im "trying" to work on a way to take fractals and using calculus , make a digital algorithm
for digital signal processing to re-create the non-linearity similar to analog gear based on the input voltage signal and biasing.


the difficult thing is needing incredibly high sample rate ( problems for real-time.)
the cool thing is the ability to create all new types of processing and fx (as well as traditional processing and fx)
we can see in this simple fractal the way analog harmonics effect analog signal and the soft clipping ,compression etc. the large portion of the image represents the regular transfer curve of the signal
think about a granular type effect as well to create a infinite portion of signal as a snap shot at any point. the fx are limitless

the real amazing quality from great vintage gear is down to each electron (well not exactly THAT far down ) with digital we have only thus far created VERY simplistic almost line drawing versions of what analog does to signal.
think about vector art for a moment

the lines are infinite detail instead of finite like jpeg or reg images
we meed am equivalent with digital audio where whn we get donw to the tiny detail of non-linearity we dont run out of bits (gets fuzzier and fizzier)
equivalent in terms of audio processing

 

Last edited: Dec 2, 2017

Again, neither. If you read the entire post I made, you would see there is zero need for these weird strawman "vs" arguments you are making.

I generally dislike everything "vintage". I couldn't give a fuck for example about how accurately a plugin emulates a specific piece of gear. But often less is better. We humans are very limited in our abilities, even the best of us. It's easy to lose focus on what sounds good when you have the possibility to record super dynamic, clean and balanced digital 24/192 sound. Many times in certain context "lofi" sounds better. Old gear / techniques, even virtual emulations can point you in the right direction.

If you want to discuss about this further, please make a coherent post that actually addresses what I said.

 

double post

 

I am out of here this is getting silly...
Spencer

 

I didn't respond to your entire post; asked what you meant by this:

What did you mean?

 

You're ignoring the limitations of human hearing. Think about human hearing as a 300x300 256-color gif. Vector graphics will appear [at best] identical to a simple bitmap. Just like with editing vids/shopping (sampling/upscaling images), higher resolution becomes important when mixing, but until you come up with a higher-resolution ear, not in the finished product.

 

Isn't it clear? There's so much more to sound than "vintage was shit" and "modern is best". It's not about the technical qualities of the sound, it is about the emotion and feel the sound gives you. If you disagree or think I understood something wrong, please go ahead and explain. But if you read my entire post it should be 100% clear my post had absolutely nothing to do with that "photoshopping and art filter on a gif file" vs "the real thing" -nonsense. I'm sorry if I sound aggressive but I'm just trying to be very clear here.

 

Man, each and every time I re-watch any movie in HD/60fps I feel completely discombobulated. Although I know 60 frames is much more fluid and "real life-like" than 24fps, I have the awful sensation of artificiality. I don't know how to explain it. Of course, it must be only a matter of custom (or lack of).

 

I maybe wrong, and it really is dependent on the type of analogue being discussed... EG natural tape compression or tube amplification.

But isn't the layman terms difference basically the number and type of overtones that colour the sound ? Transistor Amp vs Tube Amp being the most obvious example ?

I've only really replied because that fractal image reminded me of this awesome bit of 1K JS code !

http://js1k.com/2014-dragons/demo/1903

 

Not in the context of this thread, no. OP's point was fairly clear:
"So when I hear about the great new plugin emulating everything about the Vintage Sound I know you weren't there."

I assumed your post was related to this thread's topic. Wrong again. Not to say that your post wasn't full of fresh, thought-provoking ponderings on the fascinating topic of lofi production. A lot to take in, I'll give it some thought

 

The best things about The Beatles is their songs still sound great even if someone else recorded it back with Ukulele.

The point is, be it hardware gears or emulation plugins, none of those can make the songs sound great. Maybe put the blame on your songwriting for that and learn to write like the band AC/DC instead of sound like The Beatles. AC/DC records are way better than The Beatles.

 

yes for the delivery it only has to be so clear, but for the processing you get artifacts that come out effecting it.(picture an hdr image of a chrome object with different light sources and shadow sources rendered at 8k vs rendered at 1080p when you get line to what would be tiny single lines of 8k the 1080p render would not get that detail and blur and smudge it with pixelation) this is the exact tiny detail that makes analog audio what it is how it comes across do us
the fluctuations of "static" (noise floor)effecting transients the tiny amounts of compression the tiny changes in impedance the tiny etc. plus all these being effected by the actual voltage and volume of the input signal

think about doing cgi in 8k 8000 pixels and then shrinking down to 1080p delivery format. same concept.(because of the tiny detail in reflections and shadow casting)
so the reason needing to be detailed is the processing. with analog gear we are talking down to the atomic level attributes to why gear sounds so three dimensional , even though its detail is far more than we hear.
it is digital signal processing that will have swirly pixelated artifacts ( like that swirly sound mp3 has)
because we are talking about non-linearity details here.

 

Last edited: Dec 2, 2017

People say that, but no. Analog gear has a high noise floor (RF, line hum, brownian), nonlinearity (introducing harmonic & intermodulation distortion), and is bandlimited (intentionally or not). This caps how much information it contains. Think of this analog information in terms of bits: how many different voltages (shades of gray) can you stick into IN hole & be able to reliably differentiate at the OUT hole.

