CoolColJ's test pics thread
hey guys, I agree, I like those 2 as well
Like the warmth of CGSphere 1_3
I've decided to scale the "sphere" up to 2metres in size, and set the aperture size to 10mm for just a hint of depth of field. It was the same as 20mm before at the smaller scale, which is not blurry enough, and not sharp enough. A bit too much in the middle, no man's land
May try 30mm to see what it looks like
rendering the full size images of each 800x800 now to get a better feel for the lighting and details
Like the warmth of CGSphere 1_3
I've decided to scale the "sphere" up to 2metres in size, and set the aperture size to 10mm for just a hint of depth of field. It was the same as 20mm before at the smaller scale, which is not blurry enough, and not sharp enough. A bit too much in the middle, no man's land
May try 30mm to see what it looks like
rendering the full size images of each 800x800 now to get a better feel for the lighting and details
which depth of field do you prefer?
BTW MLT mode is rendering this scene much faster than Bidirectional MLT!
Backs up my feeling that with outdoor scenes with the physical sky and sun, MLT is better than Bidirectional MLT for speed, as long as you don't need accurate glass rendering and lots of caustics reflected from specular surfaces or via transmission through transparent objects. But indoors with artificial lighting it's the other way around...
That will clear enough when I post up my outdoor rendering mode comparison in my rendering modes test thread
I've already finished all the renders, mostly, just need to compile em and write it up
BTW MLT mode is rendering this scene much faster than Bidirectional MLT!
Backs up my feeling that with outdoor scenes with the physical sky and sun, MLT is better than Bidirectional MLT for speed, as long as you don't need accurate glass rendering and lots of caustics reflected from specular surfaces or via transmission through transparent objects. But indoors with artificial lighting it's the other way around...
That will clear enough when I post up my outdoor rendering mode comparison in my rendering modes test thread
I've already finished all the renders, mostly, just need to compile em and write it up
- Attachments
-
- im1189042613.JPG (73.73 KiB) Viewed 3336 times
-
- im1189041774.JPG (71.42 KiB) Viewed 3336 times
That's because they're of Alien origin
actually I leaning towards the 2nd one - in the original size renderings the DOF is more obvious, and with the first one the holes in the middle, ie on the edge of the sphere are already starting to blur! It's a 2m wide sphere, so that's a bit too much DOF, considering the camera is autofocused on those black prongs
And the tile in both versions near the camera are already starting to blur. Camera is 4m away from the sphere.
Too much DOF makes the object look small though, but I do find the lines distracting if they're too sharp....
Aperture size is 0.03 and 0.015 respectively before scaling, so with 0.1 scaling that would make it 30cm and 15cm. A bit too large me thinks
Off course if the sphere were actually 20cm like it was originally then it would be 3cm and 1.5cm.
the original size renders 800x800, 3x supersampling, 40mins render time each on a P4 3.2ghz machine. Damn, would have only taken 7 mins on a quad core Q6600 cpu
I am thinking of rendering it at 960x960 and shrink it down to 800x800 for submission. But then I can't resume the current renders and it would take a bit longer...
I'll try a render with a 0.022 aperture size
actually I leaning towards the 2nd one - in the original size renderings the DOF is more obvious, and with the first one the holes in the middle, ie on the edge of the sphere are already starting to blur! It's a 2m wide sphere, so that's a bit too much DOF, considering the camera is autofocused on those black prongs
And the tile in both versions near the camera are already starting to blur. Camera is 4m away from the sphere.
Too much DOF makes the object look small though, but I do find the lines distracting if they're too sharp....
Aperture size is 0.03 and 0.015 respectively before scaling, so with 0.1 scaling that would make it 30cm and 15cm. A bit too large me thinks
Off course if the sphere were actually 20cm like it was originally then it would be 3cm and 1.5cm.
the original size renders 800x800, 3x supersampling, 40mins render time each on a P4 3.2ghz machine. Damn, would have only taken 7 mins on a quad core Q6600 cpu
I am thinking of rendering it at 960x960 and shrink it down to 800x800 for submission. But then I can't resume the current renders and it would take a bit longer...
I'll try a render with a 0.022 aperture size
thanks, but I tinkered even more..
thinking about losing the upper orbiting things, although I may not keep the 2 antenna things
The shading angles from Cinidigo just don't look right, not so great on those 2 uprights.... I imported the same object into Blendigo, and it took much longer to load the scene in Indigo and the shading was rendered totally differently, but I can't match the lighting up properly in Blendigo. Sketchup? Well the poly counts choke the program....
