Yes. Why beat around the bush and make it about testing methods or specific frequency bands? "Emulations are not as nice as the real thing. Ever. " See how easy that is? The only reason to debate and compare is a) money b) faster workflow.

Developers who code and sell emulation versions love to tell you how a file and image on your screen is the same thing as a $10,000 compressor and that you are better off buying their 50 dollar file.

People using said plugin will of course attempt to convince others it is the same thing, too. Like they will tell you all about null tests and how "my 20$ daw sounds exactly like protools". (when playing a single wav file).

If you want to have that discussion, it is tired; but have at it. Skip the gyrations about 200 hz trying to prove a point instead of asking the question you won't ask outright.

 

I agree 100%. Well, I would probably say that taste and skill counts. The ears are not so important (as long as you have some... that aren't deaf).
But the question is not whether you can create a great mix with plugins or whether you can create a bad mix with a $200k hardware rack, but why plugins often sound so different from their hardware counterparts.

 

There's an Airwindows plugin called Interstage that is meant to fix this, or how the developer says "bring disperse digital lows and highs" to an area where they can be heard". I use on a daily basis on my bass channel and it's hard to explain its effect, but i like how it brings up the low mids area.

 

Another good example of digitized Analogness* could be UAD Luna's NonLinear Summing plug,
(and the type of workarounds Devs are doing to tackle the issue..)

I personally prefeer the API, Subtle but I'd say pretty good/positive..

Waves did a good job with their NLS plug, I like the EMI there as it's also a subtle enhancement,
whereas the SSL and Neve emulations are more apparent..

But yeah on first glance UAD's take seems to be more Refined.. and Deep?
more True?

Does this check the 200Hz box?

 

Last edited: Jul 10, 2023

Surprisingly, it seems to work with his, er, 'mixes' and it works with white noise but the delta for normal signals is so low that I didn't even try a blind test (IOW, inaudible, ATME).

 

After glancing over the source code and testing Interstage on a mix, I think a resonant high pass is going to give you the exact same effect, minus the crunchiness from Interstage's mixing in of previous samples. I'm not even sure the mixing in of the lowpass (I guess it's supposed to be a slew clipper?) is working correctly at all, but I might be wrong.

 

That's what Nebula is for, and NAM is also showing some promise with FX and studio gear modelling.

 

GuitarML / Proteus too

 

I'm having fun with NAM and Proteus, and i am an Aqcua user since a long time. Sounds really open and analog, especially for a primitive project. If you combine the Hardware section with some IR's (option available for NAM) it will get the job done pretty decent

 

That guy ows me... 500 USD? But then again, those waves plugins are 10 years old.
Focused on the saturation, the "electricity" of the sound, not the snare or 200hz.

1. If you think you can't write another note until you can afford hardware or the software catches up, you're very wrong.
That's just self-sabotage. Don't mix a song while you're creating it. "Why FINISHING YOUR SONGS Never Gets Easier." by the House of Kush opened my eyes to this.

2. Software is closing the gap to analog and is way closer than 10 years ago, but...

3. Why do we need to sound like 1970? Don't get me wrong, it's my favorite music, but we're not mixing the Bee Gees.
It's like Need for Speed, you can drive that supercar, be it in digital form.

My 2c...

 

Fairly interesting point but I personally think the problem in "narrowable" just on low mid. This shitty plugin low mid lol.

It's not an issue in 2023 lol. Most of the people does not care at all. Most listeners standards lowered because of the drive of the music industry, expanding on genres, distribution, age of reach, ect.
And all the tone / gain / wave shaping functions of plugins are fairly fixed imho, just don't look at specifically "HaRdWaRe OfFiCiAl EmUlAtIoNs" and try to understand the sound of the units you like to reproduce it with solid options. TDR offers one of the best digital sound tools available, have a look. Intelligent coding, approach on gain control signal, ect...

You need to precise your problem a little more,

- does hardware compress better than digital ? what is "better" ? what are the qualities you are after ? this is not a competition, but most of the mastering on earth is done on proper coded compression algos for a reason, which are their precision, control (so more room for improvement and creativity), fidelity to the source;

- does hardware saturates better than digital ? what is "better" ? aliasing ? grab apshaper, you can saturate the grandmother of the signal to the noise, at max quality, it won't alias an inch, but beware if you have hypersonic ears, you could hear -180dB intermodulation that could not exist in analog because of noise floor. You can even chose the shape of the transfer curve, add dynamic modifiers and so on...

- does hardware gives "tone" better than digital ? what is "tone" ? the eq curve ? any proper digital EQ 1:1 a reasonably good measured signal. So what is left to the tone ? tolerance breaking the eq response ? happens with shitty or broken equipment dude. But yeah generate some thermal noise and use an IR to have some fun !


