Guy nailed it.

 

This is pretty much exactly what I am getting going on about, and this idea is not farfetched. Just for a moment play with the idea. Not all that long ago planes flying in the air was far fetched science fiction and sonic high tech was scratching wax cylinders at narrow bandwidth. That was only 100 years after I graduated high school. The trajectory of the science of sound transmission has rocketed us to where now even cheap ass AD-DA converters are pretty darn good and almost anybody can now be in contention for a Grammy by recording in the comfort of their own bedroom. We went from big and bulky AM radios to the Walkman in a literal blink of an eye in perspective of time in regard to human scientific endeavor.

The brain is being mapped out. It is incredibly complicated bit of kit, but in a relative short time we will come to the understanding of how and why this hardware/software phenomenon ticks. Experiencing a performance without the limitations of hardware getting in the way and mucking up the process is the logical next step.

 

What are these hardware limitations that you and others are apparently trying to transcend? I'm not claiming that there aren't any, but one thing I'm curious about is what people think they're missing out on.

If we keep playing into my tongue-in-check analog-worshiping hipster idea of bypassing air vibrations, what does that gain us? Well, no reverb, for one. I think after millions of years of listening to sound through air, the human brain would probably not enjoy listening to sounds that contain zero reverb and lack the (primarily) high-frequency dampening that results in natural space.

How can we make sound better if we can already capture sound with an incredibly low noise floor, more dynamics than your ears can handle, wider frequency range than your ears can handle, and an incredibly low amount of time artifacts (no wow or flutter)? Sound staging? Okay, but that's not a limit in digital reproduction. That's a sound reproduction issue because no one wants to go around with binaural headphones all the time. That's a totally separate avenue from modern digital being able or not able to capture a signal with sufficient fidelity.

If people want to argue that most microphones or speakers impose a physical limitation on capturing/reproducing a signal then I'll buy that kind of argument, but that's not the same argument as "transcending" digital. Transcending digital (or analog for that matter) is a nonsensical statement because any digital or analog system can be improved for greater performance.

We're at a point where the improvement of audio is reaching a ceiling of rapidly diminishing returns. Quantum computing is fun to say, but what advantage is it going to gain you? You can do tons of calculations at the same time but the result has to collapse to something that can actually be listened to, either a digital expression of a wave or an actual analog wave form. At which point, we're back where we started.

Okay, so quantum computers might be able to let you model a vintage console in a VST more accurately or something, but that's just computing efficiency and has nothing to do with the method of capturing and reproducing sound in a delivery medium sense.

Can anyone identify what improvements need to be made to current digital audio tech for greater verisimilitude?

 

Last edited: Aug 19, 2016

While technology is evolving humanity will regress......

 

In my opinion, there's very little wrong with current digital as far as pure signal processing goals. The biggest flaws are in the mechanical mechanism of recording (mics) and also reproduction methods (speakers) and (primarily) listening environment (room/space).

Think about it this way: You've got a microphone that's a rough analog of a human ear. Clearly some mics are better than others, but they all have the drawback of at least some kind of physical inertia because they're dealing with sound pressure waves on a mechanical surface, right?

Then you have speakers which run into the same problem. You're taking a signal and driving something with mass and therefore inertia. Then you've got all your crossover networks and amplification problems, etc.

But then you've got your room. This is the killer. This is what's making everything sound (relatively) horrible and compromised because you're taking an original performance environment (the live venue, band in the studio, or artificial reverb space, etc.), capturing its reflections and acoustic properties and then preserving that for reproduction. That's all well and good, but then you play that back through your speakers and that original signal (already complete with its own spacial information!) gets chaotically bounced around whatever space you happen to be listening in. So you're essentially listening to two rooms at once. Anyone see a problem with that?

Bueller?

One way around this would be to have an ideal mechanical ear (a human tympanic membrane grown in a vat, or whatever) capture the signal in its original space, much as a traditional microphone would do. Then you'd take that signal and implant it directly into your auditory nerves, bypassing your listening space and your own ear and its probable flaws. Toddlers shouting right into your ear canal, right?

But this has nothing to do with analog versus digital as far as signal integrity goes. It has nothing to do with the quality of or performance of the medium and everything to do with the methodology of signal capture and reproduction. It has to do with the physics of imposing the acoustics of one space over the acoustics of another space and the unwanted step of converting to a substitute medium (microphones -- whether analog or digital recording, doesn't matter) and then back into physical space again (speakers).

 

Last edited: Aug 20, 2016

You miss my point...of course there is reverb. You would actually be remembering a Live performance as it was played in the room and recorded through the technology able to capture the actual performance as experienced by the TPDC.This could also change the way we learn, say playing a musical instrument. By experiencing an action as a memory would enable one to have a shorter learning curve in regard to any physical activity.

 

I get where you're going with the idea, but memory injection (experienced by someone else) is overkill compared to the idea of just stimulating the auditory cortex, and has some undesired drawbacks. I mean, it would be a cool technology in its own right, absolutely, but it does pose some problems.

Like you mentioned, filtering out just the audio aspects from the rest of the memory, having to screen candidates, etc. But also it requires a live person to do the recording. Meaning if I want to get a sound recording of some event underground beneath a volcano or at the bottom of the ocean or something then I've got to stick a person down there. Right now I can just toss a microphone in the middle of the woods and record 28 uninterrupted hours of bears looking for trout or whatever my purpose may be (you know, field recording).

