[Sidefx-houdini-list] Rendering DeepExr for Nuke Comp

Olly Nash ollimation at gmail.com
Fri Aug 3 06:41:13 EDT 2018


Hey Alex,

Thanks for your reply.  Yes, I think you are probably right.  It will no
doubt work straight away when tested with a sphere ; )

Olly : )


On Fri, Aug 3, 2018 at 11:23 AM Alex Doss <alexdoss at gmail.com> wrote:

> Im not able to walk u thru it, i apologize for it. Since I started learning
> Hou, Ive only used Redshift for rendering.
>
> On the deep situation. =) It might be that you are only allowed to merge
> the data by Depth or by ID.. not both, not anything else. Do u know what
> data u're setting into the file?
> My guess would be that maybe you are writing "illegal" data, so to speak
> (houdini would do whatever you ask it to), so from nuke's perspective, its
> just bollox (?).
> Have you tried opening the file elsewhere, houdini comp for example? the
> beauty EXR should be visible anyways.
>
>
> On 3 August 2018 at 11:44, Olly Nash <ollimation at gmail.com> wrote:
>
> > I'd love to get to that point - 11.1Kb per file isn't cutting the
> > mustard!   No errors apart from when loading into nuke
> > *"early end of file: read 0 out of 8 requested bytes".*
> > The only line in the RenderROP verbose read out that makes me wonder is:*
> > Deep Camera Channel: half[4] "C"*
> > What's that about?
> >
> > ah well, i'm sure i'll figure it out eventually.  Or maybe it's just this
> > .hip file, who knows!
> >
> > Olly : )
> >
> >
> > On Fri, Aug 3, 2018 at 7:59 AM Alex Doss <alexdoss at gmail.com> wrote:
> >
> > > I second Mathieu's words there. The main point around deep tech is
> > encoding
> > > data per pixel. Merging that table to the minimum data you need is
> > > essential to pull it out alive =). Otherwise things get pretty heavy
> > pretty
> > > quick.
> > >
> > > On 2 August 2018 at 21:32, Olly Nash <ollimation at gmail.com> wrote:
> > >
> > > > Thanks loads for checking it out for me, Mathieu, I really appreciate
> > it.
> > > > And thanks for the tips for once I have got it working.  Its a bit
> of a
> > > > pain to upload stuff tbh.  I've sent some files to sidefx support so
> > > > hopefully they might be able to gleen why it's not working without
> > > > producing an error the Houdini side of things.
> > > >
> > > > houdiniFX 16.5.268
> > > > Linux 2.6.32-573.3.1.e16.x86_64
> > > >
> > > > I'll update this thread when I know what the problem / resolution is!
> > > > Thanks a million,
> > > >  Olly : )
> > > >
> > > >
> > > >
> > > > On Thu, Aug 2, 2018 at 4:57 PM Mathieu Leclaire <
> mleclair at hybride.com>
> > > > wrote:
> > > >
> > > > > Yeah... just use Deep Resolver to Deep Camera Map and change the
> DCM
> > > > > Filename to a .exr and you should technically get a valid deep exr
> > > > > render. Not sure why it's not working.
> > > > >
> > > > > What you might want to do once you got your set-up working in order
> > to
> > > > > get better compression for your deep exr is to expose the "DCM
> > Z-Bias"
> > > > > and "DCM Compression" parameters so you can tweak the sample
> merging
> > > > > options. That can really help lighten up your file size. You have
> to
> > go
> > > > > to "Edit Parameter Interface..." and go into the "Render
> Properties -
> > > > > Mantra - Deep Output" and add them to the Existing Parameters to
> > expose
> > > > > them in the interface. Adjusting these can really lighten up your
> > deep
> > > > > exr file size and really accelerate the deep processing speed in
> > Nuke.
> > > I
> > > > > highly recommend it to have a good workflow.
> > > > >
> > > > > -Mathieu
> > > > >
> > > > > On 02/08/2018 11:18 AM, Martin Winkler wrote:
> > > > > > On Thu, Aug 2, 2018 at 5:10 PM, Mathieu Leclaire <
> > > mleclair at hybride.com
> > > > >
> > > > > wrote:
> > > > > >> Are you rendering your Exr in scanline mode? I haven't used Nuke
> > in
> > > a
> > > > > while,
> > > > > > It should work "just like that". No special stuff needed to
> render
> > > > > > exrs with deep data, there must be something else wrong.
> > > > > >
> > > > > > On Wed, Aug 1, 2018 at 6:05 PM Olly Nash <ollimation at gmail.com>
> > > wrote:
> > > > > >> Can anyone point me in the right direction to successfully
> render
> > a
> > > > > >> DeepExr sequence for a comp'er to read into Nuke, please?  It's
> a
> > > > .bgeo
> > > > > >> filecache of a pyro simulation.
> > > > > > Can you share the hip file (with some dummy data instead of the
> > real
> > > > > > bgeos if it's a production scene)? Also some info on which
> version
> > > you
> > > > > > use etc.
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > > > _______________________________________________
> > > > > Sidefx-houdini-list mailing list
> > > > > Sidefx-houdini-list at sidefx.com
> > > > > https://lists.sidefx.com:443/mailman/listinfo/sidefx-houdini-list
> > > > >
> > > > _______________________________________________
> > > > Sidefx-houdini-list mailing list
> > > > Sidefx-houdini-list at sidefx.com
> > > > https://lists.sidefx.com:443/mailman/listinfo/sidefx-houdini-list
> > > >
> > >
> > >
> > >
> > > --
> > > Alex Doss
> > > Mobile: +31 (0) 6 5437-2515
> > > web: www.alexdoss.com
> > > _______________________________________________
> > > Sidefx-houdini-list mailing list
> > > Sidefx-houdini-list at sidefx.com
> > > https://lists.sidefx.com:443/mailman/listinfo/sidefx-houdini-list
> > >
> > _______________________________________________
> > Sidefx-houdini-list mailing list
> > Sidefx-houdini-list at sidefx.com
> > https://lists.sidefx.com:443/mailman/listinfo/sidefx-houdini-list
> >
>
>
>
> --
> Alex Doss
> Mobile: +31 (0) 6 5437-2515
> web: www.alexdoss.com
> _______________________________________________
> Sidefx-houdini-list mailing list
> Sidefx-houdini-list at sidefx.com
> https://lists.sidefx.com:443/mailman/listinfo/sidefx-houdini-list
>


More information about the Sidefx-houdini-list mailing list