The Fight for FM

[Aroused by JarJar]

I'm really tired of this theory v. practice debate. I know what my ears are telling me,

just don't expect others to trust what your ears are telling you.

When it jibes with others' ears are telling them, and it happens over and over again for many years....

"blurred but harsh.." certainly makes sense to me, and I've heard the same description and close variations for many years, about sounds for which I have similar feelings. And it can be explained without voodoo, although for a musician even when it can't be explained scientifically, so what.

No it's not all subjective fuzz. Especially because you are a developer I think you should not let yourself be mislead by internet fog-faeries prancing about jeering at everyone who is merely using completely normal musical terms in normal musical ways.

Sure, but how the processing is done, i.e., resolution, shortcut-taking, etc., can all affect the output.

raise your hand if you've already coded an FM synth.[/quote]

You know very well that he is quite right that "resolution and shorcut-taking, etc." does have effects on the sound. "how the processing is done" would have to be clarified of course, does he mean in blocks, or what's actually patching or what, beats me.

[grymmjack]

If Toxic 2 were ever offered for sale separately, I but it in a heartbeat provided it was priced somewhere in the ballpark of Hydra.

I like FM8. Rhino 2, FM Heaven, and Blue 1.8. The 1st 3 can all import DX7 patches IIRC. Don't know about Blue. I missed picking up Sytrus during the I-L sidewalk sale, unfortunately. It's a good one as well.

toxic 2.5 sse2 version is in synapse synth pack now.

[DrWashington]

JarJar, what I meant by 'processing' is signal path, resolution, shortcuts, the whole thing.

For instance, the other night, I spent a long time playing with Reaktor, which I like very much. Then, I started playing with Sylenth1, and I just started shaking my head in disbelief.

When Reaktor has a VA style ensemble loaded, it can sound quite good. Switching over to Sylenth, I remembered instantly why I tend to try to coax more out of it than any other synth: it sounds PURE. There's very little that can be introduced to the signal chain in Sylenth that can significantly damage the final output in terms of adding nasty artifacts, questionable noises like hard to isolate clipping, etc. It can be done, sure, but it doesn't usually happen without intent.

Sylenth1 is a very, very well programmed synth. I mean, this is, to me, the very pinnacle of what a softsynth should be. A few more features like wavetables and PWM, bigger mod matrix, etc., and this would give a certain famous German hardware company a major run for its money.

Zebra is great too, but quickly gets very proc hungry, so I tend to use it a bit less. The sound is ace, though. I love it.

***

So, back to the plot: whatever "Sam" has done with the signal path in VOPM is very, very true to the YM2151. I think my RME's DACs might have something to do with it, but FM8 is running through the same thing, and, again, there's a qualitative difference. It's not as flexible as FM8, but it's a lot harder to introduce digital whispies (artifacts). It's a great, great engine, even if it ain't very easy on the CPU. If this engine were reverse engineered and expanded, I'd be really pleased. The envelopes, esp., are very sharp and stabby. I like its simplicity, but I at least want portamento and MIDI learn.

[herodotus]

"blurred but harsh.." certainly makes sense to me, and I've heard the same description and close variations for many years, about sounds for which I have similar feelings. And it can be explained without voodoo, although for a musician even when it can't be explained scientifically, so what.

No it's not all subjective fuzz. Especially because you are a developer I think you should not let yourself be mislead by internet fog-faeries prancing about jeering at everyone who is merely using completely normal musical terms in normal musical ways.

Normal musical terms used in normal musical ways are filled with subjective fuzz.

And the idea that 'blurred but harsh' means the same to you that it does to everyone else is a matter of faith.


Interesting fact: color blindness (the vision problem) was noticed for the first time by John Dalton in the late eighteenth century.

Just think, before then, everyone just assumed that when they said something was 'red' that everyone else saw the same color they did.

hmm...

[mhemnarch]

No, blurry and smeared refers to the dynamics. The transients are smeared. Sounds like jitter, you get harshness as well as losing focus/definition. It's no good to explain this, you need to hear it.

What does it mean for a transient to be smeared, in exact terms? It sounds as if you've spent some time considering exactly what you mean by these phrases, but I've never heard them used in this way before and am having trouble understanding what you're saying.

[living sounds]

No, blurry and smeared refers to the dynamics. The transients are smeared. Sounds like jitter, you get harshness as well as losing focus/definition. It's no good to explain this, you need to hear it.

What does it mean for a transient to be smeared, in exact terms? It sounds as if you've spent some time considering exactly what you mean by these phrases, but I've never heard them used in this way before and am having trouble understanding what you're saying.

Think of the waveform display in an editor. Imagine a transient, a sharp, short peak. Then "smear" or "blur" it horizontally. Slightly. What was once a very precise peak is now broader and less defined. This time domain change affects the frequency domain as well, they're always linked. The very slighr "smear" in the time domain results not only in less defined dynamics, but also registers as buildups of distortion, an amplification of certain frequencies, especially in the case of the high amplitude, high frequency transients. Many people might call this more or less subtle, but to the trained ear it is easily recognizable. Once it's in the signal, you won't get it out, and it will become more prominent with processing and make it much more difficult to mix the signal.
The solution to avoid it are high end converters as well as sources and processing without the artifects. Like VOPM.

