Minimoog Softsynth Shootout: Diva MiniV3 Monark Legend Minimonsta vs Model D

[egbert101]

<delete>

[Urs]

Starskey Carr set up a thread on this video on a certain other internet forum (rhymes with beerguts )

I'm not sure I'm ready for that. People often come to conclusions quickly and each time I spend an afternoon banging my head on a table takes a toll. Doesn't necessarily mean it is like this, but I won't take the chance.

(my alternative programme for today is about extending the realm of analogue emulations quite a bit, as I'm feeling a bit more free for interpretation)

[chk071]

I'm afraid people are so stuck in the "golden age" of over 40 years ago, that something like a Sub-37 wouldn't nearly be such a seller.

[suthnear]

It should be remembered that there were no envelopes or filtering used in those examples (unless there is some fault with his hardware), so you're hearing the "squelchiness" and "liveliness" of the raw oscillators, plus any additional hiss or noise generated in the other circuits.

There was an envelope used, though. The VCA of a minimoog is always controlled by an envelope. You can see in the videos that there is generally more release to the model D: the taper of the tail of the sounds is less abrupt. Decay is a critical component to making sounds seem natural as very few acoustic sounds die abruptly. So this parameter on its own affects our perception of a sound significantly.

Just matching controls is insufficient for proper testing since none of the developers modelled his model D. It's been mentioned several times in this thread alone that different synths of the same revision can sound somewhat different. This is due to the fact the range of each control and the underlying parameter are determined by various electronic components and there can be significant variance in these (see e.g.http://www.learningaboutelectronics.com ... -tolerance). There will be even more variation where humans are involved (e.g. when setting trimmers and the like in the factory). Components are further affected by age, temperature and handling. Thus, you get two of the same synths of the same model and vintage that sound different where the controls are matched.

Filter cutoff and resonance are quite sensitive to small variations in controls with the mini's particularly so given that it involves hand matched transistors in a ladder configuration.

[suthnear]

(my alternative programme for today is about extending the realm of analogue emulations quite a bit, as I'm feeling a bit more free for interpretation)

Yes, please. I agree that we need to build on the lessons of 40 year old designs rather than trying to slavishly reproduce them. There are very, very, very few songs that call for a perfect minimoog in 2017.

[egbert101]

<delete>

[Funkybot's Evil Twin]

Thing is, even with Repro-1 we've seen reputable outlets do A/B tests where they got everything wrong. It's really annoying when these people destroy your product and don't even take the time to read the manual. Or maybe deliberately omit the options. It's a tedious job for us to rectify the results, say, when we create better examples and post them. Thankfully that phase seems to be over, but the next one is coming for sure. (I'm explicitely not talking about Mr. Carr, he did a pretty good job with the Pro-One shoot out, but we still had to clear some things up)

I mean, nothing's stopping you guys from posting similar A/B videos yourselves.

[egbert101]

<delete>

[Urs]

I mean, nothing's stopping you guys from posting similar A/B videos yourselves.

Yeah, hmmm... not sure if we're legally clean if we advertise for our stuff using vintage synthesizers to compare our stuff to. Even NI have only used a plain audio example without naming the source. We have thought about this a lot, but investing a week or two into this for a pro-level comparison... I'd rather spend it on more interesting things.

Also, after months and months of listening A/B with each and every parameter to tweak the software, it's last thing I would like to do.

Lastly, even if we did a great job there would still be somebody raising doubts. Like, after the results of the OB-8 preset shootout with Diva were released, people suggested the author had deliberately made the analogue synthesizer sound worse than Diva. There'll always be the smell of bias and selecting examples in our favour.

That said, there's a really well done shootout between Repro-1 and a Pro-One on a German forum. Once it turned out that it was impossible to say which was which, the naysayers suggested that the Pro-One used was in bad shape and a properly set up Pro-One would sound much, much better. Yet none of the guys managed to provide any counter examples of their glorious pixiedust Pro-Ones.

So there it is. Software can not win ever. There'll always be an argument against it. Waste of time.

- U

[Funkybot's Evil Twin]

Urs, all good points. I appreciate those types of A/B comparison videos, but then again, I'm not one to go looking for excuses about why those types of comparisons are invalid (as long as they're more thorough than, "here's one sound on both, see? The same!"). This Minimoog comparison, and the recent Pro-One/Repro video, if anything, just confirm that good software is finally "there" in terms of authenticity. As I mentioned one a prior page, I think if someone played the same piece of hardware twice but told people one was software, there'd be claims of "close but not 100% there yet." Some folks WANT there to be a difference, other people WANT there not to be, but for most (outside of forums anyway), I don't think it matters one way or the other, the only question is does it sound good?

Anyway, I'm all about these hyper-detailed emulations of the classics, especially some of the more obscure ones and will buy anything you guys put out in that vain. I look at these types of offerings as being similar to those by boutique guitar makers who do hyper-detailed recreations of classics (guys like Danocaster or Ron Kirn), the difference being that I can afford your wares (those Danocasters are still in the 2-3k range). Just keep putting out great products and stay off the forums.

That said, the argument in favor of you doing an A/B video would be simple: I have no doubt that you're way more knowledgeable on these synths than your users.

[egbert101]

<delete>

[egbert101]

<delete>

[chk071]

I think what surprised me most was how much roasting Monark got, considering it was the "king" of minimoog emulations before The Legend came about. I still like Monark, and I think it sounds very analog, but I would probably prefer The Legend as my first choice (once I buy the thing).

Dunno. Maybe it's just that Monark wasn't as close to that particular unit than Legend or Diva. Monark sure sounds good to me. But then, i don't own a Minimoog. The A/B comparison demos NI posted were absolutely spot on, especially with aggressive, extreme sounds, which is where a lot of virtual analog synths fail, i think.

[Faza]

For Diva we decided to make things "sound nice", for Repro-1 we decided on "let there be bugs", but then we chose the "ideal" VCO preset as default anyway. Had we done Diva today, we would have probably added a few more gritty options.

In my opinion, such options are the perfect and best possible decisions. I think that every synth emulation should offer a "ideal" preset, that will be free defectless of defects considered as bugs and modelled first by circuit model, then allow to choose such defects individually per preset. It's like when they produce the hardware synths, they do everything by blue print and then these artifacts appear, because of the nature of the hardware synths and nature at all.

Yes, but then, implementing all these options costs a lot of time but returns little value. Add to that, it may make the software look more complex and complicated than it is, when the target audience looks for a simple and classic architecture. And what for, really? Just to win a shoot out?

Thing is, even with Repro-1 we've seen reputable outlets do A/B tests where they got everything wrong. It's really annoying when these people destroy your product and don't even take the time to read the manual. Or maybe deliberately omit the options. It's a tedious job for us to rectify the results, say, when we create better examples and post them. Thankfully that phase seems to be over, but the next one is coming for sure. (I'm explicitely not talking about Mr. Carr, he did a pretty good job with the Pro-One shoot out, but we still had to clear some things up)

In essence, if it makes a synth a better instrument, I'd rather sacrifice the chance to win a shoot out than water down the concept. After all, each software emulation is also an interpretation. People expect enhanced features, and they also expect flaws of hardware to be fixed (tuning...).

That's a very good point. I hadn't thought about it this way. Then it's perfectly fine just to skip the defects.

[egbert101]

<delete>