, as you rotate the data wheel to raise the modulator's level, the sound is more sweet and clear, but with digital stepping. The main note gets quieter while the upper harmonics get louder. Between 90-100, it never reaches the metallic overtone (unlike you would on most other FM synth engines). Instead the main tone returns but more throaty, and the upper harmonics are the ones just before they start to turn metallic.

However, on the Kronos, if you reduce slider7 to -30, this reducing of the macro briteness changes the travel range of the harmonics now. The harmonic sweep is almost the same as the MODX7 now. You can't reach the metallic overtone with slider 2 at max.

Still, even though the Kronos-adjusted harmonic sweep now contains similar harmonics, there is yet a discernible difference in the balance of harmonics to the main tone. On the Kronos, it's a sweet and gentle change, more fluid, if a bit bland. On the MODX7, the harmonics are more present, but so is the fundamental, and the overall sound is more pronounced, even if you turn the whole MODX7 down in volume. It's like Yamaha chose to compress the FM output of each FM-X voice, is my guess.

psionic311 wrote: Still, even though the Kronos-adjusted harmonic sweep now contains similar harmonics, there is yet a discernible difference in the balance of harmonics to the main tone. On the Kronos, it's a sweet and gentle change, more fluid, if a bit bland. On the MODX7, the harmonics are more present, but so is the fundamental, and the overall sound is more pronounced, even if you turn the whole MODX7 down in volume.

It's like Yamaha chose to compress the FM output of each FM-X voice, is my guess.

Sure enough, just as I posted I thought about adding a Stereo Compressor to the INIT MOD7 program. And I decided to use the Kronos rotary data dial instead of slider2, to emulate the steppiness on the MODX7.

The harmonic sweeps now sound very similar on both machines. Although the stepping is more fine and fluid on the Kronos. I guess the MODX7 data dial has less resolution.

Too bad I don't record audio or video, lol. Takes so many words to type when showing a 60 second video clip would explain more fully and easily.

Very interesting - so sounds like these desirable qualities of FM-X are easily reproducible on MOD-7, once you understand what is happening then? I guess Yamaha just decided all FM voices sound better with a little compression. Perhaps they're right. What kind of ratio did you use to true them up? I'm assuming it was fairly light compression.

Also sounds like MOD-7 might even be better because of the higher resolution. I wonder if it truly is higher res. It seems like the Kronos engines have internal smoothing everywhere the resolution might result in audible stepping. E.g. if you do a filter sweep in one of the synths, you don't hear zippering even though there are only a finite set of cutoff values - the sweep sounds analog.

I (re)noticed those macro controls on the OSC Main tab the other day; had forgotten about them, and was trying to think about what they would be useful for. There are a similar set of controls in Native Instruments FM8. There they are presented there as Easy Edit (or something like that). So basically a tool for someone who doesn't really care to understand FM synthesis to be able to tweak one of the many included presets.

I guess it could be a similar thing here, since they are mapped by default on the Tone Adj. panel. It's kind of interesting that some controls affect carriers, and others modulators, and you can change how each operator is seen by these controls, even though it doesn't actually restructure the algorithm or change that operator's role in the patch. They don't quite seem like a low level sound design tool. They are an interesting inclusion.

I just threw the default preset of the Stereo Compressor on and pumped the gain up to +10 -- it's a 4:1 ratio.

I think it's a good thing that FM-X limits the range of FM modulation. It narrows it so that there's more of sweet spot. It doesn't let you get much of those nasty FM noisy artifacts until you really push the modulation level to max.

As for resolution, I was referring to the digital resolution of the big rotary wheel of each, not the internal maths or the DAC resolution of the boards. The MODX data dial only has like 50-60(?) discrete steps from zero to max, while the Kronos rotary dial has like 2-3x that. You can still hear the discrete steps, but it's much finer. Using the data slider instead is totally smooth though. MODX has limited data entry methods. Another few plusses for the Kronos!

Like I mentioned, lowering the macro briteness limited the range or intensity of possible frequency modulations. I can see how this could be a handy tool for sound shaping envelope responses and evolving textures with much more detail than possible in FM-X. I'm guessing the FSR-1 probably had some pretty fine-grain control of certain parameters as well.

The macro briteness and macro feedback in MOD7 are overall adjustments to the sound. (I can't find any audible response from macro timbre adjustments). There are still per oscillator LPF and feedback adjustments. And phase. Multiple waveform choices for each oscillator, and waveshaping, and even 4-velocity PCM. And of course multiple mod sources and destination thanks to AMS. Another testament to MOD7 > FM-X.

One thing I'd like to mention as you explore FM programming in either FM8 or MOD7 engines.

If you aren't already thoroughly familiar with the musical harmonic series, also known as overtones, it's totally worth it to understand. FM programming with perspective of overtones helps navigate your way.

Here's a link that gets technical: https://en.wikipedia.org/wiki/Overtone

But really, it's quite intuitive if you're already musically adept on an acoustic instrument. Think guitar harmonics, changing registers on a bugle and other brass, overblowing a flute or recorder to get the octave up, etc.

