Any science to explain the “weight” or “3D depth” of hardware audio vs software that some people claim?

DSP, Plug-in and Host development discussion.
KVRian
1453 posts since 13 Oct, 2003 from Oulu, Finland

Post Mon Apr 12, 2021 11:46 am

perpetual3 wrote:
Mon Apr 12, 2021 5:22 am
whyterabbyt wrote:
Mon Apr 12, 2021 4:17 am
perpetual3 wrote:
Sun Apr 11, 2021 10:27 am
whyterabbyt wrote:
Sun Apr 11, 2021 8:33 am
perpetual3 wrote:
Sun Apr 11, 2021 7:13 am
There are lots of papers.
Should be easy for you then. Just a couple will do.
Same goes for you.
Erm, no. Im asking for your evidence for your claim.

Since you're avoiding providing any, I think its safe to say you dont actually have any.
If you’re sufficiently interested and self motivated, you would do the research yourself. I’m not here to provide academic rigor to my claims or conversations. I do that for money. But if you want to frame it as “avoidance due to lack” that’s your prerogative and ultimately your loss not mine.
That's not how things work:
In scientific circles (and any even remotely intelligently behaving circle really) the responsibility of providing the proof is on the person how makes the claim.

Otherwise I could state that "there's a pink unicorn flying around Saturn".
You say "Prove it!"
Then I say "If you don't believe it, then YOU prove that it doesn't exist!"

Since no-one can prove it doesn't exist, does not mean it exists, obviously. :)

KVRAF
2552 posts since 23 Jun, 2006

Post Mon Apr 12, 2021 2:37 pm

It's a very unscientific definition that corresponds to some of the tricks someone mentioned here. In reality, digital has accustomed us to a stability and precision that in some contexts sounds unmusical.
Some hardware objects, on the other hand, tend to be unsteady, and to sound a bit like there is an instrumentalist behind them, a human. On many levels, this is perceived as less digital.
There are lots of techniques for making a source not perfectly constant.
How to put this into words? Here you are describing a series of imperfections as if there is a new dimension. The new dimension actually exists, and it is time. It is an intuitive way to define a processing enrichment.
Where is the scientific value in this? In the measurement of all those physical characteristics that lead to these variations. For example, the nonlinearity of a transformer, the variation of nonlinearities over time, and so on. All of this is scientifically measurable, and also very variable from object to object.

Some years ago I heard of a musician that in dance tracks, where the bottom drum is usually played by a computer, he tended to add a sub line played without quantization by him. It is a process of humanization, it makes everything more natural, more similar to what we know in the analog world, that is not synchronized with the atomic clock. I think the electronic music of the early 90s is still considered special, but in fact all that often happened is that the midi of that time often trudged along randomly.

KVRian
1184 posts since 11 Jun, 2019

Post Mon Apr 12, 2021 3:25 pm

Stability and prescision ...

I must admit that I was never convinced by "drift technologys" and so on - for one simple reason: a usual synth sound is much too "defined" and "present" in a way that is very unnatural. Oscilltor drift and so on won´t change much about it. But what really does something to the 3D perception and the sounds appearance is what has just been mentioned: layering. Playing something "below". Layering separated elements systematically together will give you more control over the sounds appeareance and results in a kind of "natural deflection" compared with the technical presence of simple synth waveforms. You hear two unaligned waveforms on different levels, with different pan, filter, frequency balance, envelopes and what ever. For me the result - if done good - always sounds much better than anything you could have done "technically" - I assume that less stability and prescision are the reason for it.

ROMplers can prove that btw ;-)

User avatar
GRRRRRRR!
11392 posts since 14 Jun, 2001 from Somewhere else, on principle

Post Mon Apr 12, 2021 6:29 pm

perpetual3 wrote:
Sun Apr 11, 2021 4:52 am
Not really. Answer: non linearities, slight differences between L/R channels.
But those are things easily and often emulated in software so it explains nothing.
GRUMP wrote:
Sun Apr 11, 2021 10:21 am
Just take the time to compare with Hardware and/or HW recordings and you´ll quickly realize that the placebo discussion might result from a lack of sensitivity.
Quite the opposite. The closer you listen, the more you realise that one is the equal of the other. e.g. Many years ago I put up a blind test between my K-Station and V-Station and the results were perfectly in line with what happens when you make a random choice between two options. i.e. Around half the people picked the K-Station, half erroneously picked the V. Only one person made an informed decision - he picked the hardware based on the background noise he could hear on one sample but not the other. So I learned that next time I do something like that, I should keep the hardware plugged in, even when I am recording the software.
I have bever heard "that VST high freuqency foggy noise" from a piece of hardware ort a HW sample (And I don´t (just) mean aliasing).
And, AFAIK, I have never heard it in software, either. Although the worst aliasing I've ever heard has come from hardware samplers (Ensoniq Mirage, for example), not from any computer.
Just the opposite. Even the Virus A (1997) "shines" with a well defined, vivid high frequency spectrum.
But that synth is just software running in a dedicated box. It is 100% as digital as DUNE, which also has " a well defined, vivid high frequency spectrum". So does ANA2 and half-a-dozen other VSTi I own. Seriously, download the ANA 2 demo and play the default preset as far up the scale as you can. It remains pristine and beautiful after a lot of analogue hardware synths would have devolved into horrid squealing.
NOVAkILL : Dell G7 Core i7, 32GB RAM, Win10, Zoom U24 | Studio One | Thorn, bx_oberhausen, ARP Odyssey, JP6K, Hexeract, Vacuum Pro, TRK-01, Knifonium, Equator, VG Carbon | Uno Pro Desktop, Uno, Analog Keys, Ultranova, Rocket.

