Hardware vs. Software aka why do VST sucks?

[farlukar]

I want a VST that emulates the original hardware DSP warmth.

[D-Fusion]

Well i don't think that vst's sucks but there is some difference, take the best vst instrument you have and create a loop then add a hardware unit to play along and you will see what it's all about

Okay, done that. I actually tried it with a couple of different VSTs but they all still sounded better than the Barbie keyboard, so Im not sure what your point was.

I give up

[beej]

You know, in terms of synths anyway, I'm more and more convinced that the reason some people think software synths don't sound as good as hardware is because of one of the strong points of using a computer - the display.

More and more our attention is focused on the screen, and you are simply not responding to the sound with as much concentration as when you are just playing a hardware synth.

Try this: load up you favourite (decent) synth, assign the patch up/down buttons to buttons on your hardware controller, and some of the more meaningful synth parameters (ADSR, filter, waves etc) to knobs, and then TURN THE COMPUTER MONITOR OFF - and just play...

Did you have fun..?

(Also, don't underestimate the talent of the voicing teams from the larger companies like Rolayamakorgaha - they put serious money into voicing their instruments that most softsynth developers can't afford to do...)

[projectdan]

They say it mostly because they want hardware to sound better. They will use various arguments to support their views.

Hardware is more powerful - From what Ive read this is not true in most cases, though you developers should know more about the facts of this.

It costs more because its better - obviously a flawed argument, but I think it can be a factor that has a lot to do with it for many.

Its from more professional companies who know what they're doing better - Depends on the exact comparison.

[TRN76]

hehehe @ snooky
Maybe you should start using photoshop instead, dont you feel embarrest when you say stupid things?
No, I dont know the answer to "why VST sucks", because its a myth. I have several VST/i here that are awesome. All I'm saying is that there are still a few hardware that no software still havent archived.

[galwaypark]

Well, these dsp processors in the hardware synths run way slower than the ones in your pc. And they're much cheaper. They are just good enough to run 1 instance of the synth, with a limited number of voices, but not more. The only hardware synth I know that uses a comparably fast processor is the Neuron. And you know how expensive that was

Cheers,

Urs

Ah, I see. Thanks for clearing that up for me. I always assumed that, in this day and age, they would use much more powerful processors.

[Warmonger]

hehehe @ snooky
Maybe you should start using photoshop instead, dont you feel embarrest when you say stupid things?
No, I dont know the answer to "why VST sucks", because its a myth. I have several VST/i here that are awesome. All I'm saying is that there are still a few hardware that no software still havent archived.

Oh? I've seen plenty of software that does stuff no hardware has yet achieved.

[Ronny Pries]

Does a one legged duck swim circles?

[Kingston]

Does a one legged duck swim circles?

Initially, yes, I've seen it. Swans, too!

Poor little things. (no I didn't cause it!! )

[Kraku]

Oh how I would love to post again my KLC Polysix vs. Virus TI oscillator/unison comparison. People thought that there was something severely wrong the way I constructed the unison sound on the TI (it sounded weak and strange with the same settings as the Polysix).

[TRN76]

Sure Warmonger, me too. It wasnt what it does, it was how it sounds.
Keep on discussion guys, im sure you do better music after this.. out

[Spaceman Sounds]

After 9 months of recording nearly 12,000 samples of rare analog beasts for my sample cd I have to say that virtual analog vsti's...whilst good and very usefull (I use them all the time still for producing tracks), do NOT sound like real analog synths. They just don't.

I'm not saying vsti emulation are not worth it.... they are. But if you really know your synths and have good monitors ...dammit you can hear the difference on a pair of crappy pc monitors..... then the first thing you notice are the osscilators have nowhere near the power of a real analog oscilator. The filters do not sound the same and the whole warmth is not really captured.

We put the MS-20 vsti up agaisnt the MS-20 we were sampling at the time. They are not the same machine, running at 96khz help the vsti a lot in the bass end but even then it just wasn't the same. The same with the minimoog vsti emulations, we had 2 Moogs to a/b test against the vsti's. An old min i moog and a midimoog (rack mounted mini moog midi'd up)..... not even close.

Don't get me wrong I love vsti's, but we are a long way off from being able to really get the power of them in a vsti or dsp. It's like having skimmed milk on cereal instead of real milk. I know a lot of you will say I have no idea what I'm on about, I think I have good ears for this, and it's such a difference that anyone can hear it.

In the context of a mix a lot of the time you wont need to hear any difference, but for things like basslines and lead lines that are more naked vsti's lack the power. But I do love things like the Imposcar, Vanguard (although I have to use them at 88.2/96khz to really appreciate them). I love my uad-1 card and it's compressors... to me they are much closer to analog than vsti instruments. There is a difference between a real 1176 and a vsti uad-1 1176.... but not so big a difference as between say a real moog and a virtual one.

The best test is to just use 1 osc on a real analog and the same on a vsti. Play it really low down on a bass line and compare the vsti version with the analog using 1 osc on each. No fx no compression and maybe test the low pass filter with say 50% resonance.... move through the resonace on each... see which one makes the floor shake and which one doesnt. Notice how one needs to be much louder than the other to have the same effect in a mix.

Eq's are coming along nicely, they are sounding more and more analog. The Urs eq's are amazing at higher sample rates. Look how much power they use to get that close to the real thing. Now think about what an anlog synth filter is... it's just an analog eq really.... if you were to use the same attenton to emulation as say urs did with their eq's on the analog filter emulation, you'd soon be running out of cpu very quickly.

I don't think pc's have enough power yet to really compete. It will happen soon though

[shanecgriffo]

i rekon its just because its got knobs, and those hardware guys are real knob twiddlers

[Warmonger]

This thread is not about analog versus VSTi. It's about digital hardware versus digital software.

[living sounds]

I think it has - at least partly - to do with fixed-point processing. There are papers stating that floating-point loses precision, especially in the bass region. All hosts but one use floating-point - only Sawstudio doesn't (and is claimed by many to be superior soundwise). The Protools-plugins calculated on hardware chips (at 24-bit fixed point) do sound different than their RTAS counterparts (and arguably better, more defined and transparent, in my opinion).

My Lexicon reverbs also sound more defined and clear than the CSR - and the developer himself claims that his versions on the Scope platform (A100 etc.) are different, use different ways of modulation for instance that would cost way to much processing power on AMD/Intel systems.

So the cuprit to me seems to be a) floating-point processing and b) the x86 processor architecture.