So anything but infinite -- nonlinearity & noise make it so that increasing the input voltage by, say, 1/20 thousandth of full scale nets a corresponding output increase only ~50% of the time (garbage). The noise floor makes it so that you'll continuously see +-1/20,000 variations at the output. See the problem? And if this wasn't bad enough, there's generation loss & ... Keep in mind that processing digital = clever algos & CPU ticks; if you can think it, you can implement it (though not necessarily in real time) on any crappy laptop, Analog requires cubic $$

You're mixing things up a bit. The "swirly" MP3 stuff happens at low bitrates (as in super lossy compression, > order of magnitude smaller file than wav/bitstream equivalent). Think extreme jpg compression, not "digital art is blurry."

 

Last edited: Dec 2, 2017

{MMJ2017 said: ↑
with analog gear we are talking down to the atomic level}

you just described exactly why analog gear we are talking down to the atomic level.
all that noise floor, rf,line hum,brownian non-linearity (introducing harmonic & intermodulation distortion), and is bandlimited (intentionally or not)
is due to electricity which is electrons. creating all those things digitally but at the detail level they exist in analog is challenging
sure you can make simple "
line drawing" versions of these artifacts but the results wont hold up to be as good as the real thing

(Analog gear has a high noise floor (RF, line hum, brownian), nonlinearity (introducing harmonic & intermodulation distortion), and is bandlimited (intentionally or not).) is ALL information

its not just about how many bits to re-create the signal but how many bits to re-create the artifacts and how they re-act at the atomic level.

its not infinite, its down to the electron. the non-linearity and noise is what im talking about trying to digitally model,
we dont need to model the original clean signal we have that already when we start!
what are you calling garbage exactly?

yes that is very detailed to model at the analog atomic level.

 

Last edited: Dec 2, 2017

Not sure I get it, making a guess -- you tell me if I got it right.

If by "real thing" you mean the original signal, then digital will come much closer than analog.
OTOH, if you mean "original signal going through a bunch of analog gear, e.g. the London Philharmonic getting played on an AM clock radio," then you're right, by definition (A child's drawing would be more exact than Rembrandt's copy of that drawing, by virtue of the copy not being perfect). Is that what you mean?

Not sure how to read that. If you mean an infinitely small change at the imput will cause a corresponding (related) change at the output, then no. If you mean something like "with the input shunted (no signal), an analog amp will send billions (maybe trillions, never thought about it) of electrons through a speaker's voice coil," then sure. These electrons are pure noise; are not caused by/have anything whatsoever to do with the sound you're not feeding into the amp.

If that's what you're trying to capture, the sound of a noisy analog amp being noisy, then an analog amp is certainly the way to fly. Like that child's drawing, it's the original.
Or go hybrid -- stick a horrible headphone amp after your DAC -- that too makes analog noise. Deliciously complex analog noise that's impossible to digitally reproduce (if we're talking exact, or even down to 1/10^9999999999999999999999999 nanobar level)

 

A copy can never replicate the original. It might come "close enough for the blues", but can only imitate that which it purports to characterize. Analog circuits impart imperfections which sound a certain way and are created by physical, hard and real means. Software attempting to replicate analogue circuits only approximate reality. What is spooky to me is that people buy into the imposter as being a one to one representation of the real thing. It is not and never can be. However the fantastic fake looks, sounds, feels or tastes, it is just an illusion and depends on some kind of machine to bring it to consciousness. That is ok up to a point but it is dangerous for humanity to accept and buy into the philosophical lie that it doesn't make a difference or matter. Eat a grapefruit or swallow a grapefruit pill that might taste like a grapefruit, looks like, etc, and has synthetic nutrients, which would you rather depend on for a healthy food source?

 

Again not sure which "original" is being talked about here. Pick one:

1. The sound hitting the mic -- that's the original.

2. The sound coming from the mic, digitally encoded and decoded back to analog, using today's technology.

3. That sound recorded & reproduced with outdated analog gear, e.g. a string quartet being played through an AM clock radio. That's the original.

 

Depends which tech is used to recreate. AA being the one closest to recreate analogue gear since - in short - it takes a photograp (well thousands of it really, it's an advance convolution approach, more detailled then say Altiverb) of what the gear does and then recreate that. The photograph is already highly detailled, and it's getting better almost by the month. Soon enought there won't be much difference between the original and the photo.

 

What's AA? Analog audio? If that, it's nothing like shop/convolution reverbs.
Photoshop is 100% digital, deals with digitized images -- maths.
Convolution reverbs are too


We're not choosing between a digital copy vs. analog "original," the only choice we have is between digital and analog copies. An Edison recorder scraping an [analog] groove in a wax cylinder is the most direct recording device, but the copies it makes are [arguably] less accurate than the best digital has to offer.

 

No not at all what I'm talking about. AA as in acustica audio and its volterra tech, which is non-linear convolution. More here< :https://www.acustica-audio.com/index.php?option=com_content&view=article&id=14&Itemid=247