Software vs. Analog in 2025 – Has the Balance Shifted?

[HAL76]

I also have the Softube and the Cherry Audio Junos and me personally I think that both don´t even come close.

I own the Softube Version, and the Arturia Version

I have to question anyone who claims the Softube Model 84 "doesn't come close".



But maybe you mean 99% exactly the same means doesn't come close. It sounds exactly like my Juno used to, but like all 40 year old analog synths with voice chips in various states of decay no two will sound the same, and if you replaced your voice chips with modern ones well that's not going to sound like an original either

Everyone has their own perception and perspective. That's unfortunately the way it is. Regardless of the fact that the J6 was mentioned earlier, it's become clear to me over the years that tastes and demands are as diverse as the way people live their lives. Some jump out of airplanes with parachutes, others rave about cacophagia. Quite a few don't notice much difference, even in wine and cheese—and that's OK. What's not OK, however, is these omnipresent lectures from people who, in my opinion, are colorblind. Just accept it. The audiophile community thinks differently, and none of us is alone. Some spend >50k on legendary tools. I take a more pragmatic view. I prefer Verve Analog Machines to a room full of tape machines, tangled cables, and maintenance costs. But the Juno isn't still so popular for nothing, and anyone who's truly experienced the "Sweet Spot Machine" knows they've never seen anything like it in software.

[D-Fusion]

He was quick to move on from the Envelope test and based on his look and talking around it made it look like he was trying to cover up the flaw in the Softube envelopes.

You can clearly hear how lifeless and weird the envelope on the software sound in comparison to the Juno if you go to 26:26 in that video.

I have the same issue with the filter envelope in their Model 72 too where the envelope makes the same weak sort of muted plaf plaf sounds that kills the energy of the filter and oscillators on sounds that use short decay.
The Legend and Legend HZ are around 98% there which is good enough for me since the difference is barely noticeable so i have no issue with recreating the same sounds and use that on a track instead of my Hardware Model D.

This is where many emulations fail where they only make sure that the filter and oscillators sound very close to the Hardware and skip the details on the Envelope behavior and use a standard generic software Envelope instead.

[IvyBirds]

He was quick to move on from the Envelope test and based on his look and talking around it made it look like he was trying to cover up the flaw in the Softube envelopes.

That's because as he said envelopes are boring as are "tests" of them

You can clearly hear how lifeless and weird the envelope on the software sound in comparison to the Juno if you go to 26:26 in that video.

How do I measure "life" and "weird" in a digital software product?

The envelopes in a hardware Juno 106 and in the Softube plugin version are just software, they are code running inside of a computer, they are inanimate they don't have a life to measure

[D-Fusion]

He was quick to move on from the Envelope test and based on his look and talking around it made it look like he was trying to cover up the flaw in the Softube envelopes.

That's because as he said envelopes are boring as are "tests" of them

You can clearly hear how lifeless and weird the envelope on the software sound in comparison to the Juno if you go to 26:26 in that video.

How do I measure "life" and "weird" in a digital software product?

The envelopes in a hardware Juno 106 and in the Softube plugin version are just software, they are code running inside of a computer, they are inanimate they don't have a life to measure

The sound does break up in a weird way in that example compared to the Hardware and he was very quick to move on to something else and it could probably be closer if he took the time to add a tiny bit of attack on the software so the harmonics could get some space instead of being squashed that way.
Correct envelope behavior is just as important as the filter and oscillator sound in a clone or emulation if you want a authentic sound.

That is why i like the Legend/Legend HZ the most compared to the other Model D Plugins outthere because it is 98% there for the Model D sounds i use the most and it sounds the same as soon as i add drums to a track.

[ghettosynth]

He was quick to move on from the Envelope test and based on his look and talking around it made it look like he was trying to cover up the flaw in the Softube envelopes.

