Page 2 of 6 FirstFirst 1234 ... LastLast
Results 16 to 30 of 81

Thread: GPU Rendering - Anywhere in the development pipeline for Lightwave?

  1. #16
    Registered User rdolishny's Avatar
    Join Date
    Jul 2003
    Location
    Peterborough, Ontario, Canada
    Posts
    313
    Quote Originally Posted by TheLexx View Post
    Are you rendering Lightwave in Redshift springboarding via another software, or do you just use different softwares and renderers ?
    Sorry, I don't use Redshift at all. I just love the look and speed I've seen from Maya renderings. I would switch if it was available.
    Last edited by Chuck; 02-28-2019 at 09:31 AM. Reason: fix quote

  2. #17
    Registered User
    Join Date
    Mar 2016
    Location
    Oxford, UK
    Posts
    796
    Quote Originally Posted by rdolishny View Post
    Sorry, I don't use Redshift at all. I just love the look and speed I've seen from Maya renderings. I would switch if it was available.
    I am impressed too.
    https://www.artstation.com/artwork/r...kfh7KrmGShAKNs

    Having said that it is a thin line between LW putting resources into compatibility with another renderer, only to find LW then becomes subservient to another company when the licenses change at a whim. I have zero experience of game engines but I understand they are free and render via GPU, and LW is now better equiped for those so there may be a trick or two in that.

    Maxon chose Prorender for the licensing model and it seems very sensible NT is also looking at that, though we must remember no one has promised it. I did get the impression from the Roundtable that NT is more aware than appears on the surface. .
    Last edited by TheLexx; 02-23-2019 at 07:30 AM.

  3. #18
    Quote Originally Posted by TheLexx View Post
    I am impressed too.
    I did get the impression from the Roundtable that NT is more aware than appears on the surface. .
    They also said that they won't integrate LW because too many studios don't want it. Come again?

    Anyway on GPU rendering, the best thing to hope for is that developers are making LW as friendly to third party render solutions as possible. I don't suppose looking to ProRender as an option is unrealistic, but given they let go of Mark Granger and he was interested in GPU rendering( and now subsequently works for Otoy/Octane now doing just that) they're aren't exactly putting off the vibe of going in that direction either.

  4. #19
    Registered User
    Join Date
    Mar 2016
    Location
    Oxford, UK
    Posts
    796
    Quote Originally Posted by hrgiger View Post
    They also said that they won't integrate LW because too many studios don't want it. Come again?
    That may seem odd to some but it is not inconceivable as he said it was based on user feedback . Maybe a lot of users just have their workarounds or plain don't care. Getting away from films and games it is difficult to imagine, say a police department saying a unified app is a primary concern.

    The Brent thread went a bit pear shaped, there was a user saying he was already invested in several hundred cpu nodes, so more get-it-out-quick guys may already be doing that, so GPU would be for existing small or newer guys who are not already invested but want the firepower of GPU. Do you have any thoughts on LW using game engines for GPU rendering (am I right in thinking this is already possible now) ? Btw, if you have any dealings with Lino or Mr Granger at Otoy and could give them a polite nudge that an Octane dongle should mean permanent offline, just like a LW dongle, I would buy one tomorrow.

    Last edited by TheLexx; 02-23-2019 at 09:35 AM.

  5. #20
    Registered User
    Join Date
    May 2012
    Location
    Virginia
    Posts
    431
    Quote Originally Posted by hrgiger View Post
    They also said that they won't integrate LW because too many studios don't want it. Come again?
    That is really odd to me, because even though I know that LW isn't nearly as prevalent as it was, I know that a lot of studios (small and large) still have it in their development pipeline.

    I don't know if LW uses flex development, but this is definitely one of those situations where bringing on a small target team to get plugins developed for all of the external GPU render solutions would be massively beneficial to them.

    Making LW easier to integrate into existing pipelines would be a huge push for the product. On most large teams, they often don't care what software is used until it becomes a workflow stoppage issue. If it's flexible and easily integrated and maintained (and in this case, very affordable) then it becomes even more appealing.

    For my part, I'm just trying to avoid the necessity of making a recommendation to spend $25000+ on software and hardware before the end of the fiscal year, or at least, purchase only a third of that to get us through until a native solution is on the horizon.

