Quantum Computing and Virtual Analog

Reaktor, MAX/MSP, VST/AU, etc. A place for all things soft....

Moderators: lisa, luketeaford, Kent, Joe.

Post Reply
User avatar
Kummer
Ultra Wiggler
Posts: 846
Joined: Thu Aug 28, 2014 4:50 pm
Location: St. Louis

Quantum Computing and Virtual Analog

Post by Kummer » Sat Nov 09, 2019 2:18 pm

It's a very intriguing idea, in that future analog gear would be truly obsolete other than interface.
"We are a way for the universe to know itself" - Carl Sagan

User avatar
rowsbywoof
Veteran Wiggler
Posts: 735
Joined: Tue Mar 29, 2016 9:20 pm
Location: Los Angeles
Contact:

Post by rowsbywoof » Sat Nov 09, 2019 5:36 pm

From someone who has worked on a project using a quantum computer, and after being in the room with the modern quantum computer's hulking mass, (though haven't gotten anywhere near interfacing with it), and seeing how large of an operation you need to keep it cool, and the ridiculous amounts of space it takes to even do simple equations... We're a LOONNGGG way away from quantum computing being anything more than a neat science project. I kind of doubt they'll be down to taking up an entire normal room size in our lifetime unless something pretty miraculous happens.

User avatar
Kummer
Ultra Wiggler
Posts: 846
Joined: Thu Aug 28, 2014 4:50 pm
Location: St. Louis

Post by Kummer » Sun Nov 10, 2019 12:25 pm

rowsbywoof wrote:From someone who has worked on a project using a quantum computer, and after being in the room with the modern quantum computer's hulking mass, (though haven't gotten anywhere near interfacing with it), and seeing how large of an operation you need to keep it cool, and the ridiculous amounts of space it takes to even do simple equations... We're a LOONNGGG way away from quantum computing being anything more than a neat science project. I kind of doubt they'll be down to taking up an entire normal room size in our lifetime unless something pretty miraculous happens.
Ah, yeah I knew they were big but I didn't realize how far off they would be for home use... oh well. Maybe a quantum computer can solve death soon so we can live to see it!
"We are a way for the universe to know itself" - Carl Sagan

User avatar
commodorejohn
Ultra Wiggler
Posts: 877
Joined: Fri May 03, 2013 4:19 pm
Location: Placerville, CA

Post by commodorejohn » Sun Nov 10, 2019 5:02 pm

Obligatory SMBC link.

Remember, kids: quantum mechanics only seems like magical pixie dust you can sprinkle onto any problem to make it vanish in a rainbow of happiness and smiles.
Computers: Amiga 1200, DEC VAXStation 4000/60, DEC MicroPDP-11/73
Synthesizers: Roland JX-10/SH-09/MT-32/D-50, Yamaha DX7/V50/TX7/TG33/FB-01, Korg MS-20 Mini/ARP Odyssey/DW-8000, Ensoniq SQ-80

"'Legacy code' often differs from its suggested alternative by actually working and scaling." - Bjarne Stroustrup

c_code
Learning to Wiggle
Posts: 33
Joined: Sat Aug 16, 2014 1:04 pm

Post by c_code » Wed Nov 13, 2019 1:39 pm

commodorejohn wrote:Obligatory SMBC link.

Remember, kids: quantum mechanics only seems like magical pixie dust you can sprinkle onto any problem to make it vanish in a rainbow of happiness and smiles.
:yay:

User avatar
rjungemann
Ultra Wiggler
Posts: 830
Joined: Sat Mar 03, 2012 2:18 pm
Location: San Francisco, CA, USA

Post by rjungemann » Wed Nov 13, 2019 3:01 pm

Your intuition is good in that it is true that qubits don't model just a 0 or 1, and analog signals are somewhat similar in that regard, BUT the two are wildly different in operation. I know barely enough about quantum computing to know how little I know so I won't bungle an explanation. To be honest, I don't think anyone knows what the impact of quantum computing will look like, for music, let alone domains like machine learning.

