Overestimated synths?

[jancivil]

I bet it needs filters to be anything[/thread]

Rather than sculpting a sound by chipping away at a waveform the way a subtractive synthesizer does, a modelled instrument creates its sounds with mathematical algorithms that simulate the behaviour of the various parts of an electric or acoustic instrument. For example, a bow (an exciter, or, as Yamaha called it, a "driver") scraping across a violin string excites that string (a resonator) into vibration. Additionally, the hollow body of the violin represents another resonator.

This was heady and potentially confusing stuff to musicians weened on a steady diet of square waves and resonant filters - ...


reminds me, I was downloading the demo of Modelonia

Modelonia is designed to be the first compact multi-purpose virtual sythesizer which uses the physical-modelling technique as the only sound source and does not enhance it with multi-effects or sweepable filters (VFC in the Subctractive tradition), Modelonia sound shaping ability relies solely on physical-modelling.

I like this idea!

[jancivil]

For me, it is. If you had that experience, you would get it. I would suggest you look at
fm theory and applications
algorithm by algorithm and actually compare it with the GUI of FM8, actually thinking about it.

But that seems too much to ask, because you're so sure you know better.

You're talking about the graphic visualisations of the 32 hardwired operator arrangements printed on the plastic, without actually understanding what they represent. All of the original DX7 modulator/carrier configurations or ("algorithms" as Yamaha opted to call them) can indeed be made with the FM7/8 matrix, and quite a few more, so he really does know better.

The word is algorithms, not really an optional special new Yamaha term.

I know exactly what they represent having worked extensively with the DX7. The Yamaha algorithm is the visualization. Study it - which is why I provided the materials - and this is perfectly clear. But [both of] you know better about what my thought is (and you arrogant enough to tell us what I know or don't) despite the failure to read or think about it for a second. Typical.

[ghettosynth]

It's weird. The DX7s and D-50s went for €100 for years in the 21st century, but recently prices go up to €350.

Not overestimated but it shows what character these synths have. I think that's the whole point of a synth.

I've never seen them quite that cheap. Not even DX7s believe it or not. Were they that cheap in Europe? Granted, they have both had something of an uptick in recent years.

[Rappo Clappo]

@ghettosynth

I'm from Holland. Bought a Roland JV-90 for €150. Two years ago and I've seen several for that price since then.
But lately prizes have gone drastically up.
Somehow people have discovered they can get good money for their old synths it seems.

Lol a Roland XP-10 for €250? I bought that thing for less way back then!

[jancivil]

The Modelonia manual contains a kind of disclaimer, that some things will do some instruments better than modeling will. As far as literally capturing known instruments, one supposed they mean sampling.

It focuses in bow->string excitement and mouthpiece/wind excitement of resonators, wood box and metal tube essentially.

[ghettosynth]

@ghettosynth

I'm from Holland. Bought a Roland JV-90 for €150. Two years ago and I've seen several for that price since then.
But lately prizes have gone drastically up.
Somehow people have discovered they can get good money for their old synths it seems.

Lol a Roland XP-10 for €250? I bought that thing for less way back then!

Yeah, the JV series stuff was very cheap here for a while. I have also noticed the recent uptick in prices for old hardware in general. That said, it hasn't hit everything. My K2000R is still worthless LOL!

I bought my JD-800 new when Roland was blowing them out for about $1200 and for several years they were worth as much or more used as I paid for mine. They dropped quite low for a while, but are back up again. Not as much as they were, but not quite so bad either.

But I think that there is something to what you're saying, especially at the low end.

[.jon]

For me, it is. If you had that experience, you would get it. I would suggest you look at
fm theory and applications
algorithm by algorithm and actually compare it with the GUI of FM8, actually thinking about it.

But that seems too much to ask, because you're so sure you know better.

You're talking about the graphic visualisations of the 32 hardwired operator arrangements printed on the plastic, without actually understanding what they represent. All of the original DX7 modulator/carrier configurations or ("algorithms" as Yamaha opted to call them) can indeed be made with the FM7/8 matrix, and quite a few more, so he really does know better.