The bump map of the floor/tile background is also different in Blendigo vs Cinidgo...Cindigo is inverted!
If you look closely the black lines actually stick up, when you should be sinking down...
too many possibilities
thinking about losing the upper orbiting things, although I may not keep the 2 antenna things
The shading angles from Cinidigo just don't look right, not so great on those 2 uprights.... I imported the same object into Blendigo, and it took much longer to load the scene in Indigo and the shading was rendered totally differently, but I can't match the lighting up properly in Blendigo. Sketchup? Well the poly counts choke the program....
The bump map of the floor/tile background is also different in Blendigo vs Cinidgo...Cindigo is inverted!
If you look closely the black lines actually stick up, when you should be sinking down...
too many possibilities
- Attachments
-
- im1189154483.JPG (74.89 KiB) Viewed 3146 times
this one was rendered out of Blendigo, different lighting as I wasn't able to match it as I noted above, and differental materials, since the some objects were welded together on import. Plus the Object had to be scaled down by 0.004 even though it was exported out on C4D with 0.01 factor!!
I don't know why, but scene files exported out from Blendigo take much longer for Indigo to load and process vs Cinidigo and Skindigo!
see the difference in bump mapping of the tiles?
The yellow bits are using a blended material, so it looks different, just experimenting
I don't know why, but scene files exported out from Blendigo take much longer for Indigo to load and process vs Cinidigo and Skindigo!
see the difference in bump mapping of the tiles?
The yellow bits are using a blended material, so it looks different, just experimenting
- Attachments
-
- im1189160779.JPG (74.3 KiB) Viewed 3141 times
trying out some custom NK material using this program 
http://www.indigorenderer.com/joomla/fo ... sc&start=0
gonna use this material to finish off my Alien Artifacts scene!
http://www.indigorenderer.com/joomla/fo ... sc&start=0
gonna use this material to finish off my Alien Artifacts scene!
- Attachments
-
- im1189225085.JPG (192.63 KiB) Viewed 3051 times
now that I have the Green Alien material I wanted, I can finally go back and finish off my "Alien Artifacts" scene. Green Alien artifact been scaled down to the correct size, a bit over 1m as those towers are 10m tall....
been having trouble with creating realistic water, and I finally did it!!!
It's time to celebrate
All these rendered for about 20mins with MLT, exported out of Sketchup where I originally created the scene, I removed all the metal materials on the tower objects etc to speed up rendering and remove the artifacts from the heavy bump map reflections.
I spotted some shading error on those Tower things, see the square edge on the shadows?
The underwater shot looks cool - the refraction and caustics looks nice!
and that's with only 100 samples or so, more caustics to come later on if I left it rendering, plus MLT isn't that great for complex caustics. I'm trying this with Bidirectional MLT and with some SSS to see how it looks
been having trouble with creating realistic water, and I finally did it!!!
It's time to celebrate
All these rendered for about 20mins with MLT, exported out of Sketchup where I originally created the scene, I removed all the metal materials on the tower objects etc to speed up rendering and remove the artifacts from the heavy bump map reflections.
I spotted some shading error on those Tower things, see the square edge on the shadows?
The underwater shot looks cool - the refraction and caustics looks nice!
and that's with only 100 samples or so, more caustics to come later on if I left it rendering, plus MLT isn't that great for complex caustics. I'm trying this with Bidirectional MLT and with some SSS to see how it looks
- Attachments
-
- im1189320044.JPG (131.99 KiB) Viewed 3028 times
-
- im1189318796.JPG (125.73 KiB) Viewed 3028 times
-
- im1189317744.JPG (142.96 KiB) Viewed 3028 times
Whoops I made a mistake in my water model in the underwater pic....the bottom of the water doesn't reach down to the floor...
no wonder it looked a bit funny
---
same water model fault, but rendered with Bidirectional MLT and some SSS - much nicer caustics! And the Sun/Shadows actually goes through water model!
MLT just can't sample complex light paths and render them properly
no wonder it looked a bit funny
---
same water model fault, but rendered with Bidirectional MLT and some SSS - much nicer caustics! And the Sun/Shadows actually goes through water model!
MLT just can't sample complex light paths and render them properly
- Attachments
-
- im1189324342.JPG (149.37 KiB) Viewed 3005 times
Who is online
Users browsing this forum: No registered users and 86 guests
