1. IMPORTANT:
    We launched a new online community and this space is now closed. This community will be available as a read-only resources until further notice.
    JOIN US HERE

FM8 vs DX/TX/FS1R Operator Layouts

Discussion in 'FM8' started by sginn, Jun 27, 2007.

Thread Status:
Not open for further replies.
  1. sginn

    sginn Forum Member

    Messages:
    58
    Does anyone know if there is a chart or table or description of how to relate FM8's Operator Layouts / Algorithms to the layouts that are published by Yamaha in the docs that accompany their hardware such as the FS1R?
     
  2. Summa

    Summa Sounddesigner

    Messages:
    1,243
    When it comes to the FM8 there is no fixed operator layout, the FM-Matrix presets are just some kind of shortcut, you can interconnect the Operators the way you like...
    The FS1R is using 8 Operators with 88 Algorithms, but those algorithms contain a DX7 subset for compatibility reasons...
     
  3. sginn

    sginn Forum Member

    Messages:
    58
    Hi Summa,

    thanks for the reply

    I understand I can interconnect the operators any way I like, but when trying to replicate the sounds from Yamaha FM gear, it would help if I could learn how to interconnect the FM8 operators so that they give you the same sound as the the DX/TX equivalents. If there were some kind of side by side comparison of an example Algorithm and the equivalent matrix in FM8, that would help at least getting the fundamental sound close before additional tweaking (I would think?). I know FM8 can import the files from the TX/DX models, right?

    thanks
     
  4. Summa

    Summa Sounddesigner

    Messages:
    1,243
    Yes FM8 can import sounds of most 6 and 4 Operator FM-Synths like DX7, TX81z, DX100 ... and similar...

    Well, the FM-Matrix presets contain the original DX-Algorithms as well. It should be too hard to find a picture of the original DX7-Algorithms online and finding the related entries in the Matrix-Presets.
    Anyway, it's rather the Carrier Modulator Frequency relation than the Algorithm that are responsible for the timbre of the FM. Like Modulator->Carrier M:C= 3:1, 6:1, 9:1 are nice for bell type sounds. While Carriers with a higher frequency than the modulators tend to create resonant type timbres, the higher the carrier the higher the frequency of the "resonance" peak...
    When it comes to the algorithms, having more carriers is mostly similar to working with more Oscillators on a subtractive synth. While less carriers and more modulators in general allows to create more complex timbres. It's for instance possible to create Ringmod alike sounds using 2 serial connected Modulators on one Carriers M2-M1->C, while m2 and C are having similar frequencies, one should set M1 to a rather high or low frequency, different to the ones of M2 and C...
     
  5. sginn

    sginn Forum Member

    Messages:
    58
    ok then I should try and understand how the Carrier Modulator Frequency relations are depicted by the DX Algorithms to have a better idea which one to pick in FM8 as a starting place. Any suggestions as to how to learn to read a DX Algorithm to know what the relation is? Can you pick one Algorithm from the FS1R manual and explain what it is so I can reverse engineer the rest?

    Thanks
     
  6. Summa

    Summa Sounddesigner

    Messages:
    1,243
    The frequency relations have to real connection to the algorithms....

    [/QUOTE] to have a better idea which one to pick in FM8 as a starting place. [/QUOTE]

    I mostly start with the so called simple FM, a simple two Operator connection and add additional operators one by one to that setting...

    When I started with FM 1987 with the TX81z I found the algorithms pretty self-explanatory, so I'm not sure where to start.

    Well, let's see... FS1R Algorithm 77, Operators 4 and 8 are carriers, they're the one that are connected to the output, the one you finaly hear. Operators 5, 6, 7 are Modulators (just influencing but can't be heard directly) and modulating Operator 8. Operators 1, 2, 3 are Modulators as well and modulating Operator 4. Operator is also able to selfmodulating itself, depening on the feedback and operators level that tends to increase all harmonics until the operator wave disintegrates into noise ...
     
  7. sginn

    sginn Forum Member

    Messages:
    58
    ok with your explanation and looking at a diagram of the algorithm I see how the carrier modultation relationship is working as far as the Yamaha algorithms. I am not sure it is possible to recreate Algorithm 77 in FM8, but if I wanted to get close how would you recommend that I do so? Just looking at the list of Matrix presets doesn't give me enough information to determine what is the closest match. Also, looking at the graphical interface in FM8 ... are the Carriers represented horizontally and Modulators virtically?

    Thanks
     
  8. sginn

    sginn Forum Member

    Messages:
    58
    Summa,

    as an add on question ... since the FS1R has up to 4 voices for each preset and each performance can be comprised of up to 4 presets, how do we use FM8 to achieve similar results? Do we need to set up an environment such as in Logic with multiple instances of FM8 played simultaneously, or is there anywhere in FM8 that we can combine multiple parts?

