Overestimated synths?

[chk071]

Frankly, i gave up on the topic. There'll always be the hoard of people which will tell you that any of their super-antialiased, ZDF synths will be able to blow everything hardware, or older software out of the water. The point they're notoriously missing is that it's not so much about the technicality, but, about character, and, especially, organic sound. Which of course also has a technical background, but, antialiasing won't help, if the synth's character is simply shite, or if the oscillators sound bland, ordinary, and flat. There was a discussion on another forum, where i wrote that no soft synth sounds like a Virus, and people were immediately claiming i wanted to say that it sounds better than anything, and were arguing that it's just old DSP in a box. It seems incredibly difficult for certain people to view something ojectively. All that i was saying is that... no soft synth sounds like the Virus. I even wrote then that no hardware synth sounds like Sylenth1 either, and still someone wanted to put it down the "hardware is superior" road... people are emotional, subjective, and religious (and most definitely lack a good ear). And a supersaw is a supersaw, no matter which synth is used. They all sound the same. Right? That's why we all own 20 synths, and get GAS'd over every newly released synth. Right? Riiight.

[robotmonkey]

Diva
Omnisphere
Alchemy
Zebra

[Examigan]

Diva
Omnisphere
Alchemy
Zebra

Ouch...are you kidding?

[fluffy_little_something]

Alright.. good points. It has a clean sound overall, but it is thin when comparing the sound to a virus. I guess many people are using it somewhat like a rompler with all the soundbanks available for it. I always rather use other synths though, for example.. Predator is similar to Sylenth but has more features (minus the fourth osc), and sounds good if not better. I also dislike when the oscillators are not all visible on one page. But sylenth does have a good clean character with good quality onboard effects, so I'll give it that.

I don't really agree. I sold Predator mostly because of its inferior sound quality, be it the unison, be it the noise with certain detune and filter combinations.

[jancivil]

Chowning's goal was academic

Sure, as if you know.

[Fifty Years of Computer Music: Ideas of the Past Speak to the Future
John Chowning

Abstract. The use of the computer to analyze and synthesize sound in two early forms, additive and FM synthesis, led to new thoughts about synthesizing sound spectra, tuning and pitch. Detached from their traditional association with the timbre of acoustic instruments, spectra become structured and associated with pitch in ways that are unique to the medium of computer music.]


3 Structured Spectra and Pitch Space

There are two ways in which additive synthesis and FM synthesis have been used that merit emphasis, because they touch upon issues that are important beyond any particular means of synthesis. John Pierce and Max foresaw one way in the early years: the creation of a non-traditional scale that has a structural link to timbre, where the frequency ratios from the scale are used in the construction of the tone’s spectra. Karlheinz Stockhausen created a similar relationship between pitch and spectrum in his Studie 1 (1953). Risset, however, used synthesis in a manner not foreseen—a manner imaginative and evocative.

3.1 Constructing Spectra in the Pitch Space

The final example in Risset’s catalogue stands as a striking advance in computer music, although little recognized and little exploited. It is the first instance where pitch is used to express timbre in the same functional manner that pitch expresses melody and harmony, that is, melody-harmony-timbre all within the pitch space.


Pitch is composed sequentially as line and simultaneously as harmony, for which there are rich functional theories, but composing timbre as a collection of partials drawn from the pitch space cannot be achieved with acoustic instruments and falls squarely in the domain of computer music.

The sound potential of any instrument is vast, but limited—the partials that make up an instrument’s tone can only be partly modified by performance techniques and devices such as mutes. A clarinet and a violin can play the same pitch at the same loudness for the same duration, but they cannot be made to have the same spectrum through time—the frequency and intensity of an instrument’s partials are locked within boundaries defined by its and the performer’s physical properties.

Risset realized in his timbre studies that in creating natural sounding complex timbres by summing numbers of sinusoids (pure tones) where each sinusoid can have its own independent control over intensity and frequency through time, he had unlocked timbre from any physical constraints. He could create tones that cannot exist in the natural world, complex timbres where the partials themselves are a part of the pitch space. He composed a short pitch sequence that is heard first sequentially in time (melody), then simultaneously in time (harmony), and then again simultaneously with exactly the same pitches but now as partials associated with a single sound source, {as shown in Fig. 1. [2].

Because all of the partials die away in a similar manner, they fuse and are heard as timbre rather than harmony. The timbre is similar to that of a gong, but a gong whose spectrum is imprinted with pitch information, giving the sound an extra-natural structural link to the preceding. Risset’s was an altogether new conception, uniquely possible with computers, and beautifully framed in several of his compositions, first in Mutations (1969).

footnotes:
4 The ease with which spectral change could be coupled to effort (key velocity) is one of the reasons for the YAMAHA DX7's remarkable success.
5 The first real-time FM synthesis was programmed on a DEC PDP-15 computer by Barry Truax in 1973, while studying in Utrecht. At Stanford, Bill Schottstaedt developed a particularly powerful form of the algorithm that was used in many compositions for many years.

There's a hint at the goals. Which research went on for decades, building on eg., Risset.

This is simply significant. The instrument Yahama licensed from Stanford University was significant. It still is. It's ground-breaking. Filters has fvck-all to do with it.

Do tell yourself you know what you don't, but Dunning-Kruger Effect doesn't really need you as further proof.

[jancivil]

https://www.youtube.com/watch?v=__O39TcJEd8

[aumordia]

somebody is Mad On The Internet

A thought about being "overestimated" -- perhaps its important to keep the target audience in mind? I'm (infamously) no fan of Omnisphere, but for its target audience -- they'll TV and videogame soundtrack composers -- it's probably impossible to beat. You get zero points for originality in that space, directors in these mediums tend to want a certain sound, one which Omnisphere has a nearly limitless supply of. So is Omnisphere really overestimated? Perhaps, but it seems to me that if that's the case, it's because of certain users who proclaim it to be the one synth to rule them all, as opposed to the best synth for its particular niche.