[Aroused by JarJar]

"blurred but harsh.." certainly makes sense to me, and I've heard the same description and close variations for many years, about sounds for which I have similar feelings. And it can be explained without voodoo, although for a musician even when it can't be explained scientifically, so what.

No it's not all subjective fuzz. Especially because you are a developer I think you should not let yourself be mislead by internet fog-faeries prancing about jeering at everyone who is merely using completely normal musical terms in normal musical ways.

Normal musical terms used in normal musical ways are filled with subjective fuzz.

And the idea that 'blurred but harsh' means the same to you that it does to everyone else is a matter of faith.

Please stop making these straw men- "everyone else"? how could I possibly know what "everyone else" thinks, even in the highly unlikely case that there is anything on earth "everyone" agrees about. ( other than the undeniable universal truths like "the Powerpuff Girls rule")

Interesting fact: color blindness (the vision problem) was noticed for the first time by John Dalton in the late eighteenth century.

Just think, before then, everyone just assumed that when they said something was 'red' that everyone else saw the same color they did.

hmm...

Aside from wild and baseless assumption contained in the above... er, let's see what would be a classic bit of instruction here, hmmmm.... yes, the Shakespeare sonnet to his less-than-lovely-love would do fine....

Anyway, are you saying: "a word doesn't mean the same to everyone, therefore it means nothing to anyone."

See what I mean about "fog-faeries?"

Okay, knock yourself out. I have to work with other musicians, engineers and artists and in my world, when someone says "make the bright silvery sound louder", and one tone is a sine at 100Hz, and the other tone a trianglish tone with slightly inharmonic partials and a percussive envelope at 900Hz... um, let's see which one they're refering to...

anyway, carry on, it's quite amusing watching all the dancing around the issues and a maximal avoidance of discussing real measurable issues, or even subjective ones that don't fit the cliches.

In the China bell example for example you can hear that the tones have different attacks and different kinds of decay envelopes, there's almost a soft-clip kind of sound on the attack transient of the second one... "hear no significant difference" is the very subjective take on it, "they are different" the objective.

And it's the objective, for all the pious claims otherwise, which is what you clowns are actually trying to avoid with all this "everything is subjective" astrology shit, if I may be blunt. Otherwise I wouldn't be the only one commenting on the audible and measurable difference in envelope shapes in that example- for example.

[nuffink]

I have to work with other musicians, engineers and artists and in my world, when someone says "make the bright silvery sound louder", and one tone is a sine at 100Hz, and the other tone a trianglish tone with slightly inharmonic partials and a percussive envelope at 900Hz... um, let's see which one they're refering to...

Can I be the first to thank you for sparing the time from your busy schedule to instruct us. I know I'm learning a lot, I'm sure others are too.

[whyterabbyt]

anyway, carry on, it's quite amusing watching all the dancing around the issues and a maximal avoidance of discussing real measurable issues

oh yes, it is. of course, its actually those reliant on windbaggery like 'its more smeary blurry harsh' that are avoiding the 'real measurables', and folk like tony who are asking for them. funny that.

[tony tony chopper]

Think of the waveform display in an editor. Imagine a transient, a sharp, short peak. Then "smear" or "blur" it horizontally. Slightly.

It's the same as debates about audio engines & crap like that, you have to realize that it would be real work to make an FM synth smear transients or color the sound differently. It'd mean an EQ for the color, and to smear transients, I don't know, a noisy FIR filter or lots of allpasses. NI would ADD this, on purpose, to smear transients?

Again, FM is pretty simple. And you can divide it into different parts (again, the OPs, the envelopes & the algorithm). If you're really any interested in knowing where the difference comes from, you should test those parts separately. If an OP along produces just a sine in both synths being tested, then it's something else. The algo's maybe? Easy to test. You'll see it's the envelopes, and you will eventually be able to replicate a DX7 one in another synth with multipoint envelopes.

[living sounds]

It's the same as debates about audio engines & crap like that, you have to realize that it would be real work to make an FM synth smear transients or color the sound differently. It'd mean an EQ for the color, and to smear transients, I don't know, a noisy FIR filter or lots of allpasses. NI would ADD this, on purpose, to smear transients?

No idea what they do, I never looked under the hood, obviously. It just sounds smeared.

[nuffink]

I'd announce Sytrus 2 with 85% less smearing. Or, for only slightly more, comes bundled with a smear plugin. For that old-skool sound.

[MaliceX]

Smart question: Is there any way to replicate FM OSC feedback on a synth like Helix's signal chain? (ie: using basic modulator->carrier->carrier principles)

[tony tony chopper]

No idea what they do, I never looked under the hood

but you don't have to, you can start from a blank patch, and tweak stuff until it sounds smeared, then you know what does

[living sounds]

No idea what they do, I never looked under the hood

but you don't have to, you can start from a blank patch, and tweak stuff until it sounds smeared, then you know what does

I know the envelopes are part of it. Just a sinewave with minimal attack sounded way more punchy with the TG77.

But in the end this is for the developers to figure out. I'm busy enough working, studying, making music and building (analog) gear already to also look into this sort of thing myself. You make the plugins, you look into it!