conception of the hard work on the version 1.4

[denisgo22]

one old animation project previously made for the final render and now updated in Corona//
Render time of one frame is still too very large for commercial work's, but
quality and speed of adjustment's very help's and faster than Vray, for example///
But 1 hour for one frame is to much for this kind's of scenes,/this without DOF and MB/- for reasonable quality without visible noise///
In Vray is 15-12 minutes in HD 1920/1080 frame////
In old final render it was 5 min/frame///
I still waiting for version 1.4 or new Daily Builds with adaptivity and etc///

Posts are welcome
Thanks//

[PROH]

:)

[Christa Noel]

nice creative animation!

[mitviz]

very nice and i do admit its too much time for one scene, i guess we are all waiting for the major update to drop with all the shiny speedy new features

[maru]

What is your hardware? This scene doesn't look very complex, so 1 hour sounds like too much for a modern CPU. It probably could be optimized without using too much "fakes".

[denisgo22]

Thanks for comments Guys///
I am using comp I-7 3770 32mb RAm and renderfarm from 4 machines I-7 3770 with 16Ram/
On one comp rendertime about 58 minutes ,  16 sampling Balance, MSI 12/ 200 passes in resolution Full HD 1920/1080/ without DOF and MB /this I've been doing in post/, with small noise image/
This scene no so comlplex and heavy, without H Polygonal geometry and  large number of light's//

I tried to optimize the render adjustments and received as a result of rendertime 32-40 min/
With default material rendering of the scene taking 15-10 min 1920/1080 px
I guess, Perhaps too large reflective  and glossiness surfaces in this scene///
So in Corona not so many opportunities to fake and adjustment settings, such as in Vray, for example///

[Nekrobul]

Something is wrong there. Totaly.

Last animation i rendered was taking 1h for 100 pases with 16 PT samples 20 MSI on a single 3930k.

And i had all sorts of triky stuff there: DOF, Translusency, Opacity, displacement, Volumetrics, And 20B polys. With 8x double core X5660 it was taking 5 mins per frame where half of that was parsing due to 6 of 8 have 16GB of ram.

Btw you could try to lower MSI to 1 and you will need to render something like 50 pases instead of 200 in cost of bias result.

[denisgo22]

MSI to 1?!! Its perfect!
but I need to get a little bit quality
and realistic materials??
can still be put Max Ray Depth on 5, same in Vray???It's faster, but what result you get, this a question///

[Nekrobul]

MSI to 1?!! Its perfect!
but I need to get a little bit quality
and realistic materials??
can still be put Max Ray Depth on 5, same in Vray???

Lowering Max ray depth will cause more *destruction* then lowering MSI. Maru posted somwere tests between different MSI values.

Trust me i am dooing this since A6.

[denisgo22]

Thanks Guys///
I will try///

[maru]

There isn't much GI going on, so MSI@1 might work, unless it makes some things much darker.

I would also try increasing GIvsAA to 32-64 - it looks like you do not need much AA in your renders, especially that it's rendered in high resolution. This should make rendering less noisy in less time, within less passes.

Another, but more painful and destructive idea is to use the rayswitch material. You can use it to remove/change the way objects are visible to GI, reflections, etc.

[denisgo22]

Thanks Guys very much///
I picked up some of the options and it works in some cases///

[mitviz]

did lowering msi work? what are now the render times? i am using all defaults for rendering mostly and its taking forever sometimes

[denisgo22]

msi-4, Gi vs AA-32, rendertime without DOF and MB-- 20 min///And image some darker//but nevertheless I believe that it is better to start with default, for better quality of reflections and refraction//
With this tweaking you may lose you crown, sorry -- Corona :)
With DOF same 1 hour for reasonable quality,even at lower values of - GA vs AA  ////
Perhaps with tweaking via Override Mat it is possible to further reduce the render time, I have not tried yet///Now it is not so necessary for me/