Once you understand that, here are some numbers to plug into FM ratios:

1.0 = unison (1/1)
1.25 = maj 3rd (5/4)
1.5 = fifth (3/2)
1.75 = 7th (7/4)
2.0 = octave (2/1)

psionic311 wrote: If you aren't already thoroughly familiar with the musical harmonic series, also known as overtones, it's totally worth it to understand. FM programming with perspective of overtones helps navigate your way.

Here's a link that gets technical: https://en.wikipedia.org/wiki/Overtone

But really, it's quite intuitive if you're already musically adept on an acoustic instrument. Think guitar harmonics, changing registers on a bugle and other brass, overblowing a flute or recorder to get the octave up, etc.

Once you understand that, here are some numbers to plug into FM ratios:

1.0 = unison (1/1)
1.25 = maj 3rd (5/4)
1.5 = fifth (3/2)
1.75 = 7th (7/4)
2.0 = octave (2/1)

This is insightful! I'm away from my K for a few days but have been exploring this concept using Tenacious Frog TF7. I can definitely hear this harmonic series on the carrier. One interesting thing is that applied to the modulator, the tones produced by non-integer modulator ratios between 1.0 and 2.0 seem to actually produce a lower pitch than the original (carrier) sine wave to my ear, whereas modulator indices of 0.5, 1.0, and 2.0+ seem to produce higher pitches.

So, here's a question. This probably applies to all synths in the Kronos, really. For the past couple of years I've probably spent more time than anything with Eurorack modules. You can connect a keyboard up to Eurorack but its probably equally or more common to use a step sequencer to drive the note sequence. Some people are able to create entire pieces of music using step sequencers. I'm not really one of them, but I do find they are super nice when doing sound design. You don't have to keep pressing the notes on the keyboard after every parameter change to hear the result. It really lets you focus in on sculpting the sound without having to continually play the keyboard. It's very liberating.

In program mode, there's the common step sequencer, and some of the EXis have additional step sequencers. But to me these don't really seem to be intended to be playable/tweakable note sequencers - you need something else to trigger the amp envelope and they are not very hands on as far as making changes.

It would be really nice to have the option to have a sequencer always playing back some kind of riff while designing a sound. An external sequencer is an option, I guess. I'm not especially inclined to use a DAW, but maybe something like a Beatstep Pro or SQ-1. Or even an iPad app. I wonder if using Edit Program from internal Song Sequencer mode would work here...

What are others doing? I guess Karma is an option too.

You totally can use the Kronos EXi sequencers to play a pattern while you sculpt a sound. From another of my recent posts:

Here's a step by step for future readers.

From Program mode:
EXi1 tab ==> Pitch EG/Mod tab

Under the OSC1 & Sub Pitch Modulation section:
1) set AMS1 to Common Step SEQ
2) Set its Intensity to 25

Under the OSC2 Pitch Modulation section:
1) Only needed if your timbre contains 2 active Exi OSCs
2) Repeat steps 1 and 2 from OSC1 above

Go now to the Common Step Sequencer and set it to loop only 2 steps:
1) Common tab ==> Modulation Tab ==> Common Step SEQ
2) Mode = Loop
3) Start Step = 01, End Step = 02

Notice in the Value section that there is a gray vertical stripe along steps 1 and 2. We will now set these 2 steps to play your note on step 1 and an interval of a fourth for step 2.

Very important key bit of knowledge:
A semitone has a value of 4 in the Kronos step sequencer when the Pitch Mod AMS intensity is set to 25. (see page 178 of the Parameter Guide).

Since we want an interval of a fourth, which is 5 semitones above the original note, the value we want is 5 x 4 = 20. Therefore we set Step 1 to 0 (the original base note you will play on your keyboard), and we set Step 2 to 20 (an interval of a fourth from the previous step).

Value fields
Step 1 = 000
Step 2 = 020

Now when you press a "D" note on the keyboard, you should hear it flip back and forth between a D and a G above it.

The default Duration is a Quarter note = 1. We want 16th notes for each step for this song "Feels Like the First Time". Change the duration of each step so that 16th note = 1.

A quicker way to change all step durations in one go, instead of laboriously setting each step by hand, is to use the shortcut that divides in half each time:

Duration (lower right corner)
1) Press the /2 button a couple times to divide in half each time.

Don't forget to set the Tempo of the song to match, and save the program.

Keep in mind that besides the Common Step Sequencer, there is also a separate Step Sequencer per Exi engine. So for a program made up of Exi1 and Exi2, you have a total of 3 different step sequencers to use as AMS sources. Have fun!

psionic311 wrote:You totally can use the Kronos EXi sequencers to play a pattern while you sculpt a sound.

Thanks for the write up. Yep, you can do that but I guess my point was that it is probably more trouble than it’s worth for this application. You’ll still need to hold a key down to hear it. You’d have to set up a tempo synced square LFO to trigger each EG that’s in use by the program for each new note (which could be a lot with MOD-7) to really hear the what the program sounds like on each note after the note on event. And then you’re still going to have a pattern with no rests. Every note is on. So it will likely get tedious to listen to pretty quickly.

Next time I’m front of my K I’ll see if you can do this by editing the program in sequencer mode. That seems really ideal since you could just record a little loop. It could be polyphonic and have chords. And when it gets tiresome you just go record a new one. I just don’t know if you edit a program that way and write it, if it’s permanently writing the program.

Or you just turn KARMA on.