That's because as he said envelopes are boring as are "tests" of them

Well then, I guess that's that. It's the kind of thing that goes hand in hand with these kinds of sloppy evaluations delivered with youtube star confidence. It's well known that the Juno 60 envelopes are a bit snappier than the Juno 106. It is because the 106 envelopes are done in software. However, don't misread that, it is a contextual distinction. At that time the relatively slow intel microcontrollers had a lot of work to do to sum up all of the modulation amounts and deliver them to the VCF/VCA. Slushy EGs are common problem with software EG synths of the day, Juno 106, Jupiter 6, Matrix 6/1000, and even more dramatic, the Matrix 12.

Today, of course, computers are much faster and it's generally not a problem to use low spec microcontrollers of today as envelope generators and certainly desktop CPUs, in general, shouldn't have a problem here.

However, regardless of the actual differences between the model and the vintage synth, it is sloppy analysis to dismiss EG quality as "boring" when evaluating said differences. The EG can have a dramatic impact on sound quality. That kind of statement is one of the reasons that I don't pay much attention to Starsky, he's not really technical or thorough enough for my tastes. I get it though, that accent, that confidence, it's convincing for many.

Interestingly, the IR3R01 EG IC used in the Juno 60 sells for $250 to $300 on ebay. I assume that these are being sold slowly to Jupiter 8 owners who need them for repair. These are not being remade, there is little demand for them in terms of repair and no reason at all to use them in DIY projects when the very excellent AS3310 is available.

[El°HYM]

Chimpanzees are also 98% there; yet they look & sound nothing like homo sapiens.

[HAL76]

While training I just remarked that this dicussion here has reached a point where it´s time to say sorry and leave things as they are.

If you still hea a difference between hardware in software there´s "something wrong with you", while if you have the main problem of software users that is that it "doesn´t sound as you want it to" and you ask yourself "which plugin can do that?" it´s all just your fault. [...]

that´s so paradox. Just as adults showing videos of influencers to prrove theirt claims.

The differences sum up btw. You often don´t hear that much of a difference if you play just a single sound. So if you ask yourself once again when something is wrong with your mix which plugin does the trick better think twice and don´t go on feeding the VST cheaters.

[IvyBirds]

Chimpanzees are also 98% there; yet they look & sound nothing like homo sapiens.

it's physically, and logically impossible for something to be 98% similar to something else and then be said to be nothing like it

Of course it's like it, it's by definition 98% like it

[IvyBirds]

While training I just remarked that this dicussion here has reached a point where it´s time to say sorry and leave things as they are.

If you still hea a difference between hardware in software there´s "something wrong with you", while if you have the main problem of software users that is that it "doesn´t sound as you want it to" and you ask yourself "which plugin can do that?" it´s all just your fault. [...]

that´s so paradox. Just as adults showing videos of influencers to prrove theirt claims.

The differences sum up btw. You often don´t hear that much of a difference if you play just a single sound. So if you ask yourself once again when something is wrong with your mix which plugin does the trick better think twice and don´t go on feeding the VST cheaters.

But that's not the issue is it. The issue is in July of 2025 no two 40+ year old hardware synths of the exact same make and model will sound exactly the same. Pretty close sure but not exactly the same, and that even goes for things like the envelopes on hardware Juno 106 units. This was the case also 40 years ago in July of 1985 also but even more so now as calibration and age of components has an effect and they all drift over time

Is the end goal of software synths just to sound exactly the same as hardware? If so which hardware unit are we talking about? If it's the Juno 106 which exact unit are we talking about when it comes to things like oscillators and envelopes and how was it calibrated, to what standard, and with what test equipment? Also what is the skill of the person doing the calibration?