But I think there IS room for synths to use more analog circuitry in various hybrid approaches. I have heard of one synth prototype based off of FPAAs (field-programmable analog array), a circuit consisting of a bunch of analog components whose structure can be rearranged. It's a sibling to FPGAs which are equivalent but for digital signals, AND FPGAs are currently being used by released synths. That's one potential example. FPGA oscs going into analog filters is another approach in use since the '80s. Digital waveshapes through analog waveshapers is another (used by Modal synths IIRC).

As an aside, the history of analog computing is fascinating https://spectrum.ieee.org/computing/har ... g-computer
My modular (ModularGrid). Check out my new synth-pop album, Umbram.

francoprussian
Wiggling with Experience
Posts: 346
Joined: Sun Jul 26, 2015 7:54 am

Post by francoprussian » Thu Nov 14, 2019 6:54 am

Very interesting article there rjungemann.

User avatar
Kummer
Ultra Wiggler
Posts: 846
Joined: Thu Aug 28, 2014 4:50 pm
Location: St. Louis

Post by Kummer » Mon Nov 18, 2019 4:33 pm

Thanks for the thoughtful reply rjungemann & rowsbywoof! Its interesting to think about. As far as I am aware quantum computing is capable of creating truly random numbers. Extrapolating from that and as far as I understand it, a true "virtual" analog model could then be recreated in software as it's the random properties of the analog components (movement of electrons and what
not) that differentiate analog from the inherent quantization of digital technology (I have yet to check out the link rjungemann). As far as sound quality is concerned we seemed to have arrived at a point where the granularity of digital is so fine that we ourselves are unable to detect the difference in sound quality or character between an analog hardware device and a well coded digital representation of it. So no big deal, but I was just wondering if, in the future, quantum technology would allow for anything and everything to be simulated (at the subatomic level), to where the there is really no difference between whats real and whats virtual.... maybe we are there already... ? :hmm:
"We are a way for the universe to know itself" - Carl Sagan

User avatar
Kummer
Ultra Wiggler
Posts: 846
Joined: Thu Aug 28, 2014 4:50 pm
Location: St. Louis

Post by Kummer » Mon Nov 18, 2019 6:39 pm

Just found this out, but apparently there is debate among physicists at to whether or not reality is actually analog or digital in nature... seems the current consensus is that it's more analog than digital... anyways... :party:
"We are a way for the universe to know itself" - Carl Sagan

User avatar
GrantB
Super Deluxe Wiggler
Posts: 1316
Joined: Fri Apr 09, 2010 5:58 pm
Location: Oly

Post by GrantB » Tue Nov 19, 2019 2:50 am

You know what would be really next-level is to skip quantum and make a fully ANALOG computer to imitate analog sounds. If only there were some clever way to change the connections between the circuits, you could have a huge variety of realistic analog sounds. :sadbanana: :hihi:

User avatar
rjungemann
Ultra Wiggler
Posts: 830
Joined: Sat Mar 03, 2012 2:18 pm
Location: San Francisco, CA, USA

Post by rjungemann » Tue Nov 19, 2019 5:23 am

@Kummer I know you knew it's more complicated than discrete vs continuous values, but I was afraid to go too deep lest I said something incorrect haha. I wrote a wall of text for this message and then cut most of it for that reason... If anyone sees inaccuracies let me know.

(Sidebar, most of what I know about quantum computing comes from poking around with IBM Q, which you can run code on real, hosted quantum computers or in a simulator, and there are tons of docs.)

My understanding is that quantum computing is currently frequently used in a way which looks sort of like constraint solving to me, where you can set up some initial conditions and it will converge on some state with some probability.

I don't think anyone is currently using it for DSP but that certainly doesn't mean it's not possible or not useful. It's still very early days.

Digital random number generation is something I know a little bit more about. My personal opinion is that current computers probably generate "good enough" random numbers. There are "secure random" numbers used for cryptography in modern computers, which are sampled from "sources of entropy", and if these weren't sufficiently random, then cryptography wouldn't work. BUT these sources of entropy can be exhausted if sampled too frequently, so for most use cases, a pseudorandom number generator is "seeded" from a secure random number generator, and periodically re-seeded.

