How to do this ?
-
My earlier 'Proximity' tool [part of Matrix-Proximity] did this swapping of three types of component definition with affix [+/~/-] depending on each instance's distance from the camera's eye... BUT you had to do it scene by scene - adding detailed instances near the camera at each view point then rendering each in turn etc... It keeps poly count down but is cumbersome...
-
I thought he was asking about having a component replaced dynamically from a set of already reduced versions...
-
That was it Matrix Proximity. I knew I was wrong with XRef. But what does XRef do?
-
@thomthom said:
I thought he was asking about having a component replaced dynamically from a set of already reduced versions...
Or that.
-
He was asking for automation.
Proximity is that, but manual... BUT you could adapt the idea to [semi]automatic but it'd be a massive overhead in processing the changes...[Xref Manager lets you update external SKPs that are inserted as components, if they've been changed since they were originally added [or last updated] into this SKP etc...]
-
Yea, I think that auto-update when the camera moves will in all likely hood cause too much of a performance hit. And there is no knowing if making changes to the model might interfere with other plugins or whatnot.
But being able to define a set of distance ranges and components - which would be adjusted on the click on a button, that would be interesting for staging scenes for rendering from various angles. -
Proximity does that [has done for years] - you set the distances and if any components have the matching 'proximity' types with the affix +/~/- [near/mid/distant] and they are within the specified distance ranges then they get swapped in that view to be the appropriate type...
Thus any detailed trees are used near the camera in that view and simple ones further away... -
thank you for your interest into this.
I will try the poly reduction techniques this morning and test TIG's plugins
still I go with ThomThom :
*by thomthom on Tue Jul 05, 2011 11:24 pmWhat he's talking about is LOD ( Level Of Detail ). He want to use low poly mesh when the camera is far away and high poly when the camera is close. Automated.
(...) being able to define a set of distance ranges and components - which would be adjusted on the click on a button, that would be interesting for staging scenes for rendering from various angles.*
this is exactly the issue, i am into film making these days... with lots of statues / artifacts / buildings.
very large set. -
So it is related to animation making you want this LOD feature?
Flythrough, or do you use an animation plugin? -
it will be fly through
no animation involved
yet we need motion blur, but it'll be post prod I imagine
-
You could envisage a version of Proximity linked to a Scene-Observer so as a scene changed the components would get swapped around but it will be sloooooow... Perhaps making it as image frames rather than a movie would be best ?
-
well that is the idea
image frames : 25 fps
render with ???? not decided yet
processed in Adobe Premiere or equivalent
-
Why not try OpenSpace 3D ?
It's in French made by Frenchy people
I have made a little video for Sketchup to OpenSpace3D -
c'est ta voix sur la video Pilou ?
le commentaire est hilarant !je ne crois pas que ça fasse le match pour nous.
en fait on va faire des mouvements de caméra dans un modèle et des rendus plan par plan dans un moteur à déterminer
-
we are aiming at producing about 3 movies of 3 mn each
4500 frames per movie
13500 frames total
we are looking for an engine that could run a render farm any idea ?
-
Normally I recommend V-Ray for SketchUp, but the current version has a bug in their distributed render engine so it is currently unpredictable.
Thea support render nodes - another one I have tried. And I think you can place proxies in Thea, so you can place low poly components in SU and replace them with high poly in Thea. (If I'm not mistaken. I've not gotten around to play around with Thea that much yet.) Other members here can confirm or ridicule me.
-
@Bertier
Oui, on ne se refait pas, s'il l'on n'est pas un imitateur de premier ordre!
Oui effectivement c'est peut-être pas le logiciel le plus approprié
Il est plutôt fait pour de la balade interactive ou d'exploration quitte à faire des enregistrements
vidéos de l'écran
Par contre pour de la vidéo à fond les manettes il y a Lumion,
mais pour de la sculpture je ne sais pas ce que cela donne!
Advertisement