The oscillators and envelopes of the Juno 106 for example are digitally controlled but you still need to make sure the computer is sending the right voltages and frequencies and the components are responding properly. Roland produced a manual that has information on how to calibrate them. You need a multimeter and an oscilloscope and they need to be able to measure very tiny differences in DC current something many software solutions fail badly at and are inaccurate, and of course hardware oscilloscopes also need to be properly maintained and calibrated

The person doing this calibration also needs to know what they are doing and you have to do it exactly the same across all 6 voice chips and all the envelope generators, then because the chorus is so important you have to make sure that is also calibrated and has the correct bias set which again is a calibration issue

Now if you or your tech has all of the right skills, tools, and test equipment to pull all of that off your Juno 106 will sound different than other Juno 106 units

In the end I would be perfectly happy with something sounding even 80% similar to a Juno 106 in plugin form, but the reality is my software probably comes to within 98-99% accuracy on the vast overwhelming majority of patches and 90-95% on the others and I am fine with that, as I am able to save thousands of dollars, not be limited to just 6 voices of Polyphony, and not have to deal with the failing components and calibration issues that 40+ year old electronics bring to the table

Now for some people that 1-2% difference on the vast majority of patches and 5-10% on the others is significant enough that they are willing to deal with the financial hit, the maintenance and calibration issues, and the risk of catastrophic failure of components to actually own a hardware Juno 106 in July of 2025, awesome then they absolutely should own one but let's all have the integrity to admit that that something that sounds 98% the same sounds 98% the same and it's incorrect to pretend it sounds "nothing" like the thing it sounds 98% like

[ghettosynth]

The differences sum up btw. You often don´t hear that much of a difference if you play just a single sound. So if you ask yourself once again when something is wrong with your mix which plugin does the trick better think twice and don´t go on feeding the VST cheaters.

Yep. I recall a thread on some forum, maybe a mailing list. Someone was asking about how to the general vibe of a Boards of Canada sound and the response was "a room full of analog hardware." Yes it was a bit flippant, but, we shouldn't dismiss it so quickly in the same way that we don't dismiss the sound of analog consoles and outboard vintage gear. In isolation emulations are very convincing. However, if that were all there is to it, then why do we need plugins to simulate subtle channel differences?

[HAL76]

The differences sum up btw. You often don´t hear that much of a difference if you play just a single sound. So if you ask yourself once again when something is wrong with your mix which plugin does the trick better think twice and don´t go on feeding the VST cheaters.

Yep. I recall a thread on some forum, maybe a mailing list. Someone was asking about how to the general vibe of a Boards of Canada sound and the response was "a room full of analog hardware." Yes it was a bit flippant, but, we shouldn't dismiss it so quickly in the same way that we don't dismiss the sound of analog consoles and outboard vintage gear. In isolation emulations are very convincing. However, if that were all there is to it, then why do we need plugins to simulate subtle channel differences?

From a sales perrspctive: because we need arguments and USPs to convince customers. A sales manager knows that he just needs to convince the people. The potential customers. Not the scientific audience or the pope. Just people who are willing to pay for the advantages of the product over others.

I don´t know if you see the same you tube advertisements where you live but here in germany we frequently hear "read this book and get rich", "I can show you how to loose kilos without changing your eating habits" or "this tablet will clean your toilet and you don´t have to do anything".

The VST companys usually don´t promise too much, they just say "modelled after" and their peergroups exaggerate it for them ("Sounds just like the hardware!"). But the behaviour seems just as suspicious to me as the behaviour of the people behind the mentioned advertisements.

But I definitely shouldn´t generalize too much. There are good plugins and even if it´s just because they are so easy to use in comparison, take no room and don´t eat cables

[ghettosynth]

The differences sum up btw. You often don´t hear that much of a difference if you play just a single sound. So if you ask yourself once again when something is wrong with your mix which plugin does the trick better think twice and don´t go on feeding the VST cheaters.

Yep. I recall a thread on some forum, maybe a mailing list. Someone was asking about how to the general vibe of a Boards of Canada sound and the response was "a room full of analog hardware." Yes it was a bit flippant, but, we shouldn't dismiss it so quickly in the same way that we don't dismiss the sound of analog consoles and outboard vintage gear. In isolation emulations are very convincing. However, if that were all there is to it, then why do we need plugins to simulate subtle channel differences?

