Sound Resolution Upscaling

This forum is for discussing Reason. Questions, answers, ideas, and opinions... all apply.
Post Reply
User avatar
Chizmata
Posts: 918
Joined: 21 Dec 2015
Contact:

18 Jan 2021

hey everyone, i was wondering if this is or will be a thing in music/reason sooner or later?

proper upscaling is a new thing for game visuals, DLSS 2.0 is a technology that allows games to be displayed at high resolutions while maintaining a high framerate. the key to this technology is ai-based upscaling of a lower resolution. it has basically no quality loss compared to native resolution. yes, its as voodoo as it sounds but works.

now for a DAW that could mean that you could be able to compose at 192k Hertz rate while not needing not much calculation power at all. Maybe you could pack your rack until it bursts and have it run stutter-free because you have an AI upscale the sound resolution from 22k to 96k (DLSS 2.0 can absolutely upscale 4 times, 1080p to 2160p/4k, 4 pixels instead of one).

is this technology already here, in the making or not useful in music after all?

User avatar
Loque
Moderator
Posts: 11176
Joined: 28 Dec 2015

18 Jan 2021

Great times than. I let the customer upscale to whatever he/she needs :-)

Nothing heard of it yet. But i already saw image scaling adjusted to audio, but not AI based.
Reason12, Win10

User avatar
Billy+
Posts: 4157
Joined: 09 Dec 2016

18 Jan 2021

It's not up scaling, but over sampling is a thing already...


User avatar
Chizmata
Posts: 918
Joined: 21 Dec 2015
Contact:

18 Jan 2021

Loque wrote:
18 Jan 2021
Great times than. I let the customer upscale to whatever he/she needs :-)

Nothing heard of it yet. But i already saw image scaling adjusted to audio, but not AI based.
you can see a really quick example here:


theres a lot more in depth stuff on youtube

User avatar
Chizmata
Posts: 918
Joined: 21 Dec 2015
Contact:

18 Jan 2021

Billy wrote:
18 Jan 2021
It's not up scaling, but over sampling is a thing already...

yes oversampling is exactly what you could achieve easily with the upscaling tech

User avatar
adfielding
Posts: 959
Joined: 19 May 2015
Contact:

22 Jan 2021

I think the key thing here is that DLSS is a consumer-side solution. As such, DLSS serves no real purpose for anyone on the developer/creative side.

I don't think it would be technically possible to use AI "upscaling" in a professional audio capacity, at least not in the way you're describing. Even on the consumer side, audio systems are (typically) considerably less computationally taxing than rendering real-time graphics - speaking in terms of pure playback as opposed to real-time audio systems, I would argue that the bigger issues with audio are with regards to destructive mastering practises rather than any logisitcal or technical consideration.

I think DLSS is a wonderful solution for rendering consumer-side real-time graphics, but I don't think there exists the same need for users to reduce the load of real-time audio systems in consumer-based media (i.e. games/film/TV).

Proboscis
Posts: 1004
Joined: 28 Aug 2019

22 Jan 2021

It would be great if Reason upscaled to get away from 'the reason sound'. It's 2021, and time for them to get up to speed with the default sonic output.

I'm hoping they code in the Reaper sound when R12 drops.

User avatar
bxbrkrz
Posts: 3814
Joined: 17 Jan 2015

22 Jan 2021

NVIDIA DLSS 2.0. Is it an open platform? Can Intel, Apple, Samsung, AMD, Amazon, Raspberry Pi be able to use it, without using a NVIDIA graphic card? If the answer is 100% yes, then it's a great tool.
757365206C6F67696320746F207365656B20616E73776572732075736520726561736F6E20746F2066696E6420776973646F6D20676574206F7574206F6620796F757220636F6D666F7274207A6F6E65206F7220796F757220696E737069726174696F6E2077696C6C206372797374616C6C697A6520666F7265766572

Post Reply
  • Information
  • Who is online

    Users browsing this forum: No registered users and 26 guests