The word is algorithms, not really an optional special new Yamaha term.

I know exactly what they represent having worked extensively with the DX7. The Yamaha algorithm is the visualization. Study it - which is why I provided the materials - and this is perfectly clear. But [both of] you know better about what my thought is (and you arrogant enough to tell us what I know or don't) despite the failure to read or think about it for a second. Typical.

I find your attitude uncomfortable, why do you prefer to act hostile instead of trying to learn something out of this trainwreck? We are not discussing opinions, but rather simple facts.

Yes, there is word algorithm, and such are present in FM synthesis. Algorithm is a set of instructions or rules which result in an outcome, and it impossible to procude a correct audio waveform if given only a picture such as printed on the DX7 case. The pictures do not describe the algorithms, just a part of them- the arrangement in which the operators are placed in the actual algorithm that produces the sound. In other words, which operators modulate (modulators) and which output sound (carriers).

The mathematical principle, the core algo, is the same in all 32 DX7 operator configurations. Only thing that changes is the arrangement of operators. They couldn't figure out or did not have the technological means to implement a freely editable matrix like in FM7/8/Octopus/etc, so they provided those 32 preset configurations which have a decent amount of variety to cover a lot of sonic ground. In FM7/8 you can arrange the operators arbitrarily with 1-6 carriers, and any of them can feed back into itself. I've had too many jallushots and champagnes to count the exact number, but it's even bigger than your ego anyway.

[ghettosynth]

Let's talk brass tacks, which DX7 sounds will have been better purely through applying a filter to it.

This is specious reasoning as I've already pointed out. Since DX7 sounds would have been created without a filter, the question of whether those sounds would be better with a filter is moot. The question that is interesting is whether or not the DX7 could create better sounds with filters.

Again, this is trivial to understand. Simply look at the patches in FM8. Since the FM8 is fully capable of creating pure DX7 patches but has a filter, then if there's no reason to use the filter, it shouldn't be used or it will be used in lazy ways.

I've already talked "brass tacks" which you have yet to respond to with actual effort and citing a Zappa patch or your own is hardly evidence of anything. I've given you a challenge with two patches, both from the factory library. One should be easy to recreate for an expert on FM such as yourself. Go for it, I can post the algorithm if it will help, maybe you can just describe how you would make the changes? If FM as implemented makes filters unnecessary, then recreating a simple three op patch that includes a filter with six FM ops and no filter should be a trivial exercise, no?

Of course, if the filter is truly superfluous, then you should be able to recreate the patch with only three ops and no filter, right? Anything greater than that is, in effect, arguing that in a practical implementation, a filter adds some value to any particular fixed FM architecture. If you can do it in six though, then we can debate the value of a filter with respect to the number of operators and the resulting complexity.

As I've already pointed out, there are many many patches where the filter is used just on feedback or to add character to the output. Further, the saturation module is also used in the same way. Think about that for a minute, it's not about just getting an analog filter sweep. The filter and saturator are being used to shape feedback spectrum. This is absolutely no different in terms of challenging Chowning's original ideas than allowing different waveforms as modulators and carriers. Why, if FM as envisioned by Chowing was so universal, would you need to add different waveforms? The reasoning is simple and if you'd let go of your prejudice you might be able to see it. It grants flexibility that, even if it's possible to reproduce the signal without this feature in theory, it's either considerably more complex or beyond the abilities of the instrument.

FM, particularly as envisioned by Chowning and implemented in the DX series was overestimated, that is the subject of this thread. Subtractive technology, however, was underestimated at the time, at least in terms of value to musicians by Yamaha. It is still widely used in almost all new synthesizer products. Where FM has continued to be used it is almost universally used in a hybrid model with subtractive elements and abuse of its original conception of only using sine waves. Research has conclusively shown that finding FM parameters becomes non-trivial as the complexity of the model increases. Had FM not been overestimated at the time we might have seen phenomenal hybrid synths that would almost certainly fetch a mint today.

There is absolutely no reason whatsoever to mentally separate different synthesis methods and treat them as silos in the context of a product, and the market today is largely telling you that.

