Just export to fuckin 96 Kbps MP3. You'll love it. You're welcome

 

Don't laugh or hate me, but,
I think there is one person that could dig super deep and explain why hardware is super hard to replicate, and why it requires tons of processing power...:

Uhs heckmann, or rather uhe.

Why ? Because before he was a software dev, he was... an industrial designer...
... and he knows exactly what it takes, to make a good, great, or perfect emulation of a circuit...

Take Diva, which contains several modules that emulate oscillators, filters, etc, from many old hardware synths.
When designing and coding Diva's algorithms, there must be a never ending fight between.... audio quality VS cpu usage. (Not even talking about supersampling...).
The first versus of diva, all customers were complaining about the cpu usage.
Then, some upgrades were released, and the vsti required les and less cpu.
Did uhe improve some algorithms ? Maybe. Did they reduce the quality, and directly, the cpu usage ? Probably for sure.
Maybe the audio quality was already so high, that even reduced by 10, 20 or 30%, it would still sound good enough..?

Just like a video game, that would use top quality assets, perfect antialiasing, perfect visuals, perfect lighting, with a ton of raytracing, would probably run at 1 fps, even on a 10k computer,
If Uhe was using the best algorithms as humanly possible, and the most faithful emulation as possible, and he didn't care about cpu usage,
Maybe 100% of 16 cores at 5GHz wouldn't be enough, to play 1 single note, of a Diva patch...?

 

Joint Stereo?

 

So, chain = EQ -> IRloader?

What's this?

 

There are different colours of noises. White noise is a random number added to the signal that is equally loud across the frequency spectrum and doesn't change its behaviour across time. Other colours of noise are filtered or generated in a way that they're not the same across the spectrum, but they're still behaving the same across time. Velvet noise is different in that it is a sequence of impulses forming an impulse response. VN adapts to the incoming signal in time and amplitude (and if one wanted to, in frequency). It's not 'static background noise', it's '(re)active background noise'.

Check out the examples for this particular type of velvet noise: https://www.audiolabs-erlangen.de/resources/2018-DAFx-VND

OVN15 - to my ears - subtly improves the original track every time, never sounds worse and beats white noise (WN) every time.

 

Last edited: Jul 13, 2023

Hence comes the technology, the approach. How to find different ways to solve nodal analysis or any other methods to find current, or phase values coherent with the reality.

Some examples may be the neural network of the Chowtape (STN) for outputting hysteresis values (even posted here afaik), or the "transform" (uneducated guess : probably modifying filter coefficients ?) on eq's like Kirchhoff or Crave to match analog phase response without oversampling, or even the optimized internal bandwidth ~400kHz sample rate of TDR products at max settings.

Both allow to output, to "obtain" real physics sound with less ressources needed with previous techniques.

And don't fear the derisive kind of topic lol. It's a fascinating subject and it's normal to attract all kind of people to exchange on this topic, the stake is huge. Even people which crystallized their ego on some machines. People are happy to know new things even if they prefer mocking you. It's an human thing.

 

I preferred OVN15 in the first test, OVN30 in the second one. There's someone here who got some of them .wavs?

Edit: I searched on the internet for gaussian white noise and for velvet noise, but I couldn't really find any audio files

 

Last edited: Jul 14, 2023

Same difference "less beefy in the low mid area"

There is something special about low mids and below that no one has discovered yet...

 

The low mids problem...

Maybe those developers should ask these genius hardware manufacturers. They seem to have worked that out already.

 

There is no

There's nothing special or unknown about hardware. It's built by engineers, not magic, and what every component does is known and documented. This thread has no examples of the low end issue you're talking about nor any blind tests or frequency analysis demonstrating its existence. This happens on every plug-in forum. Objective A/B tests are always missing from the discussion.

 

That's just your opinion, the FACT IS software has a difference that can be heard in that area.

Also what is also missing form your response is your test, your hardware and YOUR results. I've posted people who actually have access to the hardware, people who evidently have the cash and experience. So why would I bow to your "objective bla bla bla" when someone else who has the hardware, experience has tested it for me and confirmed what I've said?

I don't get why I should ignore what I'm hearing and seeing for you? please explain that logic

 

Last edited: Jul 14, 2023

Hmmm... let's say joint... something. I'll leave it at that

 

And do not forget that shitty sound from the Antares mic modeling. I am telling you, that will become a classic. Nothing else in the market compares to it. It sounds so fucking bad that it is as unique as it can be.

 

For years, developers have been selling plugins which they claim are an "exact reproduction" of some rare or super expensive hardware that makes it almost impossible for the regular Joe to compare it to. Moreover, there is always a lot of wiggle room if developer X claims to have coded the most accurate emulation of a LA2 compressor. If "experts" contend that the plugin does not sound like their LA2, the developer simply will say that no 2 pieces of analog gear sound exactly the same. The only exception being emulations of digital hardware, like Softube's Weiss mastering EQ. Finally, developers know that the majority of their customer base NEVER have heard , much less used, their plugin's hardware counterpart. Therefore, right off the starting line they are in good shape. Indeed, I will point out to my comment above. Do you really think that a $150 plugin will turn your $100 Shure SM57 into a $15k mic? Really? Not so! And please, do not come here telling me that such and such Grammy winner swears by it. Those are audio mercenaries. Give them a couple of bucks, publish their photograph, and they will swear that you are Jesus or Ghandi.

 

Or like Acustica Audio is doing it by using 'modded' hardware.

 

Yes yes yes... modded, indeed

 

proteus solved all the saturation needs for me, cant think any better solution, so that is that