It seems like low mids is almost impossible to get right when emulating hardware. Why is this?
And why do they keep releasing plugins that have this problem

 

One or two examples would be useful.

 

I suspect you have an entirely different problem in your system. Need details.

 

Cause computers are inherently Cold Digital Scientific machines, made by Nerds, devoid of any Soul,
so anything Warm and cozy like the "Woman Tone" is philosophically and cosmically pretty far apart..

Afterall it's all 1's and 0's, pure Quantization matrix..
so in order to be Warm and Organic they they need to be Rounded off, given some warmth, fuzz and raw unprecision..
And that's what plugin devs are tryin to do/solve..


(Ass talkin here..
But yeah, Nembrini Clon/Minotaur comes to mind,
while imperfect? it's got some Digitized warmth right there..)

 

Last edited: Jul 9, 2023

It's more or less the exact opposite

 

just slap tal-dac on it and saturate carefully, problem solved.
(osirus is no joke, have you tried it?)

 

It is indeed
but isn't..

It's the 1 and 0 paradox and dichotomy right there..
So gimme my 0.5s nigg, gimme my .33^s, Pi's and Phi's, gimme unnormalized un-Plancked & unlimited Mojo,
that's what a properly Antialiased Woman Curve shall be like,
Sweet, Warm and Mushy beyond the squares

 

Every example has the same difference. The hardware is thicker. Even if it's some other reason that has nothing to do with the hardware, why can't they compensate?

 

What are you talking about?

 

theres a 1959 original Fairchild on ebay, you can solve the problem (at least for 1 track, 20k dollars per track)

 

The question is why is 200hz below so hard to code/emulate

 

Because you can not charge $2000 for a Software based device.
In my opinion from a purely technical sight you can not even charge 1c but that's a different discussion.

Pulsar modular got the best and thickest and blah sound by far at the moment. They charge the most.

Maybe you should try the non-emulation software then if the emulations do not deliver. They could have a thicker sound.

 

Also BIAS Amp comes to mind.. nice try

A bit impractical in the end tho, you'd Never finish tweaking it..
so things like Neural Wong/Nolly, or Nembrini Extacy go on Top on my list.


As far as Compresors go, I donno..
Softube CL-1B or Summit TLA100-A, or maybe ELi Arousor could be examples of good (or even great?) Tone,
and I guess praxis when doing a VST emulation of a piece of hardware..?

Also what about Dynamics,
Elysia Alpha Master doesn't Color the signal too much, but the Dynamics, boy.. what that does is pure Magic

 

So the guy really said let's compare hardware and waves plugins from 10 years ago

 

ok seriously now

it's unuseful to open a thread talking about the difference between software and hardware in midrange

you noticed a lack of energy in that point? Pretty fair.

So you think that the silkness of an API EQ for the 10kHz is perfectly emulated by soft plugins? do you think that the only one difference is in midrange, low end and top end are finally matched hardware/software?

Well I dont think so, and many people agree

For me, the BIGGEST DISTANCE hardware/software is not in dynamics or EQ, but is in the ambience/reverb simulation. Thats a point in which software is still FAR AWAY than simulating lexicon machines, plate EMT, BX20 spring and so on and so forth

The point is: it's unuseful to ask software developers "to match more the hardware", we just need to understand that software and hardware are 2 different worlds and we have to make both work in its own way

Let's then talk "how to get a good midrange with soft tools", that would be interesting

 

I'm inclined to agree with you, I just didn't want to be the asshole who called someone out on that likely being the issue when I don't have much experience with hardware (at least not high-end mixing units.)
But, now that we got that out the way, my initial basis for it being an issue on your end @gotnofriends is that I find it incredibly difficult to believe that a grip of highly regarded and sought after mixing engineers—many of them old enough that they have a genuine love affair with their consoles and hardware units—would willingly part ways with their tubes and diodes and knobs and switches for a bunch of digital plugins, if they're even half as cold and dead as you're making them out to be. In the same vein, if digital is truly lacking to such a detrimental degree, labels could be require that a mix engineer use hardware to ensure that a track comes out proper quality, but I've never heard of, watched, or read of any insistence on using analogue gear.

I like the fact you seem to be making the most of the opportunity to discuss around this topic (as my opinion after my above comment should make it clear where I stand... well, clear enough, anyway.) I agree, there is often times a missing bit of, as you put it, ambience. It's not yet quite an answer to that, but I like what I've seen with the neural nets emulating the more active/fluid responses of hardware. Past to Future Reverbs has been mapping hardware units for Proteus and Neural Amp Simulator, both initially made for guitar amps neural IRs, but as it turns out, they work equally well with any... I'll call them static feeds, like tape machines, compressors, limiters, etc. (as opposed to delays, reverbs, etc.) I think we're going to see a jump in that missing mojo once neural nets (or just as likely, new, similar AI approaches,) become fully realized and, of course, devs can figure a way to lighten their CPU load, as they're fairly taxing; as they are, it's hardly a real option to use them on a track-to-track basis, as you would with say, pre-amps or tape machines.

 

You're not wrong. Ever since I've been doing this shit ('97), hardware synths have just produced buttery low-end. From NL1 to Waldorf Q and Moog Slim Phatty n' Minitaur or Vermona Mono Lancet, the list is longer. Real metal xfmrs (iron, steel, nickel) also seems to iron out an unruly low to low-mid. An ADDA roundtrip may have a role as well. Experiment and find what works 4 u.

 

I would like to point out that the comparison method in this video is flawed (I'm not speaking of engineers of course).

Why?

- a required condition in any statistical comparison is that test conditions must be the same, or at least recreated as much as possible as the reference (in this case same monitors, eq'ed room, etc.).

- the tester must provide the same audio sample(s) to each of the tested engineers.

- to gather statistical data, the number of provided samples must be enough (it depends from the type of test, but surely 1 sample is not enough (I would say 10 would give a sufficient estimate compared to the 50% coin toss of the clip).

Also, it's not clear if the tester adjusted VSTs replicating knobs or by ear, and that would add another bit of uncertainty in the test system.

Summing up, it seems more of an occasional relaxed game among friends than an evaluation test.

Yep, beside that, also ears get old, this is something alwasy neglected, but unfortunately true for anyone.

 

@eXACT_Beats_ what does your thesis on me and the industry have to do with my question?

 

Last edited: Jul 10, 2023

Exactly, it's clear is day.

 

Last edited: Jul 10, 2023