    It would also be great to have more third party GPU render options to present; not to knock Octane, but preferably options that are perpetual and don't include a dongle.

  6. #21
    Quote Originally Posted by TheLexx View Post
    That may seem odd to some but it is not inconceivable as he said it was based on user feedback .
    User feedback and suggesting a significant number of studios are saying that LW somehow being integrated would be showstopper are two different things. The fact is, if you're still using LW today, of course most of those people have either resigned to the split app approach or prefer it. The people who didn't like it have moved on so of course all these years later you're going to see a larger percentage of LW users who say they don't mind the split app approach. But then I would suggest that LW is keeping up with the strategy of appealing to current uses without the regard for what might attract new users to the software.


    I don't really think Mark or Lino are in a position to suggest Otoy licensing policy. But then I don't really see a lot of issues with the way things work now either...

    - - - Updated - - -
    Last edited by hrgiger; 02-23-2019 at 11:38 AM.

  7. #22
    Nothing ado about much Exception's Avatar
    Join Date
    Mar 2003
    Location
    Netherlands
    Posts
    5,560
    I have to add that the GPU based denoiser already helps to cut rendertimes by over 2/3rds in many cases. While this is a far cry from GPU based rendering, it's a significant gain. I have found using LW's native renderer to be preferable over Octane in most cases, except where long complex animations are involved. The time spent on setting up Octane at this point in time, and the effort needed to fix unexpected incompatibilities, bugs, or missing features just isn't worth it most of the time in my case.

    I hope NT keeps extending GPU support based on the way they handled the denoiser. Perhaps CUDA functionality and such can be slotted in to assist the native renderer for select functions as well?

  8. #23
    Medical Animator mummyman's Avatar
    Join Date
    Aug 2009
    Location
    Connecticut
    Posts
    1,039
    Quote Originally Posted by Exception View Post
    I have to add that the GPU based denoiser already helps to cut rendertimes by over 2/3rds in many cases. While this is a far cry from GPU based rendering, it's a significant gain. I have found using LW's native renderer to be preferable over Octane in most cases, except where long complex animations are involved. The time spent on setting up Octane at this point in time, and the effort needed to fix unexpected incompatibilities, bugs, or missing features just isn't worth it most of the time in my case.

    I hope NT keeps extending GPU support based on the way they handled the denoiser. Perhaps CUDA functionality and such can be slotted in to assist the native renderer for select functions as well?
    I haven't looked at the denoiser yet in LW. Does this work only on still frames? Or submitting an animation to a render farm...it will work?

    - - - Updated - - -

    Quote Originally Posted by Exception View Post
    I have to add that the GPU based denoiser already helps to cut rendertimes by over 2/3rds in many cases. While this is a far cry from GPU based rendering, it's a significant gain. I have found using LW's native renderer to be preferable over Octane in most cases, except where long complex animations are involved. The time spent on setting up Octane at this point in time, and the effort needed to fix unexpected incompatibilities, bugs, or missing features just isn't worth it most of the time in my case.

    I hope NT keeps extending GPU support based on the way they handled the denoiser. Perhaps CUDA functionality and such can be slotted in to assist the native renderer for select functions as well?
    I haven't looked at the denoiser yet in LW. Does this work only on still frames? Or submitting an animation to a render farm...it will work?

  9. #24
    Quote Originally Posted by mummyman View Post
    I haven't looked at the denoiser yet in LW. Does this work only on still frames? Or submitting an animation to a render farm...it will work?
    I almost never render just stills, only long animations. If I give one of my clients a 300 frame shot, he wants it slowed down to 900 frames. So animated sequences are critical for me.

    I've tried it on a few sequence of frames and it failed. The denoise effect across frames resembles dancing mpeg compression. The manual also specifically mentions that it can't be used on a render farm. Whether it's actually "can't" or "shouldn't be" I'm not sure, but results aren't promising for animations.

    As a potential Octane user, I'm skeptical of their denoise over a sequence of frames until I've actually tried it. For those who use it, does it work?

    Regards
    Last edited by Imageshoppe; 02-24-2019 at 06:18 PM.