Also, TPDC only applies to a live venue and doesn't work well with the recording process or synthetic music. Right now I can create something on the computer (never even listening to it!) and distribute it to other people. If we're going with the TPDC idea then I'd have to play the music out of speakers (or some other means of reproduction) into an environment which is going to color whatever I have generated in-the-box. That means LESS fidelity, not more.

Let's assume I have tinnitus but want to record a live concert of Radiohead for an ultra-HiFi pay-what-you-want download (or whatever). I'd have to hire someone with perfect hearing to come stand in one spot and never go to the bathroom, turn their head, etc. That's not very practical because right now I can just go plug into the mixer and get a consistent product.

 

Last edited: Aug 20, 2016

I remember my direct to disk vinyl JBL speakers and mac and Crown amps ! Now i get mp3's and an ipod -yeah the future is great.

 

I feel like many people here are asseerting that Quantum computers are the magic bullet to making everything related to computing better.

Let's see if this is the case:

QC = Quantum computer
CC= classical computer

Qubits DON'T store an infinite number of states. "A single qubit can represent a one, a zero, or any quantum superposition of those two qubit states"

Lets say you have an awesome 10s mono 44/16 guitar solo and wish to process it using a QC. We use 16qubits instead of 16bits to represent a sample.

If you use qubits to store 0 or 1, it's no diffrent to a CC, and is most likely slower.

If you use qubit superposition, your awesome guitar solo has turned into, most likely, white noise before you even do anything.

Why? using superposition, there's a 1/44,000x16x10 chance that what's in the QC is what you put in, or <0.1%. This is worse than soot paper, https://en.wikipedia.org/wiki/Phonautograph where you store about 60% of what you put in, or even an LP (~95-98%). Your

So inherently, QCs are NOT good for anything that are inherently definite to start with (like audio).

SO what are QCs good for then?

perhaps some kinda fuzzy problems, but so far, the only things they're better at, is simulating themselves, which is quite useless.

so far: "The overwhelming part of classical calculations cannot be accelerated on a quantum computer.[89]"

There are a lot of hurdles to making one do anything close to what resembles definite numbers, faster than CC.

• scalable physically to increase the number of qubits;
• qubits that can be initialized to arbitrary values;
• quantum gates that are faster than decoherence time;
• universal gate set;
• qubits that can be read easily.

ref: https://en.wikipedia.org/wiki/Quantum_computing

RE "digital staircasing", or discreteness, with qubits, you're still gonna get it. And for EVERYTHING recorded, there is nothing worth >20bits. WHY? We're hampered by the analogue world through electrical, thermal noise, etc.

Also, the analogue world is not completely limitless. There exists planck time & length. https://en.wikipedia.org/wiki/Planck_time

"the reciprocal of the Planck time can be interpreted as an upper bound on the frequency of a wave."

And you'll hit into physical and natural limits way before you hit planck F. I've studied HR classical recordings. At 96/192k, there comes a point when the high frequencies of the instruments just sinks into the noise floor.

 

Last edited: Aug 20, 2016

Digital Audio is not stair stepped, you guys need to stop doing grade school math thinking that what comes out of a DAC is not what went in.

If you want to talk about band limited recording and harmonic distortion, per-ringing in linear phase filters, aliasing in filters and quantization error and you have the hearing of a bionic bat then maybe there is a discussion to be had. Though the gear one uses to record signals is also band limited and no linear as well.

junh1024 is probably talking above a lot of people heads here but now he has let the rabbit out of the hat, and there it is in text.... quanta are everything and so in a way the universe itself it a fixed number scale of quanta and their interactions.

The improvements areas that can be improved are human hearing, the space in which you listen to things, and the playback systems that delivers the sound.

Also digital doesn't mean MP3 or whatever flavor of shit lossy codec is you favorite flavor this month for playback or streaming. Most of the morons on this planet think the radio sounds good, so to them 128Kbps is pristine. Flac is lossless and so are all flavors of PCM or DSD what you put in with the physical limitations of the inputs is what comes out according to the physical limitations of the output stage. Even crap DACs are fine for listening to music in all but a very controlled and well calibrated environment.


Tape is cool, am old enough to have bough records and tapes when that's all that was out there to be bought and I still am so old I still have a HiFi, a truntable, and a tape deck to listen to those old recordings. There is something there and it sounds nice but so does digital and digital is much cleaner when done right.

The crap you are getting today at -6LUFS with a mountain of bass on the bottom has nothing to do with digital quality, it has more to do with human stupidity and what people like or what people think people like. As long you keep buying it, then the people who make it will assume that nothing is wrong with it and keep making it the same way.

 

Quantum computers are not classical computers. It's like apples and squids.They solve different types of problems.I think D-waves approach is best suited for these types of devices.

Google has just funded there own quantum computer lab aimed at A.I. research.First thing on there agenda reduce decoherence time of there q-bits.

 

Last edited: Aug 20, 2016

I really wouldn't want to call DSD lossless. DSD is a very asinine way of capturing audio that PCM has done perfectly well for decades. When you stay within the DSD domain, it's lossless, but that's the end of the story. If you want to play it back, you'll need to convert to PCM, which isn't lossless, and you may need to apply additional DSP such as LP EQ so that you don't damage equipment more than you need.

But I do agree with most of your other points.

You summed up one of my main points pretty nicely ispz, thanks.