The last thing we could not rip off an hardware unit would be its sole identity and that's perfect. The definitive analog unit value is its unicity.
But can we replicate it, has it was another one standing ? yes, of course. The very same ? Never.

I'm surprised people here are still hardware supremacists, when I touched my first gear (outboard gear, consoles, reverbs) years ago I really realized I idealized a lot of stuff about analog electronics. Undeniably more "forgiving" than ever, as I like to say when I compare analog gear against digital audio processors. Superb sounding, particularly if it's dead expensive. But no better, just different. Except when emulations are shitty, but that makes sense as there is a fidelity notion toward a particular unit behavior underlying when talking about emus.

I have always followed a critical, strict and exigent hearing, towards a wide variety of genres, and honestly, apart from some obscure units, or difficult beasts like the 670, I don't think there is an existing sonic superiority in hardware anymore when processing audio. You cannot of course emulate speakers and acoustic, but I think you have my point.


When I hear AI inference and interpolation superpowers (like the Travis Scott models actually capture its vocal chain near perfect), I just think it's about time some cleaver nerds actually shatter the mixing industry which is actually crossing the next step of rotting with learned models of hardware. It's a leap back to stupid black boxes but seems hyper promising. Maybe some other clever guy could make a model learn electronics and tolerances to suggest something even more interesting ? Dreaming is free.


Ah, and yes, analog gives more depth. What is depth ?

Spoiler

Transient definition when processing, (el famoso "true to the source", the analog "clarity"), stereo decorrelation, decorrelated noise, "punch" (low and low-mid signal in the transient phase does not get squared, electronic dynamic response (subtle modifiers on the path of audio like waveshaping or tone shaping, ect...)

"Why is 200hz below so hard for devs"
"200hz is hard for devs but not a dead-end. There a working solutions taking care of your sound below 200Hz. The problem ain't the devs. Plugins companies are the problem, they will just squeeze dry the branding marketing till a next fracture in the industry like a revolutionary product will threaten them. They use the same tech since 20 years lol.

 

Analog saturation compresses and EQ's naturally, and that's why many professional mixing and mastering engineers use hardware.

https://kmraudio.com/blogs/interview/dave-eringa-interview-part1

https://kmraudio.com/blogs/interview/dave-eringa-interview-part-2

 

Last edited: Jul 13, 2023

most people want to hear how the plugin sounds, matched (by ear) and that's exactly what he did and most other people do, and devs actually.

My thesis is not that plugins suck. That's your assumption so battle with that in privacy and respect yourself

 

The overwhelming majority of digital saturation methods naturally compress too. Digital saturations are just less 'imperfect' than their analog cousins. No component tolerances, no sagging, much less noise, no stereo decorrelation (or crosstalk for that matter), not sensitive to changes in temperature, no 'natural' high/low-pass.

Give it a couple years and people will grow nostalgic for that early 2000s "digital warmth" that can't be recreated with modern plugins because magic. The sound of the 2030s is just too damn clean, I hate it already!

 

hi guise, i want to make an old school edm trax like the ones in the '10s decade!! how do i emulate plugins like sylenth1 and nexus? thanx

 

This is always an interesting topic. Thanks to everyone for their insights

 

Can you explain this? Isn't thermal noise just white noise?

 

Already happened, IIRC with the S950 converter sound. Most recent example HoRNet ADDA The sound of early digital devices. You can now crappify your sound for just 12€, ain't that great?

 

Mostly, yes. White Gaussian noise for thermal noise in general (even if Gaussian - type of density distribution ≠ White - constant spectral density).

For an handmade analog seasoning of an eq curve,

You can load an impulse of your choice into an IR (you can play with its length if you're adventurous but keep full control of it because it will mess with your transients wildly) and blend this signal into tiny amounts with the equalized one. You will break the filter response and simulate tolerance inaccuracies that makes sometimes filters circuit do not respond perfectly. This is a fun (at least to me) and a creative way to process sound, but again, disclaimer, that's just strange wizardry, some of those won't finish on any master.

Note, those are caricatural in terms of sound. This may give a boner for those who mix with eyes but hurt people mixing with ears.

Just some stupid examples

Spoiler

Magic decorrelated lowend


Faulty grandma shelf


Realistic 20k box tone

 

Oh, that's clever. I'm working on my own "analog" EQ, but my implementation for tolerances is much more naive than that you came up with - basically shifting freqs/changing Q/gain + mixing in velvet noise for better stereo decorrelation.

That's a great idea though. Might steal.