It would be a really cool experiment to generate noise from an analog source and from a digital source and compare the raw sounds, and also the sounds of both used as modulators!

All of that said, I own a few analog synths and a small modular, and I still feel like they have some unquantifiable spark. So what do I know :lol:
My modular (ModularGrid). Check out my new synth-pop album, Umbram.

User avatar
UltraViolet
Wiggling with Experience
Posts: 415
Joined: Fri Dec 14, 2018 7:10 pm
Location: Edinburg, Ohio
Contact:

Post by UltraViolet » Tue Nov 19, 2019 8:27 am

GrantB wrote:You know what would be really next-level is to skip quantum and make a fully ANALOG computer to imitate analog sounds. If only there were some clever way to change the connections between the circuits, you could have a huge variety of realistic analog sounds. :sadbanana: :hihi:
How about a bunch of these? :

https://www.futurlec.com/Zarlink/MT8816AEpr.shtml

I've been wanting to build something with these.

User avatar
commodorejohn
Ultra Wiggler
Posts: 877
Joined: Fri May 03, 2013 4:19 pm
Location: Placerville, CA

Post by commodorejohn » Tue Nov 19, 2019 9:33 am

A reasonably complex pseudorandom number generator is definitely good enough for audio purposes, however short it might fall for cryptography :lol: IMHO the real thing that analog synthesizers bring to the table naturally that digital synthesizers have to work for isn't so much generic-random instability so much as subtle, unintended deviations from the norm - the Moog sawtooth isn't much like an ideal sawtooth, the Korg35 filter turns from fairly-normal into a growling, shrieking maniac when overloaded or with the resonance cranked up, etc. These things aren't so much "random" per se as just different from the textbook-definition model of how these things should work - but they're wonderful.
Computers: Amiga 1200, DEC VAXStation 4000/60, DEC MicroPDP-11/73
Synthesizers: Roland JX-10/SH-09/MT-32/D-50, Yamaha DX7/V50/TX7/TG33/FB-01, Korg MS-20 Mini/ARP Odyssey/DW-8000, Ensoniq SQ-80

"'Legacy code' often differs from its suggested alternative by actually working and scaling." - Bjarne Stroustrup

User avatar
Voltcontrol
Common Wiggler
Posts: 149
Joined: Thu Jan 25, 2018 1:31 pm
Location: Sprang-capelle, The Netherlands

Post by Voltcontrol » Tue Nov 19, 2019 10:02 am

Quantum Computing and Virtual Analog. It's neither here nor there. It's all over the place.

User avatar
GrantB
Super Deluxe Wiggler
Posts: 1316
Joined: Fri Apr 09, 2010 5:58 pm
Location: Oly

Post by GrantB » Tue Nov 19, 2019 1:34 pm

UltraViolet wrote:
GrantB wrote:You know what would be really next-level is to skip quantum and make a fully ANALOG computer to imitate analog sounds. If only there were some clever way to change the connections between the circuits, you could have a huge variety of realistic analog sounds. :sadbanana: :hihi:
How about a bunch of these? :

https://www.futurlec.com/Zarlink/MT8816AEpr.shtml

I've been wanting to build something with these.
I was thinking these:
https://www.patch-point.com/banana-patch-cables

User avatar
Kummer
Ultra Wiggler
Posts: 846
Joined: Thu Aug 28, 2014 4:50 pm
Location: St. Louis

Post by Kummer » Thu Nov 21, 2019 11:07 pm

Thanks again for the replies!

Don't get me wrong I love analog synths (and digital) and own many, I'm just projecting out into the future and thinking that this whole debate (which has thankfully largely gone away) about the differences between analog and digital will one day be a moot point because for all intents and purposes we will have the real thing, in virtual form.
"We are a way for the universe to know itself" - Carl Sagan

Post Reply

Return to “Music Software”