Why even modern VST Synths can't sound like 20 year-old Hardware VA Synths?

[recursive one]

I actually think the term "sound quality" should be dropped here. Most software synths these days have adequate sound quality in purely technical terms, and so do the hardware synths, otherwise they would not have become so popular. It's about the differences in character. How does a synth behave in a mix, does it tend to sound soft van delicate or rough and edgy, etc.

[BONES]

As an oldschool trance lover and an Access Virus owner I can totally see where you are coming from. The demos you posted sound awesome to my taste and I agree that it's not really that easy to get all these sounds form software with 100% authentic character - you can do something kinda similar with Spire or Dune, but then you process it, layer it, compress the hell of it and then it doesn't really matter what was the original synth

I was thinking something similar about the sounds in those videos. I only scrubbed through but all the sounds I heard seemed terribly generic to me, like I could use any old synth for them. I didn't hear anything that actually impressed me, that would make me want to buy any of those instruments.

But i think it's not the matter of the lack of skills in the developers, rather the lack of interest and demand. These sounds went out of fashion 5-10 years ago, trance sounds different these days, the sound of Spire (mainstream trance) or Serum (psytrance) is the new norm.

I find that a very hard concept to wrap my head around. I can't think that Spire and Serum sound so different from one another that they wouldn't be equally useful in two different styles of the same genre, let alone dozens of other genres. After all, they are synths, they should be hugely versatile and the areas where they are different would be a tiny fraction of their overall ability. And by tiny I mean probably 99% the same and 1% different. The whole concept of "if you want to make this music, use that synth" is ridiculous to me. I use all kinds of different synths to make the same style of music and I doubt any of them were made with what I use them for in mind. So does this come from a lack of ability/knowledge or because the presets these thigs ship with are slanted more towards one thing than other things?

[recursive one]

Serum is good for FM and wavetable manipulation, Spire does niser soundsing supersaw. Therefore psy guys use Serum more often and trance guy often prefer Spire. That's it.

[freshalex]

I think the first point is that hardware synths have defined the sound. Software came after them, so software has to sound like hardware - not the other way round. So hardware sounds real and software sounds a bit different.

Another important point is that hardware manufacturers simply made better products because they only had 'one shot'. So all details were better thought out and rounded off. And it was 'squeezed out the last bit of sounds and effects'.

Exaggerated: If you take a hardware synth and turn off all the effects and tuning, it usually sounds like a cheap freeware trumpet

[anomandaris1]

Serum is good for FM and wavetable manipulation, Spire does niser soundsing supersaw. Therefore psy guys use Serum more often and trance guy often prefer Spire. That's it.

Serum is actually terrible for FM (unless you want very basic sound effects type timbres out of it), but it can actually do great supersaws etc. There are many-many soundbanks for trance/house etc for it.
Spire can also weird sound effects and dubstep noises, but it takes more work. I don't think that there are many such preset banks for it.
I would say that preset designers and buyers probably defined what are the synth niches. (Just like after Hans Zimmer "Dark Zebra" presets were released, the synth became the "go to" for all movie trailer/underscore composers - along with Omnisphere, of course - with most new preset banks in the same vein)

[recursive one]

Serum is actually terrible for FM (unless you want very basic sound effects type timbres out of it)

Depends on what kind of FM you need (and your skills i guess). I have no problems with making nice FM sounds in serum for my needs.

Sure it can do decent supersaws but it's somewhat easier in Spire.

Ofc almost any synth can be used for almost any genre, it's just certain things are easier done in certain synths (less tweaking, less processing). If your goal is to make actual tracks it's important.

[anomandaris1]

Serum is actually terrible for FM (unless you want very basic sound effects type timbres out of it)

Depends on what kind of FM you need (and your skills i guess). I have no problems with making nice FM sounds in serum for my needs.

Yeah, and you make psytrance, right? I am sure that your definition of "nice FM sounds" has nothing to do with mine. For hyperrealistic instruments, even 8 operator synths may be too limited. The only way to get somewhere close to this with Serum is to use the resynthesis option. No way to get close with wavetable or FM, or manually with its limited additive engine.