From a sales perrspctive: because we need arguments and USPs to convince customers. A sales manager knows that he just needs to convince the people. The potential customers. Not the scientific audience or the pope. Just people who are willing to pay for the advantages of the product over others.

I don´t know if you see the same you tube advertisements where you live but here in germany we frequently hear "read this book and get rich", "I can show you how to loose kilos without changing your eating habits" or "this tablet will clean your toilet and you don´t have to do anything".

The VST companys usually don´t promise too much, they just say "modelled after" and their peergroups exaggerate it for them ("Sounds just like the hardware!"). But the behaviour seems just as suspicious to me as the behaviour of the people behind the mentioned advertisements.

But I definitely shouldn´t generalize too much. There are good plugins and even if it´s just because they are so easy to use in comparison, take no room and don´t eat cables

Of course, and get rich quick sells like hotcakes here as well, but, you have to see the other side. Dave Grohl and others of the same mind are convinced that the hardware, taken as a whole, is better. He made a movie about it even.

[zerocrossing]

Chimpanzees are also 98% there; yet they look & sound nothing like homo sapiens.

Did a chimp write this, because that’s the level of intelligence I’m seeing.

[zerocrossing]

The differences sum up btw. You often don´t hear that much of a difference if you play just a single sound. So if you ask yourself once again when something is wrong with your mix which plugin does the trick better think twice and don´t go on feeding the VST cheaters.

Yep. I recall a thread on some forum, maybe a mailing list. Someone was asking about how to the general vibe of a Boards of Canada sound and the response was "a room full of analog hardware." Yes it was a bit flippant, but, we shouldn't dismiss it so quickly in the same way that we don't dismiss the sound of analog consoles and outboard vintage gear. In isolation emulations are very convincing. However, if that were all there is to it, then why do we need plugins to simulate subtle channel differences?

We don’t… unless we want to sound like a band that’s been living off residuals for a very long time. I can also point out Grammy Award winning pop albums that have out sold every single BoC album by a factor of 10 that were made on a UAD Apollo and a laptop.

Of course, [checks notes] none of us are doing anything as well as either scenario… so… moving on.

Again, with software, you could quite easily use a different channel strip plugin on every single channel. I’m unclear as to how that would make a track better, but go you. I also don’t think “subtle channel differences” contribute much to your overall mix. I’m hoping you adjust everything by ear, so any differences would be moot.

I will say that a good preamp emulation after a plugin instrument, even if it’s not an analog emulation, can help make it pop. One thing I notice when I’m putting a plugin under a microscope and comparing it to one of my hardware synths is that it’s unfair not to compare apples to emulated apples, and that means skin and stems.

[DrGonzo]

He was quick to move on from the Envelope test and based on his look and talking around it made it look like he was trying to cover up the flaw in the Softube envelopes.

You can clearly hear how lifeless and weird the envelope on the software sound in comparison to the Juno if you go to 26:26 in that video.

I don't think he wanted to cover up anything and my suspicion is that he is focusing on the whole.

When it comes to noticing details like 1:1 envelope behaviour, I first want to congratulate you for developing a personal taste and being able to notice things like that. As much as this is a super power it's also a bit of a curse, getting yourself hung up on details like that. I've had that seeking-for-perfection trait in me as well and looking back it never ever did me anything good. That ruined my love for the Ensoniq EPS16+ with its crackling envelopes. Cast a wide net. Focus on what works rather than what's perfect. Today I admire the EPS crackling envelopes. They add to its character.

Speaking of Junos - partly because they are fairly simple synths with a great sound - I love the Arturia JUN6. But I also love the Cherry one. Vastly different sound.I can clearly audiolize/visualize their differences in sound and decide which one would work best.

20 years ago I was hanging out in the Muffwiggler forum trying to build myself a Eurorack Juno clone (mono of course) and the most common advice I got was: BUY AN EFFING JUNO instead of wasting your time.

Seriously. Isn't EVERY clone or replication inferior? It wouldn't surprise me if Black Corporation Deckard´s Dream isn't even capturing every nuance of the CS80.