Softube Model 84 (Juno 106)

[chk071]

"I see a lot of issues in his comparison videos" what issues precisely?.

He only compares very simple patches (or the raw oscillators, which are very similar in most synths anyway... but even when he compares the raw oscillator sounds, there's often differences in terms of brightness). He doesn't match volumes precisely in many cases. He often matches parameter values, neglecting the differences. I've seen at least two videos where he didn't match the tune of filter keytracking precisely. He also often compares sounds where the resonance levels aren't matched (as I said... probably rather going by parameter value than by ears).

Again, comparisons are very tedious. I wouldn't want to do them as well. You can get a very bad impression, if the sounds aren't matched precisely. And, if you DO match them precisely, there's always the risk that you're trying too hard to match everything, also kinda invalidating the comparison.

Creating exact copies of hardware in software is not impossible, the Weiss stuff that Softube does is the exact same code apparently as was the Access Virus which was on the older PT systems.

We're talking about emulations of analog gear, don't we?

[woodsdenis]

The only “ analog” is the filter in a 106, The osc are digital. Are you saying that like a Minimoog. 106’s all sound different. In the vid he shows the osc waveforms on an oscilloscope as well as playing them for comparison. But apparently you can hear differences. If you really can then fair enough, perhaps you should hire your golden ears to software developers.

[chk071]

I don't know why you feel the need to get defensive.

It doesn't matter if the oscillators are DCO's. It's all running through a VCA and an analog filter.

[woodsdenis]

Not being defensive at all, don’t know how you get that from my comments. You just seem to be such an expert on such matters.

[chk071]

Not being defensive at all, don’t know how you get that from my comments. You just seem to be such an expert on such matters.

Find the contradiction.

[Rastkovic]

The only “ analog” is the filter in a 106, The osc are digital.

Nope, that's wrong.

DCO only means digital controlled oscillator, it's not a digital oscillator....

https://en.wikipedia.org/wiki/Digitally ... oscillator

[woodsdenis]

Not being defensive at all, don’t know how you get that from my comments. You just seem to be such an expert on such matters.

Find the contradiction.

Sorry can’t.

[chk071]

Then I'm sorry for you.

[woodsdenis]

Then I'm sorry for you.

Thanks

[vitocorleone123]

The only “ analog” is the filter in a 106, The osc are digital. Are you saying that like a Minimoog. 106’s all sound different. In the vid he shows the osc waveforms on an oscilloscope as well as playing them for comparison. But apparently you can hear differences. If you really can then fair enough, perhaps you should hire your golden ears to software developers.

While of course everything in the software Model 84 is digital, a DCO is NOT a digital oscillator. It's a Digitally Controlled Oscillator. It's analog.

[NikkiA]

Specifically, it's a ramp generator that is reset digitally when it reaches the right point for the desired frequency, the square wave is then derived from that by a simple analogue comparator.

The Juno oscillator design is an ingenious invention that is almost purely analog, and even the 'digital' aspect of the reset signal could in theory be done in the purely analogue domain.

[vitocorleone123]

Specifically, it's a ramp generator that is reset digitally when it reaches the right point for the desired frequency, the square wave is then derived from that by a simple analogue comparator.

The Juno oscillator design is an ingenious invention that is almost purely analog, and even the 'digital' aspect of the reset signal could in theory be done in the purely analogue domain.

Thanks! I learned something today. Nicely presented, too (short and to the point).

[cyberheater]

Specifically, it's a ramp generator that is reset digitally when it reaches the right point for the desired frequency, the square wave is then derived from that by a simple analogue comparator.

The Juno oscillator design is an ingenious invention that is almost purely analog, and even the 'digital' aspect of the reset signal could in theory be done in the purely analogue domain.

Yep. Totally correct. I’ve built a Juno dco saw core reset from an arduino. Very raw and buzzy.

[BONES]

I didn't, either. After the MiniMoog, I reckon it is the most over-rated synth going. But Model 84 is actually pretty good. Definitely worth getting while it's cheap.

So unless you think his ears are made of cloth,

That is exactly what I think.

So you can hear differences in that YT video that the rest of us can't? Seems unlikely to me. That's why it's not purely anecdotal - he is showing us and telling us, so if you think the YT quality might not be good enough to show up any important differences (I won't go into what a stupid notion that is), he is there with his studio monitors to confirm that what you're hearing is accurate enough. Like I said, I don't give a shit one way or the other but I find his video compelling. Someone would have to prove it wrong for me to stop believing that it's accurate. When you simply proclaim that it's not, that feels like your bias to me.

And as a researcher who has been in the business for 30 years now, I need no random KVR member to tell me what counts as anecdotal evidence or not.

Clearly you do because an anecdote is an unsupported story, whereas this guy supports his story with actual evidence - video with audio. It would be an anecdote if he simply posted on this forum that he'd done some side-by-side testing and they both sound the same but he does more than that - he backs it up with video and audio evidence so that we can judge for ourselves. We don't have to take his word for it, so it's definitely not anecdotal. He doesn't apply scientific rigour to his testing but I can't find fault in his testing regime, can you?

What do you think samplesizes are for?

Other things. What do you think he should do, buy 50 licenses of Model 84 to confirm that they all sound the same? Or 50 Junos to confirm that what he is seeking is unobtainable? i.e. That there is a greater difference between individual Junos than there is between any one Juno and Model 84? All that would show is that the entire debate is pointless (which we kind of know it is). If it doesn't sound like your Juno, presuming you have one, maybe that shows your Juno is busted?

A double blind ABX test with a valid samplesize would fare much better, but it better be tested with headphones too for me to think it is constructed properly.

No it wouldn't. Again, all it would confirm is that what is sought is unobtainable. The problem with a blind test is you don't get to see how hard they've tried to make them sound the same. You have to make a big assumption that they've done the best they can, which is a much bigger leap of faith than is required with this video.

No emulation sounds "the same" as the hardware. There's way too many nuances in a synthesizer's sound to capture them all digitally.

That's absolute f**king bullshit. FFS, The Legend even models the different behaviour between a 110V US model and a 220V international model. It doesn't get any more subtle than that. You get into areas where two different hardware units will sound less alike than the digital recreation, which means they have absolutely nailed the sound.

[fairlyclose]

You get into areas where two different hardware units will sound less alike than the digital recreation, which means they have absolutely nailed the sound.

This is true. People make the silliest claims about analogue and digital recreations. In reality for any of the old synths there is variation in sound, particularly with age but it sure was the case back in the day as well. It follows pretty simply that a digital recreation just has to sit within the range of possible behaviours of the analogue devices it aims to model.