Hi Torax, Deep images store multiple samples for each pixel, when you use deep from image you're taking a regular non deep image with or without z values and resampling it into deep data. Each pixel will then have exactly one sample - if the input image had a Z value it will be stored at that depth, if not it will be stored at a depth provided by the node parameters.
If you use deep to image on multiple images it will then create multiple samples per pixel - so you're effectively creating a volume of deep data as if they were slices from a volume render ( though the z values may not all lie on the same plane ). This works well if you need to create deep elements from live action footage and is similar to how you would layer multiple smoke elements in a comp, but allows you to integrate it nicely with CG rendered deep data. If you're generated volumetric data from a renderer it would be tedious to render out all the slices and combine, you'd be much better off using a renderer that allows you to output the deep information. Although Nuke ships with a reader for PRMan dtex files hopefully over the coming months more readers will be made available for Houdini, 3delight, etc. and eventually OpenEXR 2.0 will be released - though I imagine there will be a waiting period while all your favourite renderers ship with OpenEXR 2.0 support. Cheers On 2011-08-25, at 4:13 AM, Torax Unga wrote: > Wow is taking me a while to reply.... busy times. > > ...The resulting deep image is not all that deep ( exactely one sample per > pixel )... > > "Not all that deep"... do you mean that I will still have issues like aliased > edges on overlapping geo (like the standard depth comp technique). > > ...sample smoke sims and then merge them into a real shot using deep image > tools... > > When you sat "sample" are you talking about rendering "slices" of the smoke > element that can later be assembled into deep samples? It would be great to > see an example of this... lets say with Maya or Houdini fluids. > > The deep from frame nodes creates the deep information based on time, the > resulting data is frozen in time (isn't it?)... that wont work for moving > smoke. > > Maybe I'll just have to wait for this technology to mature more... no plugin > writing skills here. > > T > > > From: Stuart Bruzek <[email protected]> > To: Nuke user discussion <[email protected]> > Sent: Friday, August 19, 2011 10:30 AM > Subject: Re: [Nuke-users] Deep Images coming from Houdini > > sorry for the repost. > > On Fri, Aug 19, 2011 at 10:30 AM, Stuart Bruzek <[email protected]> > wrote: > I am not sure if i understand your question properly. If you wanna know of > you can create deep images from regular images then yes you can provided you > give them a depth matte. The resulting deep image is not all that deep ( > exactely one sample per pixel ) but is usable with all the tools and can be > combined. > If you ask if you can render sample smoke sims and then merge them into a > real shot using deep image tools yes you can, i am just not sure how the > foundrys implementation on 2 or 3D transforms is... if not its one of THE > things to have >> have to write some plugins ;) > > JO > > On Fri, Aug 19, 2011 at 3:42 AM, Torax Unga <[email protected]> wrote: > same > - Show quoted text - > > _______________________________________________ > Nuke-users mailing list > [email protected], http://forums.thefoundry.co.uk/ > http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users > Reply > Reply to all > Forward > Invite Johannes Saam to chat > Reply > Jonathan Egstad [email protected] to Nuke > show details 9:34 AM (54 minutes ago) > Try the DeepFromFrames node to get multiple z slices. > > On Fri, Aug 19, 2011 at 9:34 AM, Jonathan Egstad <[email protected]> > wrote: > Try the DeepFromFrames node to get multiple z slices. > > -jonathan > > On Aug 18, 2011, at 9:23 PM, Johannes Saam wrote: > >> I am not sure if i understand your question properly. If you wanna know of >> you can create deep images from regular images then yes you can provided you >> give them a depth matte. The resulting deep image is not all that deep ( >> exactely one sample per pixel ) but is usable with all the tools and can be >> combined. >> If you ask if you can render sample smoke sims and then merge them into a >> real shot using deep image tools yes you can, i am just not sure how the >> foundrys implementation on 2 or 3D transforms is... if not its one of THE >> things to have >> have to write some plugins ;) >> >> JO >> >> On Fri, Aug 19, 2011 at 3:42 AM, Torax Unga <[email protected]> wrote: >> same >> >> >> >> -- >> vfx | Johannes Saam >> web | www.johannessaam.com >> work | www.fuelvfx.com >> _______________________________________________ >> Nuke-users mailing list >> [email protected], http://forums.thefoundry.co.uk/ >> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users > > > _______________________________________________ > Nuke-users mailing list > [email protected], http://forums.thefoundry.co.uk/ > http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users > > > > _______________________________________________ > Nuke-users mailing list > [email protected], http://forums.thefoundry.co.uk/ > http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users > > _______________________________________________ > Nuke-users mailing list > [email protected], http://forums.thefoundry.co.uk/ > http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
_______________________________________________ Nuke-users mailing list [email protected], http://forums.thefoundry.co.uk/ http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
