Results 1 to 10 of 10

Thread: Nvidia AI DENOISER VS Native VPR

  1. #1
    Registered User 3dhotshot's Avatar
    Join Date
    Mar 2009
    Location
    Durban
    Posts
    63

    Nvidia AI DENOISER VS Native VPR

    Correct me if Im wrong but removing noise from a render isn't anything great as with a fast machine native lightwave vpr does this very well for many years. I smell smoke and mirrors on nvidia side only to suck up more sales for their high end cards....

  2. #2
    obfuscated SDK hacker Lightwolf's Avatar
    Join Date
    Feb 2003
    Location
    Stuttgart, Germany
    Posts
    13,473
    Well, the point is to remove noise without spending hours more on rendering. Also, you don't really need a high-end card to run the denoiser.
    Plus, the interesting part is that it's based on deep learning as opposed to traditional, more generic, noise reduction algorithms.

    Cheers,
    Mike

  3. #3

    buy a used fast PC, and add noise reduction on top of that.

    Ai denoisers might not be Fantastic, but they do help to cut down render-time quite a bit.

    a good test >
    https://www.youtube.com/watch?v=CmhbesUmqcU

    LW vidz   DPont donate   LightWiki   RHiggit   IKBooster   My vidz

  4. #4
    Registered User ianr's Avatar
    Join Date
    Oct 2006
    Location
    Chiltern Riviera
    Posts
    1,220
    Quote Originally Posted by Lightwolf View Post
    Well, the point is to remove noise without spending hours more on rendering. Also, you don't really need a high-end card to run the denoiser.
    Plus, the interesting part is that it's based on deep learning as opposed to traditional, more generic, noise reduction algorithms.

    Cheers,
    Mike

    It would be nice if Antii comes up with a port of this (Nvidia) for the next cycle release?

  5. #5
    Registered User
    Join Date
    Feb 2014
    Location
    Sweden
    Posts
    67
    I'd prefer gpu agnostic solution rather than Nvidia only, like what Octane is trying to do from what I hear.

  6. #6
    obfuscated SDK hacker Lightwolf's Avatar
    Join Date
    Feb 2003
    Location
    Stuttgart, Germany
    Posts
    13,473
    Quote Originally Posted by Morgan Nilsson View Post
    I'd prefer gpu agnostic solution rather than Nvidia only, like what Octane is trying to do from what I hear.
    Rolling your own is extremely complex and time consuming (if you want it deep learning based). I don't think NT have the resources for that.

    Cheers,
    Mike

  7. #7
    Registered User
    Join Date
    Apr 2016
    Location
    Barcelona, Spain
    Posts
    482
    Have tested Ai denoise with some Lw renders and is not working fine with adaptive, I suspect if Antii makes something it should be adapted to a some sort of progressive unbiased engine. Which could also be great.
    English is not my native language so please be patient.

    Salvador Ureņa
    http://urenasalvador.wixsite.com/portfolio

  8. #8
    Registered User 3dhotshot's Avatar
    Join Date
    Mar 2009
    Location
    Durban
    Posts
    63
    Interesting stuff octane looks like a winner ...

  9. #9
    obfuscated SDK hacker Lightwolf's Avatar
    Join Date
    Feb 2003
    Location
    Stuttgart, Germany
    Posts
    13,473
    Quote Originally Posted by Asticles View Post
    Have tested Ai denoise with some Lw renders and is not working fine with adaptive, I suspect if Antii makes something it should be adapted to a some sort of progressive unbiased engine. Which could also be great.
    Just don't expect it to be deep learning based then.

    Cheers,
    Mike

  10. #10
    Registered User ianr's Avatar
    Join Date
    Oct 2006
    Location
    Chiltern Riviera
    Posts
    1,220
    Just 'Lobby P.M.' him (Antii). Nicely of course...

    Or NT have buy in..either way!

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •