recreating Bell-like synth

[fanodu1]

For the longest time I've been wondering how some of the sounds in the Tekken 3 soundtrack were made, or rather what specific synths/effects were used because to this day it sounds so distinctively Tekken. One of the sounds that stand out in particlar is this "bell"-like synth in Law's theme that can be heard from 1:25 onwards: https://youtu.be/qZCDO10Kw4w?t=86 Any hints would be greatly appreciated!

[discoDSP]

Those bell hit sound very FM-ish to me. So you know what to look for to start with: a FM synth.

[BlackWinny]

Yes. Any FM synth. Bells are some of the most famous sounds of the FM synthesis since its creation in 1969.

[McLilith]

Maybe play a gamelan sample with bit reduction — maybe 8 bits or a bit less? Then use EQ to filter to taste?

[fanodu1]

Those bell hit sound very FM-ish to me. So you know what to look for to start with: a FM synth.

Yes. Any FM synth. Bells are some of the most famous sounds of the FM synthesis since its creation in 1969.

You're probably right, I figured it might be either FM or maybe a wavetable synth. It definitely doesn't sound analog. I guess I'll dig through my dexed presets and find a similar one and tweak it from there since I'm not very good at creating FM patches from scratch.

Maybe play a gamelan sample with bit reduction — maybe 8 bits or a bit less? Then use EQ to filter to taste?

Yeah it does kinda sound like that. I'll try tweaking some samples to get a similar sound.

[Benedict]

Possibly FM but I think they are more likely Additive (which FM actually is) or maybe even a contemporary Physical Modeled form of synthesis as they sound pretty detailed under that bit compression. I also expect that the sampling process has had an impact (and god knows what fx may have gone in the middle).

[McLilith]

Possibly FM but I think they are more likely Additive (which FM actually is)

I've been an electronics technician since 1982, and I'm familiar with what frequency modulation is on a technical level. I've never heard anyone claim frequency modulation is actually additive synthesis. Do you have some sort of link that backs up that idea?

I can assure you, your local FM radio station is broadcasting a single radio frequency that is being frequency modulated at an audio rate. They're not using additive synthesis. FM and additive synthesis are not the same thing.

[Benedict]

Possibly FM but I think they are more likely Additive (which FM actually is)

I've been an electronics technician since 1982, and I'm familiar with what frequency modulation is on a technical level. I've never heard anyone claim frequency modulation is actually additive synthesis. Do you have some sort of link that backs up that idea?

I can assure you, your local FM radio station is broadcasting a single radio frequency that is being frequency modulated at an audio rate. They're not using additive synthesis. FM and additive synthesis are not the same thing.

Well FM (or Phase Modulation as I'm sure you know) has nothing really to do with FM radio from a synthesist's perspective so let's remove that connection right now.

Yamaha style PM, commonly called FM, is a Sine wave - no overtones. To get more complex tones one sine wave is multiplied by another which adds more complexity to the wave shape. The extra harmonics weren't there before the "wave shaping" occurred. To make a very complex sound you use several sine waves affecting each other. Hence a form of Additive process as you stat with a Sine and end with a Saw. In Subtractive you start with a Saw then shave harmonics off with a filter.

This is the same core reasoning Roland's D-50 style was referred to as Linear Arithmetic. As you Add layers of sound to get new ones.

[imrae]

So you would consider saturating the mixer in a Minimoog to be a form of additive synthesis?

You're pretty far from standard nomenclature here.

[Benedict]

So you would consider saturating the mixer in a Minimoog to be a form of additive synthesis?

Yes it technically is seeing you are adding overtones at that point in time to create your tone instead of taking them away, even tho your Model D is indeed mostly a subtractive style synth. Unless of course, you are using one OSC to audio-rate Mod the other then we are back to additive - that you will probably then subtract from .

You're pretty far from standard nomenclature here.

I know, but that is a result of fashion (and sheer laziness) which really has no impact on reality .

I do understand, and it took me only a few decades to realize this but you are always adding or subtracting and they both go on all the time in most synths (or even in real instruments like when you overblow a Saxamaphone to make it growl or scream - you are waveshaping that sucker to add extra harmonics).

[imrae]

It's much easier to communicate when people don't change the meanings of words.

"Waveshaping" processes generally refer to manipulation of the waveform with a (relatively slowly-varying) transfer function. I don't believe that to be a good description of overblowing a saxophone.

Additive synthesis generally refers to processes that directly manipulate partials; this could be implemented by FFT methods, mixing simple waveforms or even waveshaping with Chebyshev polynomials in principle. But it's not a good description of typical PM synthesis in which many harmonics appear in a more complex fashion.

[imrae]

(To be clear I do see the value in breaking synthesis into processes that add and remove harmonics. We just have to avoid the word "additive" as it has a more specific usage.)

[Kwurqx]

About additive/spectral synthesis....

Additive synthesis is quite rare in synthesizers. Especially where you get control over the individual per partial amplitudes. In hardware Kawai is one of the manufacturers that gave it a try in their K5 en K5000.

Compared to subtractive synthesis or AM, PM, FM it is often a lot of work to set up a spectrum of tens or hundreds of partials and link then to modulators/envelopes.

Some additive/spectral synths offer spectral analysis/resynthesis, e.g. by converting a sample or picture into a spectrum (for further processing).

A free synth where you can set (not modulate) both amplitude and phase of individual partials is Seaweed Audio - Fathom (MONO is free).
https://www.kvraudio.com/product/fathom ... weed-audio

There are other and polyphonic free options, but most that I know of are 32-bit. Like Harmonaut and Oatmeal (use Lumina skin).
https://www.kvraudio.com/product/harmonaut-by-tonebytes
https://www.kvraudio.com/product/oatmeal-by-fuzzpilz

There is also the convert-image-to-sound approach. E.g. the free Fragment
https://www.kvraudio.com/product/fragment-by-fragment

Then there is the spectrum-but-not-individual-partials approach. Like used in the free Tunefish (and derivatives).
https://www.kvraudio.com/product/tunefi ... in-control

You can use any FM synth as an additive synth. Commonly the operators will produce at least a sine. But you will have limited numbers of partials (commonly 6). Commonly you get multiple carrier algorithms. Each carrier can be seen as an additive component. But you can combine FM and additive to gerenerate more complex spectra with one ore more modulators and then use multiple carriers. E.g. the free Yamaha DX7 emulation Dexed. Or Oxe FM.
https://www.kvraudio.com/product/dexed- ... l-suburban
https://www.kvraudio.com/product/oxe-fm ... e-software

There are commercial options for additive/spectral synthesis/resynthesis too off course. But not too many. E.g. Loom and Iris.
https://www.kvraudio.com/product/loom-i ... technology
https://www.kvraudio.com/product/iris-2-by-izotope-inc