Yamaha GS1 Electric Piano patch

Oh, you missunderstood me, I was tinkering with my computer not with the sound, that's why I came up that late with the patch...
But still, to built it took about two hours or so...

The spice isn't part of the algorithm, it's part of quantisation noise the old FM-Synths tend to create probably due to the lower enginge resolution. I guess a S/H module between the operators would be helpfull

 

What is S/H module?

According to http://members.tripod.com/kmi9000/kmi_multim.htm, the GS1 has 8 operators. Does that mean its impossible for the FM8 to fully replicate GS1 sounds since it only has only 6 operators?

Is it possible to implement more operators, maybe up to a limit of say, 48, to have an equivalent of TX816?

 

S/H stands for Sample and hold, since I'm pretty lazy here's a description of it's function

http://en.wikipedia.org/wiki/Sample_and_hold

Not necessarily, if I'm not mistaken it's rather working with 2x4Operators and a fixed 2 Modulator 2 Carrier algorithm. Due to that fixed operator design I'd guess that in lot's of cases some of the operators are simply turned off...

The TX816 is simply module with an equivalent of up to 8 x DX7 stacked. There were also smaller versions with less TF1 modules avaiable.

http://www.sequencer.de/syns/yamaha/TX816_TX416_TX216.html

So simply stacking a few FM8 Sounds, for instance in Kore or similar software, should do...

 

It'll be good if FMx can implement > 6 operators. Many great sounds in modern hardware synth are created by layering simpler sounds. It's a big inconvenience to create seperate FM8 tracks in DAW just for layering. Using Kore cost $

Back to my point of the unpredictability of FM synthesis. Even for a seemingly simple GS1 EP1 sound, it takes a lot of experimentation/guess work to tweak the algorithm and freq ratio and there's no guarantee that the replication is completely accurate.

I was deeply drawn to FM sythesis in the 1980's because of a Japanese Jazz Fusion group - Casiopea. During FM's heyday, Casiopea's keyboardist was using 100% FM synth - DX1, DX7 and TX816 with KX88, and the sounds were amazing! I tried replicating some of those sounds but had only limited success. I was quite certain that those FM sounds were so good because they were layered sounds from the TX816.

 

You haven't forgotten the Digital adjustment on FM8's master page have you?

 

Haven't tried that, but at least on the FM7 this usualy was not much help. It's more effective using the saturator to add some dirt, but with two almost independet sounds in one patch, I'd need two of them. Anyway, the FM8 was never built to emulate an GS1...

 

Don't get me wrong, but I must admit that I'm not a big friend of layering sounds, I'm rather a friend of creating more complex sounds all along, as well as I prefere more sophisticated unisono modes, with lot's of variaton parameters.
Layering is far too much endless trying out what sounds may work or don't work with each other, I'm rather a friend of target oriented soundprogramming. I love to experiment but not to dice...
Still I have no influence on this, so it doesn't make any sense trying to confince me... I'm just the sound guy...

Well, when it comes to layering only I think there are a few more affordable but less mighty alternatives avaiable...

Well, it's the same when trying to replicate sounds between different subtractive synths. Filters and oscillators sounds different, thus it will not sound the same. Does this make subtractive synthesis unpredictable, I don't think so...

I don't know the band well enough, but sometimes things that seems to sound like a DX are simply from a different sound generator. Synclavier did a pretty sophisticated FM and PPG is able to create FM alike timbres as well.

 

Summa, you sound like you have been playing with FM for a long time. Have you any recommendations for learning how to develop a sound in a predictable way?

I have had a fascination with FM for a long time but when I looked at the DX7 not long after they came out the interface lost me completely and I sort of dropped interest.

I found FM7 about 18 months ago after a period away from music and a lot of the programming nightmare floated away. Now I would like to learn beyond the basics. Did you learn by trial and error?

 

I learned more while making sounds for others than in the time I was creating patches for myself only...

Well, quite some but far too much for a posting, probably enough for a tutorial or even a book.
General helpfull is to add chorus while creating the sound, since it helps to detect the differences between the timbres, what is a good advice for all sorts of non-subtractive synthesis forms...

When it comes to FM I prefere working with editor software, otherwise it's simply too much paremters to keep them all in mind So my first encounter with FM about 1987, programming a TX81z using the front panel, wasn't much fun either, but later I participated alot from the experiences I got at that time...

I had a good 2nd start about 1999, using FS1R together with Sounddiver. I think when it comes to FM it's not good choice to start with a low level 4 OP maschine, one get's frustrated by the lack off posibilities pretty soon. So it's a good thing to chose a compromise between complexity like for instance FM7/8...

I had no choice, most of the books that covering FM are pretty detailed about how to make certain sounds, but rather quiet, except a few rather hard to understand formulas, when it comes to explain how FM is ticking...

 

Thanks for that reply Summa, guess I should start listening a bit closer and experimenting.