User avatar
KVRAF
1666 posts since 4 May, 2012

Post Mon Apr 12, 2021 7:10 pm

Electronic circuits often require feedback to perform tasks that are otherwise trivial ITB: Op amp implementation requiring feedback Vs straight multiplication, for example. There might be something to measure in phase introduced by inductance - and electromagnetism in general can create complex relationships between electronic components and generate noise within the audio frequency range.

User avatar
GRRRRRRR!
11392 posts since 14 Jun, 2001 from Somewhere else, on principle

Post Mon Apr 12, 2021 9:30 pm

All of which can be emulated.
NOVAkILL : Dell G7 Core i7, 32GB RAM, Win10, Zoom U24 | Studio One | Thorn, bx_oberhausen, ARP Odyssey, JP6K, Hexeract, Vacuum Pro, TRK-01, Knifonium, Equator, VG Carbon | Uno Pro Desktop, Uno, Analog Keys, Ultranova, Rocket.

AnX
KVRAF
9209 posts since 17 Nov, 2015

Post Mon Apr 12, 2021 9:42 pm

I think the answer to the question is no, unless I missed the scientific proof someone linked to?

User avatar
KVRAF
2538 posts since 4 Sep, 2006 from 127.0.0.1

Post Mon Apr 12, 2021 9:53 pm

in my mind, it's as simple as this:

if the "magic" of <analog/hardware> can be recorded digitally, and it is still there when that recording is played back, THEN that digital signal could be emulated and synthesized digitally, with enough effort.
It doesn't matter how it sounds..
..as long as it has BASS and it's LOUD!

irc.freenode.net >>> #kvr

User avatar
KVRian
1283 posts since 10 Oct, 2018

Post Mon Apr 12, 2021 11:16 pm

Then again, one still has to mix the sound. Which compromises any single sound when mixed together. No matter how you listen to it.
Then you have to add reverb. Without the appropriate reverb you never are going to get the same sound you hear from the recordings approximately.

I won't start about compression/saturation or other fx.

KVRAF
4369 posts since 22 Nov, 2012

Post Tue Apr 13, 2021 4:45 am

yeah, someone said long ago that a great mix includes both, and i always agreed with that. the virus was a digital synth, waldorfs are digital synths, but they are flowing through analog circuits before the output. it's more about the behavior of the instrument. the mixing needs can be overcome with a simple push of a button these days. it's all AI.

KVRAF
28088 posts since 11 Apr, 2010 from Germany

Post Tue Apr 13, 2021 4:50 am

I think what this thread needs is less talk, and more sound examples. Post some sound demos of "weighty", "3D" sounds. They do exist IMO. Instead of asking for a science to explain them, or anything else esoteric, people should hear what you mean. That's what sound is about.

Things like "flat", "plasticky", "thin", "fat", "organic" definitely exist. It's just hard to put them in words. Need moar sound.
Plugins and a DAW. On an operating system. Which runs on a computer.

KVRAF
1696 posts since 28 Sep, 2012

Post Tue Apr 13, 2021 5:06 am

chk071 wrote:
Tue Apr 13, 2021 4:50 am
I think what this thread needs is less talk, and more sound examples. Post some sound demos of "weighty", "3D" sounds. They do exist IMO. Instead of asking for a science to explain them, or anything else esoteric, people should hear what you mean. That's what sound is about.

Things like "flat", "plasticky", "thin", "fat", "organic" definitely exist. It's just hard to put them in words. Need moar sound.
Agreed. And I actually outlined a process by which someone can emulate the process ITB and conduct a scientific experiment themselves.

KVRAF
4369 posts since 22 Nov, 2012

Post Tue Apr 13, 2021 5:15 am

yes, record your analog to tape, then convert it to digital, before compressing it to mp3 and post it for us, so we can stream it from the web through out digital converters no less!

over apple ear pods!

KVRAF
28088 posts since 11 Apr, 2010 from Germany

Post Tue Apr 13, 2021 5:33 am

Like all the music these days, regardless whether they were made with digital or analog gear. :P

The things I listed have nothing to do with analog or digital anyway. There are enough digital synths, which model those things pretty well. I'm sure there's still a difference to a real analog in many situations, but, I wouldn't count top end which goes to infinite to those at all, especially as noone is capable of hearing it anyway.

Just like vinyl. You much rather hear all the crackling and humming than the top end (which is limited by the physical properties of the speakers anyway).
Plugins and a DAW. On an operating system. Which runs on a computer.

KVRAF
4369 posts since 22 Nov, 2012

Post Tue Apr 13, 2021 8:36 am

there is a huge debate in the music community over wether we only listen to overtones. just saying.

Return to “DSP and Plug-in Development”