[ghettosynth]

Chowning's goal was academic

Sure, as if you know.

As if you would know that I don't. His goal was academic, and frankly, I don't think that you really have the experience to understand what I mean by that. In fact, your copy paste only supports that idea, it does not refute it. Stanford's goals in licensing the technology were different as were Yamaha's goals in building an instrument.

None of this whining has anything to do with the fact that we're talking about "overestimation." More was expected of FM than it was able to deliver, at least within the time frame of its popularity and in the context of a practical musical instrument. This has nothing to do with whether or not it was groundbreaking, it certainly was. However, as a practical method for synthesis within a commercial instrument, e.g., the DX7, it was certainly overestimated. I don't need to prove this to you, the market has done that nicely.

We're back to analog filters today, and were only a decade or so after the initial FM instruments, because, as practical synthesis goes, they are an improvement over FM alone.

This is trivial to understand if you grok what the point of FM as a practical synthesis method actually is and what value analog filters provide beyond that as musical tools rather than as largely theoretical constructs to support an academic goal. Moreover, this has nothing to do with FM being difficult to program, per se. FM is a fantastic and rather easy way to think about synthesis, up to a point. The issue here is really rather simple, beyond a certain point, the method becomes impractical as the mapping between the control parameters and the desired change in spectra becomes non-trivial. This is well supported by the literature. In other words, it doesn't actually make sense to use FM to synthesize that which is more easily obtained through the use of a filter.

What we have learned is that when you add filters to FM you get a better instrument. That is a fact that no amount of hagiographic whining will change.

[ghettosynth]

Frankly, i gave up on the topic. There'll always be the hoard of people which will tell you that any of their super-antialiased, ZDF synths will be able to blow everything hardware, or older software out of the water. The point they're notoriously missing is that it's not so much about the technicality, but, about character, and, especially, organic sound. Which of course also has a technical background

Really? So what is that technical background?

You are painting a false dichotomy along with a straw man.

[ghettosynth]

somebody is Mad On The Internet

A thought about being "overestimated" -- perhaps its important to keep the target audience in mind? I'm (infamously) no fan of Omnisphere, but for its target audience -- they'll TV and videogame soundtrack composers -- it's probably impossible to beat. You get zero points for originality in that space

No doubt, but, the product title itself, the marketing efforts, and it's borderline rabid fan base, are all that you need to understand that more than a few people "overestimate" Omnisphere

[robotmonkey]

Diva
Omnisphere
Alchemy
Zebra

Ouch...are you kidding?

Indeed. Putting Alchemy on the list was mistake.

[Urs]

Never really clicked with the Uhe selection of Synths which come across more like
a technical exercise in how much cpu power they can drain from my computer.
and then there's the placebo thing going on where everyone swears
that anything they make sounds infinitely better than anything else.

Radio: "There's a wrong-way driver on the motorway..." - "One? - Hundreds!!!"

Take Hive for example, I am not even sure it stacks up well against Sylenth considering the era they were created in. Certainly not the Sylenth killer some have suggested.

Serum killed Sylenth. Hive had nothing to do with that. I swear!

I have a feeling I am going to regret saying that especially here on kvr

If you think you'll regret "saying" something, just wait till one day you figure it out and see what you missed out on

[chk071]

Frankly, i gave up on the topic. There'll always be the hoard of people which will tell you that any of their super-antialiased, ZDF synths will be able to blow everything hardware, or older software out of the water. The point they're notoriously missing is that it's not so much about the technicality, but, about character, and, especially, organic sound. Which of course also has a technical background

Really? So what is that technical background?

You are painting a false dichotomy along with a straw man.

Of course "organic" sound has a technical background, as it takes technical measures from developer's side to achieve it. What i was trying to say is that, of course, to sound organic, the dev has to work on it, and implement stuff alongside the obvious "sound-improving" things like antialiasing, and, especially, that this work should not fall short. Because, frankly, i've heard soft synths which are super antialiased, and sound bland and boring. There are dozens of factors which are important for an organic, alive sound, apart from antialiasing, or ZDF filters. Actually, i would even go as far as to say that this factors are some of the least important. Especially considering that some of the cult VA hardware synths alias quite a bit, and don't have ZDF filters.

[ghettosynth]

Frankly, i gave up on the topic. There'll always be the hoard of people which will tell you that any of their super-antialiased, ZDF synths will be able to blow everything hardware, or older software out of the water. The point they're notoriously missing is that it's not so much about the technicality, but, about character, and, especially, organic sound. Which of course also has a technical background

Really? So what is that technical background?

You are painting a false dichotomy along with a straw man.

Of course "organic" sound has a technical background

Yes, you said that already, what is it? In fact, define "organic sound."

[recursive one]

Frankly, i gave up on the topic. There'll always be the hoard of people which will tell you that any of their super-antialiased, ZDF synths will be able to blow everything hardware, or older software out of the water. The point they're notoriously missing is that it's not so much about the technicality, but, about character, and, especially, organic sound. Which of course also has a technical background

Really? So what is that technical background?

You are painting a false dichotomy along with a straw man.

Of course "organic" sound has a technical background

Yes, you said that already, what is it? In fact, define "organic sound."

I think it's the task for the developers, not the synth players, to define what techncial fetaures of a sound engine contribute to the perception of the sound as organic/stale, fat/thin, punchy/weak etc. As far as people mostly agree on that certain synths have certain of these qualities I think it should be possible to correlate "organic sound" with some measurable parameters.