  10. #25
    Almost newbie Cageman's Avatar
    Join Date
    Apr 2003
    Location
    Malmö, SWEDEN
    Posts
    7,650
    Quote Originally Posted by hrgiger View Post
    User feedback and suggesting a significant number of studios are saying that LW somehow being integrated would be showstopper[/SIZE][/COLOR]
    What? Really? I need to see a source for this information. LW, as far as I know it in our mixed pipeline, is not a showstopper.
    Senior Technical Supervisor
    Cinematics Department
    Massive - A Ubisoft Studio
    -----
    Intel Core i7-4790K @ 4GHz
    16GB Ram
    GeForce GTX 1080 8GB
    Windows 10 Pro x64

  11. #26
    Quote Originally Posted by Cageman View Post
    What? Really? I need to see a source for this information. LW, as far as I know it in our mixed pipeline, is not a showstopper.
    Maybe you misunderstood. Someone from NT said they won't unify LW because too many studios don't want it. I mean first, are there a significant number of studios even using LW, even in a mixed pipeline and secondly what about unification would make it unusable? But it's a moot point since LW will never be unified. They went on to say that unification was about procedural modeling. Which again is a strange statement since procedural modeling is only a small part of overall unification benefits.

  12. #27
    Quote Originally Posted by Cageman View Post
    What? Really? I need to see a source for this information. LW, as far as I know it in our mixed pipeline, is not a showstopper.
    Maybe you misunderstood. Someone from NT said they won't unify LW because too many studios don't want it. I mean first, are there a significant number of studios even using LW, even in a mixed pipeline and secondly what about unification would make it unusable? But it's a moot point since LW will never be unified. They went on to say that unification was about procedural modeling. Which again is a strange statement since procedural modeling is only a small part of overall unification benefits.

  13. #28
    Super Member
    Join Date
    Sep 2003
    Location
    Washington DC
    Posts
    3,627
    Quote Originally Posted by hrgiger View Post
    Maybe you misunderstood. Someone from NT said they won't unify LW because too many studios don't want it. I mean first, are there a significant number of studios even using LW, even in a mixed pipeline and secondly what about unification would make it unusable? But it's a moot point since LW will never be unified. They went on to say that unification was about procedural modeling. Which again is a strange statement since procedural modeling is only a small part of overall unification benefits.
    OK Ill bite...

    Why not support both?
    Have old school poly style modeling in Modeler and procedural modeling in Layout?

    The way both styles of modeling achieve results is distinctive enough that this could prove to be feasible.

    -R
    This message does not reflect the opinions of the US Government, CG Networks or CGTALK.com. The opinions expressed on this posting are on my own volition.

  14. #29
    Medical Animator mummyman's Avatar
    Join Date
    Aug 2009
    Location
    Connecticut
    Posts
    1,039
    Quote Originally Posted by Imageshoppe View Post
    I almost never render just stills, only long animations. If I give one of my clients a 300 frame shot, he wants it slowed down to 900 frames. So animated sequences are critical for me.

    I've tried it on a few sequence of frames and it failed. The denoise effect across frames resembles dancing mpeg compression. The manual also specifically mentions that it can't be used on a render farm. Whether it's actually "can't" or "shouldn't be" I'm not sure, but results aren't promising for animations.

    As a potential Octane user, I'm skeptical of their denoise over a sequence of frames until I've actually tried it. For those who use it, does it work?

    Regards
    Thanks for the information!

  15. #30
    Registered User
    Join Date
    May 2012
    Location
    Virginia
    Posts
    431
    Quote Originally Posted by hrgiger View Post
    Maybe you misunderstood. Someone from NT said they won't unify LW because too many studios don't want it. I mean first, are there a significant number of studios even using LW, even in a mixed pipeline and secondly what about unification would make it unusable? But it's a moot point since LW will never be unified. They went on to say that unification was about procedural modeling. Which again is a strange statement since procedural modeling is only a small part of overall unification benefits.
    I definitely misunderstood what you were talking about; I also thought by integration you were referring to integrating LW into their production pipeline (per a general discussion on expanding GPU render support or native GPU render support to increase the appeal of Lighwave) - not unifying Modeler and Layout.

    To get a definitive answer on a significant number of studios, I think we would first have to define what is considered a studio; then gather data on the use of Lightwave.

    For my part, I'm just always surprised where Lightwave pops up in the development of major film and television productions (the recent Academy Award winner Green Book, for example).
    Last edited by RPSchmidt; 02-25-2019 at 07:20 AM.

Page 2 of 6 FirstFirst 1234 ... LastLast

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
  •