Diva Vs. Real Analog

[Z1202]

Simple example: http://s000.tinyupload.com/index.php?fi ... 1233960507

Can you reproduce anything even remotely like this entirely generic sounding typical filter in software?

http://vk.com/wall161447783_2955
The first sound should be pretty close to the original example, the others are more random variations. Notice that I have no idea which filter was demonstrated by aciddose above and which settings were used

[aciddose]

This sounds exactly like my own failed attempts unfortunately

It has too much odd harmonic vs. even and doesn't exhibit any feed-thru or the other more complex non-linear effects in the signal path.

The filter I used is identical to the KORG35, only reverse the direction of the transistors and remove the biasing circuit (use a single resistor instead, it improves key-tracking somewhat.)

It also uses a typical output -> high-pass -> r voltage divider -> diode clamp feedback circuit rather than one like the MS-10.

[wagtunes]

It's all about the music and what works for YOU...PERIOD.

Everything else is just BS posturing.

[aciddose]

Of course it's about what works for me. Software isn't the same and I find it really offensive when people claim "the differences are tiny" because for me they're absolutely huge. It's like being stomped on the face by godzilla!

[Mutant]

Simple example: http://s000.tinyupload.com/index.php?fi ... 1233960507

Can you reproduce anything even remotely like this entirely generic sounding typical filter in software?

Here, the first 8 (too lazy to do more).
https://app.box.com/s/mjbwb9momgdcgcr0m0ahdan47w48p2fz
As always i am not claiming it is a perfect match (and it some cases even the waveform looks totally different), but for my purposes it is close enough.

Just happened that my mother was around, i gave her the headphones to listen to the clip and asked just 1 question.
"Can you tell me what it is ?"
The answer was "It is a synthesizer".
Singular, not plural.
She is not only not a synth nerd, but also quite old, so maybe not the best example of how non-nerds can't tell the difference and that there is software on one side and hardware on the other side.

[aciddose]

Identical problem as Z1202's, although also sounds radically different (yours has a certain weirdness I can't even identify the source of), I immediately knew the left channel was different without even having to mute the right.

Did you apply a distortion/waveshaper on top of the left channel?

[Z1202]

It has too much odd harmonic vs. even and doesn't exhibit any feed-thru or the other more complex non-linear effects in the signal path.

As for the odd harmonics, most of the example I posted was driven by a square wave anyway. Could you point out to a specific time in your example where the even/odd harmonic balance is to your taste? Also, what is the feed-thru effect you are referring to?

[chk071]

How would you be able to replicate that sound if not even the filters or waveforms of the respective synths are modelled after those of the synth the sample was used from? Especially when the goal is "to really nail it". How, if the filters and oscillators differ? And where is the musical use of all this? Or the proof that analog is better than virtual? And what does that even mean? If analogs are as different as is always proclaimed, you'll even have a hard time replicating the sound with another model of synth XY being used for the sample.

[aciddose]

Just look at the shape of the resonance, it's nearly a saw relative to yours.

Also the sound of it, it "latches" harmonics far less giving it a "smoother" timbre without even needing to realize the even harmonic content is far greater. I'm speaking of the content on the oscillation from the filter, not the input. The input is irrelevant to this.

I'm speaking from the point of view of having spent thousands of hours attempting to duplicate this with all the knowledge and experience available to me. I have better results with my own software than these posts... I don't mean to say you haven't invested the same or greater effort, just that this is by no means simple to achieve in my opinion.

I find I would rather attempt to run in front of a bus than claim I've been successful in coming anywhere near the example I posted with any of my attempts.

Although I was motivated in my efforts by some posts by Shy and another member I don't recall the username of (he demonstrated his own work toward achieving the same goal which radically changed my view on the possibility of achieving this. Possible, although difficult.)

[Urs]

hehehe, threads take funny routes sometimes...

@aciddose: I don't find it difficult to replicate the behaviour of the Korg35 based filter in an MS-20 Rev1. We got the HP wrong, but the lowpass is spot-on. It's more difficult to replicate an unbuffered, diode-based Sallen-Key such as the Steiner-Parker version.

[wagtunes]

Here are my personal pros and cons for both hardware and software synths and why I choose Diva over any hardware synth in the final analysis.

Hardware

Pros

Like being able to touch the synth and manually work the controls.
Like the sound

Cons

Hate the space they take up. 100 plugins (no space) 100 analog synths (a whole house)
Hate having to bring them in for repair
Hate all the cables you have to have, not to mention your hardware based audio system since everything is OTB
Hate my electric bill at the end of the month after having these things on all day everyday.
Hate having to actually have to physically go down to the store to demo these things.
As a rule, more expensive than VSTs

Software

Pros

No space taken up
No repairs
No cables or plugs
No electric bill if I have 100 plugs activated inside a project outside of my PC being on.
No having to go down to store to demo. Can download and do in minutes and then buy immediately.
Perfectly fine sound.
As a rule, cheaper than analog synths.

Cons

Can't touch them. Don't have that "cool" factor.
If OS crashes, have to reinstall every single one. Not looking forward to this the first time it happens.
Some OS or DAW conflicts. Buggy at times. Something I've never had with an analog synth.

Final Analysis

Taking all of the above into consideration, VA synths win hands down. There is no contest for me. I don't have the space or the money for all the analog synths that I would want.

But like I said, for each individual it's what works for them.

[Mutant]

Did you apply a distortion/waveshaper on top of the left channel?

Tube driver only on one of the 8 notes, the other 7 are pure Xhip.

I immediately knew the left channel was different without even having to mute the right.

Sure, as i said, they are not a perfect match.
But i think that if you added the usual chorus, delay, reverb, EQ, comp, limiter, then not even your superior ears would be able to spot the difference.

And the even/odd harmonic problem: yes i noticed it, i tried to tweak the oscillator, but i only got even more odd harmonics. I failed there.
And there is the high part of the spectrum where on your analog there are some "waves", impossible to replicate.

[aciddose]

hehehe, threads take funny routes sometimes...

@aciddose: I don't find it difficult to replicate the behaviour of the Korg35 based filter in an MS-20 Rev1. We got the HP wrong, but the lowpass is spot-on. It's more difficult to replicate an unbuffered, diode-based Sallen-Key such as the Steiner-Parker version.

Can you demonstrate that? My attempts have all been a miserable failure to make anything coming anywhere near run in real-time without significant aliasing.

[Urs]

hehehe, threads take funny routes sometimes...

@aciddose: I don't find it difficult to replicate the behaviour of the Korg35 based filter in an MS-20 Rev1. We got the HP wrong, but the lowpass is spot-on. It's more difficult to replicate an unbuffered, diode-based Sallen-Key such as the Steiner-Parker version.

Can you demonstrate that? My attempts have all been a miserable failure to make anything coming anywhere near run in real-time without significant aliasing.

It's in Diva, the Sallen-Key type is called "Bite". Try "Divine" mode.

Alternatively, try Cytomic's "The Drop" for a probably even more accurate model.

[aciddose]

https://soundcloud.com/aciddose-1/sallenkey

Does it cover the crossover distortion from the two amps and the feed-thru both signal and CV, forward via the feedback path?

I'd be amazed if so because that is super expensive.