[BlitBit]

Bones already mentioned hardware processing factor. I've heard an opinion that a lot of character of hardware gear is in digital-to-analog converter, which is obviously non-existent in virtual synths and, either way, tricky to emulate as far as I know.

Should be resolved by running your softsynth through a digital-to-analog convertor before listening to it, then.

Yes, you could for example use the digital-to-analog converter in your audio interface to do that.

Everything that comes out of the DAW goes through a DAC so that's very likely not a factor if there even is any in the first place.

[clipnotic]

Modern processors have computing power like supercomputers of 2000
But almost all modern VST synths sounds "lifeless" and thin..

I worked much with analog and digital hardware and meanwhile I don't miss them. That statement is nonsense, because many VSTs have much more power than much hardware. It's not the tool, it's the user!

[ShawnG]

Bones already mentioned hardware processing factor. I've heard an opinion that a lot of character of hardware gear is in digital-to-analog converter, which is obviously non-existent in virtual synths and, either way, tricky to emulate as far as I know.

Should be resolved by running your softsynth through a digital-to-analog convertor before listening to it, then.

Yes, you could for example use the digital-to-analog converter in your audio interface to do that.

Everything that comes out of the DAW goes through a DAC so that's very likely not a factor if there even is any in the first place.

I think I read somewhere years ago (probably here) that a some of these early 00s VAs used cheaper DACs that were used on mobile phones rather than the ones on the audio interfaces (which obviously vary themselves) so that might color the sound at least a bit? dunno

[BlitBit]

Bones already mentioned hardware processing factor. I've heard an opinion that a lot of character of hardware gear is in digital-to-analog converter, which is obviously non-existent in virtual synths and, either way, tricky to emulate as far as I know.

Should be resolved by running your softsynth through a digital-to-analog convertor before listening to it, then.

Yes, you could for example use the digital-to-analog converter in your audio interface to do that.

Everything that comes out of the DAW goes through a DAC so that's very likely not a factor if there even is any in the first place.

I think I read somewhere years ago (probably here) that a some of these early 00s VAs used cheaper DACs that were used on mobile phones rather than the ones on the audio interfaces (which obviously vary themselves) so that might color the sound at least a bit? dunno

If this was true then it would fall into the "sounds different due to technical restrictions" category.

To be frank I don't really know how DACs are actually implemented in hardware. If someone knows and wants to share this knowledge that would be great. If I remember correctly an ideal DAC would convolve the digital signal with a sinc function which would produce a perfectly bandlimited (at Nyquist frequency) and continuous waveform. Because this is not possible you always have to go with compromises which will always result in having to let go of the perfect bandlimiting which in turn will introduce aliasing.

The main problem is that OP just dropped some videos where the synths are drenched in effects and most importantly without providing any information about the full signal chains. One way to drastically alter the sound of an electric guitar through an amplifier is to change the cabinet (or the impulse response if you are in the box). So if these synths are sent through a (power) amp and then through a cabinet this will drastically alter the sound. The microphone used to mic the cabinet in turn will also change the frequency response again. So, after all there are lots factors at play here and it does not make sense to compare hardware which might go through a long signal chain until it arrives in the box to soft synths which are in the box immediately.

I guess I will now go and experiment a bit with sending some soft synths through different cabinet impulse responses to see how it sounds.

[recursive one]

One thing i noticed about the Virus Ti was that it had certain "hardness" in its sound. Don't really know how to explain, the sound is really punchy and coherent. It sounds like there's maybe a very subtle waveshaper or mutliband compressor somewhere in the signal path.I'll try to provide some examples later.

There's no DAC involved as the sound goes over the USB.

Some softsynths do have similar qualities, like Repro or Massive X.

[Jeff McClintock]

Why even modern VST Synths can't sound like 20 year-old Hardware VA Synths?

Blind listening test or stfu.

Seriously, someone please post the audio from a 20 year old synth, and also post audio from a (good) VST emulation (like Arturia or similar). Both though the same signal chain. But DON"T TELL US WHICH IS WHICH.