    Thanks
     
  9. sginn

    sginn Forum Member

    Messages:
    58
    and I guess one more question about the Algorithm conversion subject ....

    just knowing the Algorithm arrangement from the FS1R Algorithm charts does not tell you what frequencies each of the operators (carriers & modulators) are set to, but I suppose if I bring up an FS1R editor, I can determine this information, right? However, if I do this, is the numeric information displayed for the FS1R directly transportable to the FM8 operator or will there need to be another mathematical conversion between the two architectures?

    Thanks

    P.S. NI could sure help answer all these questions by updating FM8 to the level where the FS1R is ... 8 operators, voiced & unvoiced (using formants), multiple voices as well as AN1X type filters, and a Performance feature! :)
     
  10. Summa

    Summa Sounddesigner

    Messages:
    1,243
    Well, FM8 is not supposed to recreate 8 Operator Synth algorithms or FS1R Sounds...

    Chose a DX7 algorithm and I'll explain...

    As I already stated, the carriers are the operators that are connected with the output. Other than on the Yamaha FM Synthesizers every carrier can be a modulator as well...

    You could use Kore to do that...

    Yep, the editor will show you the frequency relations...

    Since both are using Ratio Frequencies no conversion is needed. Levels values probably differ. Never tried a conversion between those machines, since they only share the DX7 but the rest differs a lot...

    Why should they, the FM7/8 is a lot more common than the FS1R...
     
  11. sginn

    sginn Forum Member

    Messages:
    58
    Sure FM7/8 is more common than the FS1R, just the same as the TX & DX variants were more prevelent than the FS1R. But since the FS1R is essentially backwards compatible with the DX, it obviously has quite a bit more to offer than the DX. So should NI just keep the FM8 where it is today, and not even consider adding the more advanced features of the FS1R because it is more common than the FS1R?

    thanks
     
  12. Summa

    Summa Sounddesigner

    Messages:
    1,243
    I think the problem is that only a few ppl. understand FM fully and even less are into formant synthesis as well, that makes it pretty hard to sell a synth and find good sounddesigners for the factory presets. I find complex synths to be pretty challenging but I think most ppl. tend to buy things that look easy to use...
    Other than that used FS1Rs are pretty affordable now, after the first price drop I bought a new FS1R at the end of 1999 for about 850 Euros, and got a 2nd one about two years ago for about 320 Euros, since I really love that synth...
     
  13. sginn

    sginn Forum Member

    Messages:
    58
    I agree ... I am in the camp of not fully understanding FM. And although formant synthesis and working with FSEQ's are way over my head, it would still be nice if FM8 had 8 ops, multiple voices and performances like the FS1R does.

    I have been trying to find ways to move from my 8 space rack of synths to software equivalents and have been successful in finding adequate software replacements for all my synths except the FS1R and the VL1-m. Some of the sounds I play on the FS1R are basically beefed up DX & TX layered sounds, which I think I might be able to reproduce on FM8, but others are going to be extremely difficult unless NI adds some of these other desireable features.

    Thanks
     
  14. musos

    musos NI Product Owner

    Messages:
    395
  15. sginn

    sginn Forum Member

    Messages:
    58
    I have Autosampler and it works pretty nice. However, the thing I like most about the FM synths in general and the FS1R specifically is their responsiveness to control by wind controllers. With samplers, you have to deal with issues like velocity switching, etc. that is not handled as well on an EWI. I really have thought about using Autosampler to recreate some of the sounds from the FS1R, but I just don't think EXS24 will succesfully reproduce the sound and playing experience.

    Yesterday, I was playing a sound on the FS1R called Fat Glide. It is programmed specifcally for the windcontroller and its so much fun to play. I brought up Sounddiver to take a look at it, and learned that it didn't use FSEQ's and thought it might be a good candidate for FM8, but then learned that it was 8 Op instead of 6, and it had 2 voices layered. Oh well ...


    thanks
     
  16. Summa

    Summa Sounddesigner

    Messages:
    1,243
    Is it part of the presets? What number, I hardly had a look at them...
     
  17. sginn

    sginn Forum Member

    Messages:
    58
    I think its called WX Fat Glide and it shows up as patch number 5. But, if you are referring to the algorithm preset, then I think it is number 17, but will let you know for sure when I get home. In any event, the algorithm did use 8 ops.
     
  18. sginn

    sginn Forum Member

    Messages:
    58
    I just checked the Algorithm sheet here from work and confirmed that it is #17 in the FS1R.
     
    • Funny Funny x 1
  19. sginn

    sginn Forum Member

    Messages:
    58
    Summa, did you get a chance to look at my last response to your message?

    thanks
     
  20. Summa

    Summa Sounddesigner

    Messages:
    1,243
    Yep, but just a short look, due to my workload. It doesn't sound too complex, so one might not be able to convert it parameter by parameter but it should be possible to come very close.
     
Thread Status:
Not open for further replies.