My attitude about this comes, in part, from my own experience with FM used with and without filters, an experience that you, admittedly, lack. You might consider gaining some experience with this before preaching to others.

[Rappo Clappo]

"FM, particularly as envisioned by Chowning and implemented in the DX series was overestimated, that is the subject of this thread."

Well, I just noticed this thread is in the section dedicated to software.
But anyway, I don't how it is for Yamaha, but the manuals of the old Rolands are -how should I call it softly- not to recommend.

And the so called polyphony of the JV-90 has become a bit of a joke. Although it still looks cool (that's why I bought it, actually).

[ghettosynth]

"FM, particularly as envisioned by Chowning and implemented in the DX series was overestimated, that is the subject of this thread."

Well, I just noticed this thread is in the section dedicated to software.
But anyway, I don't how it is for Yamaha, but the manuals of the old Rolands are -how should I call it softly- not to recommend.

And the so called polyphony of the JV-90 has become a bit of a joke. Although it still looks cool (that's why I bought it, actually).

A tangential but somewhat on-topic point is that a lot of the early manuals are hilarious. Especially the SH-101 and TB-303 manual. Superb Jinglish to be found there.

While we're on the 303. It was totally an example of an overestimated AND underestimated synth. Roland believed that guitar players would see it like drum machines, as, guitar players often purchased drum machines as practice aids, a fancy metronome, or to lay down demo tracks. The reality is though, it was so hard to program that it just wasn't worth the effort for most. I think that it was probably easier for most guitarists, and about the same price, to just buy a bass guitar. Of course, if you didn't have a multitrack you were limited to ping-ponging with two tape decks.

In any case, in that sense, it was overestimated and was really a failure as a product.

However, it was picked up by electronic musicians and made wildly popular, precisely because of the weirdness of its design in trying to mimic bass guitar, something it actually did quite poorly. Roland could not see the future nor did they really have any insight at all as to where the music scene was going and could not have predicted the popularity of the 303. In that sense, it was underestimated.

[Rappo Clappo]

Yeah the 303 was discovered by an acid genius. I remember. No one knew how to program it or how to use it.

Oh, those were the days of pure magic!

[Richard deHove]

When it was released the 303 was the cheap weak tediously complex little bass sequencer you could use with their cheap tinny little 606 drum machine. Both were what you bought if you couldn't afford anything better. And at least they could pair up so you didn't need to build custom voltage gates so they could talk to each other. The only step lower on the ladder was buying Casio. So at least you could turn up with the 606 and 303 and be taken seriously. Not so with Casio gear (unless you were going for the Da Da Da look).

[xoxos]

j.o.s.' korg z1 does exceptional physical modeling (of a different calibre than modelonia which afaia is closer to perry cook/STK methods) and it has a filter

really though i was reading one of these threads about a week ago and logged in one day and found a couple pages of depthy meaningful posts about making music. i was really busy making something that day so i didn't read beyond skimming. i was amazed. it's like i had this wonderful dream then i woke up to "[prime audio industry meme] yes or no"?

i feel like, i should help my friends realise what a farce participating in this thread is because of all the baiting and to get back to their other things of saving the entire species so they can help in something meaningful. you can see these souls nailed to an asteroid billions of years from now trying to make aliens crash so they can share eternally in the glory of some shitty mass produced synthesizer some tech guy made trying to pay for stuff.

[ghettosynth]

i feel like, i should help my friends realise what a farce participating in this thread is because of all the baiting and to get back to their other things of saving the entire species so they can help in something meaningful.

Well yes, but, nothing is meaningful, life sucks, and then you die. Well, feeling great is meaningful, in the moment, choose your own poison.

[jancivil]

"FM, particularly as envisioned by Chowning and implemented in the DX series was overestimated, that is the subject of this thread."

Well, I just noticed this thread is in the section dedicated to software.

I did not know that! So does this apply to all the threads regarding software vs hardware and all the posts regarding comparison particularly with say the Minimoog?

I didn't know our hijack got to be 'the subject of the thread'; it was for a bit.