Let see how the armchair experts fare.

[christian f.]

I can name hundreds of "classic" tracks with outstanding synth sounds , sounds that were the reason emulating vintage synths became the trend and possibly even the biggest hook to sell soft synths. Those tracks were made in the 80s and 90s. Can you name some note worthy. "classic" tracks with outstanding synth sounds from the past 20 years ? The trend on kvr now, Arturia OBXa , upcoming Synapse. OBXa , Arturia vs synapse OBXa and even polls about this. I think its self explanatory. If soft synths were so great, we wouldn't even be talking about vintage analogs vs softsynths anymore or still comparing originals with emulations.

[BONES]

You can hear the difference. It's not the kind of difference that would make one of them much better than another one or matter a lot in an actual mix but it's here. I'd say Virus is thicker and in your face, the sound keeps more integrity upon modulations. Minuscule differences but they can be heard and probably can be meausred.

Where I can hear a difference it is exactly that - a slight difference, like the patches aren't quite identical. I certainly can't hear that one sound consistently better or worse than the other, just that some sounds are closer than others. They both sound great to my ears. I'd probably buy Viper if I ever saw it on sale.

Perhaps the higher performance of modern digital systems is precisely the problem.

I won't quote your whole post but if what you are suggesting is true, then there would be plenty of people still holding on to their low-res VSTi from 20 years ago but I think we all acknowledge that VSTi get better and better and nobody remembers the old days of Superwave P8 or Synth 1 with anything other than nostalgia. Although we may still have some respect for those instruments, the caveat is always "for back then", it's never absolute. i.e. I have heard plenty of people say how much they still like Synth 1 but nobody has ever tried to suggest it sounds better than Thorn or DUNE or Serum.

People are like ducklings. We “bond” to the instruments we were exposed to during our freshly hatched state.

Speak for yourself. There is nothing from my under-25 years that I would rate at all. Korg Delta would probably come closest, closer even than the Mono/Poly I had, but no-one is ever going to make an emulation of that and I don't blame them. All the persistent memories of those days are of frustration, not joy.

ESQ-M was the first synth I owned that I would still rate today and be interested in a VSTi version of and I was probably 28 or 29 when I got that. My first V/A was the MOSS board on my Trinity but I wouldn't rate that above the ROMpler CS1x I had at the same time (around the time I turned 40). Korg Prophecy, OTOH, felt far more magical and I might be interested in a VSTi version of that. I had a K Station and I love my V-Station - getting great sounds out of it is effortless. I also had an Alesis Micron but I don't know that I'd want a VSTi version of it.

The one V/A I know I would absolutely, 100% want as a VSTi is Ultranova and I was in my 50s when I got that.

Late 70s this was my centerfold:

This is the one and only vintage synth I have ever lusted after. But when Korg re-released it without patch memory, I totally lost interest. Korg's VSTi is all the Odyssey I need and I love it.

Not everyone want's to sound like Hans Zimmer. The fact that Zebra was used on Dark Knight doesn't automatically make it better than any synth that was not used at Dark Knight.

You don't have to want to sound like Hans Zimmer to have some respect for his experience and ability and, therefore, the decisions he makes around his process.

Serum is good for FM and wavetable manipulation, Spire does niser soundsing supersaw. Therefore psy guys use Serum more often and trance guy often prefer Spire. That's it.

But if you can load your own wavetables into Serum, then presumably you could load in the world's most awesome supersaw sound and beat Spire's, couldn't you? And FM is FM, the only thing that would make one different to another would be the actual functionality. e.g. DUNE's multiple algorithms will beat TRK-01's simple FM a lot of the time.

Ofc almost any synth can be used for almost any genre, it's just certain things are easier done in certain synths (less tweaking, less processing). If your goal is to make actual tracks it's important.

Absolutely. Agree 100% but I don't see how it necessarily applies to two synths that are far more similar than they are different. e.g. I can see how you might not wat to buggerise about patching up an FM synth in PhasePlant but I can't see much that's different between Serum and Spire, which makes me think it probably does come down to presets.