DX7 Data-Sheets > Zebra?

[hakey]

Minor details like this aren't that important anyway. Just listen to some of Howards DX pianos from the factory lib, they are awesome and don't care about this stuff (I think).
This was more like a "Let's see if I can do it" thing.

Yep.

Still, there's plenty to be learnt from reverse engineering stuff.

And it would be nice to have the perfect recreation of that E Piano 1, and some other classic DX patches, in Zebra.

[Bronto Scorpio]

Go Dennis! Don't think you're wasting your time - you have contributed a LOT to your fellow community of musicians

Thank you very much
I'm just happy I have some people here to discuss this kind of stuff with!
My girlfriend and my cat just don't care anymore

Minor details like this aren't that important anyway. Just listen to some of Howards DX pianos from the factory lib, they are awesome and don't care about this stuff (I think).
This was more like a "Let's see if I can do it" thing.

Yep.

Still, there's plenty to be learnt from reverse engineering stuff.

And it's a lot of fun

And it would be nice to have the perfect recreation of that E Piano 1, and some other classic DX patches, in Zebra.

Indeed! That would be quite interesting!

Here is a very quick shot at the first algo.
MSEG1 controls the volume of OSC1 and ENVs 1 to 4 control the volumes of FMO 1 to 4.
I've set the FM depths to 100, which seem to be "a little" high.

Edit: To actually control the volumes of the operators you'll have to adjust the volume mod depths.

Edit 2: Whoops
I forgot 1 operator

Edit 3: Am I dumb or isn't algo 1 even possible?
I used this image as reference:

Cheers
Dennis

[Gonga]

DXulator, DXManager3 (from FM Alive) and FM8 will all load DX patches, and have graphical representations of the envelopes for each operator with nodes you can drag around. I thought this might help those of you who don't own FM8. DXulator is free, while DXManager is $15 US.

[hakey]

Am I dumb or isn't algo 1 even possible?

No - you're not dumb. I am (I missed that one, and algo 2 as well).

All algos, bar 1, 2, 4 and 6, should be possible... I think.

Edit: and algo 3 and probably a few others...

Edit: 1, 2, 3, 4, 6, 19, 21, 22 aren't possible. Any more?

[Bronto Scorpio]

Am I dumb or isn't algo 1 even possible?

No - you're not dumb. I am (I missed that one).

All algos, bar 1, 4 and 6, should be possible... I think.

Yeah, when I first checked which algos would be possible in Zebra ages ago I also only found 2 which aren't possible.
This one isn't that obvious because it doesn't have any feedback between operators.

Cheers
Dennis

[hakey]

Of course, "Zebra is very limited in terms of FM".

[hakey]

DXulator is free

cool

[hollo]

Also the key tracking (free Break Point is missing)

ModMappers?

All Zebra oscillators have key-tracking. And you don't really need to change breakpoint positions (change the level instead)

Unless I'm missing something, I don't think it's that simple - eg, where scaling on either side of a break point is exponential.

In the Osc - key - levels we can mimic some kind of exponential too, by adding
more points to draw levels.
Thanks Howard!
If needed ModMappers can do it more fine.
Thanks Hakey!

GO for it Dennis! You'v found this phase thing already - i'm just going to check it.

[pdxindy]

great thread!! I love this stuff!

[pdxindy]

Also the key tracking (free Break Point is missing)

ModMappers?

All Zebra oscillators have key-tracking. And you don't really need to change breakpoint positions (change the level instead)

Unless I'm missing something, I don't think it's that simple - eg, where scaling on either side of a break point is exponential.

In the Osc - key - levels we can mimic some kind of exponential too, by adding
more points to draw levels.
Thanks Howard!
If needed ModMappers can do it more fine.
Thanks Hakey!

GO for it Dennis! You'v found this phase thing already - i'm just going to check it.

I don't remember being able to add points to those curves?

[hollo]

Yes i was wrong.

There is 4 point, so you can make a poor mans exponential on one side and a linear on the other.

[hollo]

Here is a very quick shot at the first algo.
MSEG1 controls the volume of OSC1 and ENVs 1 to 4 control the volumes of FMO 1 to 4.
I've set the FM depths to 100, which seem to be "a little" high.

Edit: To actually control the volumes of the operators you'll have to adjust the volume mod depths.

I put Osc2 (sinus) to the top of the second line, and the volume can come from MSEG2.
because it's op6 have feedback, i should set FMO3 to FM self+
(op6=FMO3)

There is algorithm 1. Do you think is OK?

[Urs]

Well, hehe, I guess all algorithms can be done if feedback is set to zero.

But here it comes: Where feedback isn't set to zero and spans only a single oscillator, use a normal oscillator with spectroblend sine. Use "Scrambler" as OscFX, which is plain simple (but antialiased) FM feedback on the wavetable. Should deliver pretty good results with Resolution around 7 to 8

So here you go, all but 2 algorithms

[Urs]

Another trick: If you run out of lanes (don't think you will, but...) you must not forget that the DX have a mono signal path. You can pan modulators hard right and left and keep a carrier stereo to create two paths in one lane, and use Channelmixers to make them mono afterwards. This way you can exceed the number of available operators in DXs pretty nicely

[hollo]

Wow a VOICE calling us from Australia!

Are you sure you all right on your holiday Urs?
We can free you up when you want to come back and make some
nice FM algorithm!

Hope you having a good time! This was really something new
about FM! I'll try the scrambler trick.

Cheers