I agree assuming you go for compositing, my need is simplier but as a paradox, it requires more controls. I'd like to avoid the usage of bilboards to insert a background image like landscape as it cast shadows and receive light and sometimes it's a real problem.FrankPooleFloating wrote:I am (personally) not so sure bro -- the way I work, I would never tweak the image to
my scenes - but always tweak the scene to the image... that I am rendering for. This feature
was begged for by me and some other blokes who do a ton of compositing... this really is
all about compositing.
And by not having controls, this actually helps ensure that you will never accidentally render
your stuff incorrectly... to me, it just seems that with the current bg image, this is essentially
a non-brainer. I love it!
OctaneRender™ Standalone 1.12
Forum rules
NOTE: The software in this forum is not %100 reliable, they are development builds and are meant for testing by experienced octane users. If you are a new octane user, we recommend to use the current stable release from the 'Commercial Product News & Releases' forum.
NOTE: The software in this forum is not %100 reliable, they are development builds and are meant for testing by experienced octane users. If you are a new octane user, we recommend to use the current stable release from the 'Commercial Product News & Releases' forum.
- enricocerica
- Posts: 1012
- Joined: Wed Nov 25, 2009 7:32 pm
- Contact:
Modeling system : I7 32GB Windows 10 & Fujitsu Celsius H720
GPU : 1x Gigabyte GTX580 3GB + 1x MSI GTX780 3GB + 1x PALIT GTX780 6GB +1x Asus Stix GTX1070 8GB
http://www.myline.be
GPU : 1x Gigabyte GTX580 3GB + 1x MSI GTX780 3GB + 1x PALIT GTX780 6GB +1x Asus Stix GTX1070 8GB
http://www.myline.be
As a side note for the node graph, I think it looks less visually appealing the way the nodes look now.
It looks like you are color coding different types of nodes, which I think is awesome, but visually, it could be more polished (like the rest of the UI is).
It looks like you are color coding different types of nodes, which I think is awesome, but visually, it could be more polished (like the rest of the UI is).
Intel quad core i5 @ 4.0 ghz | 8 gigs of Ram | Geforce GTX 470 - 1.25 gigs of Ram
- FrankPooleFloating
- Posts: 1669
- Joined: Thu Nov 29, 2012 3:48 pm
Well, hopefully (gets on knees and prays to the saint of cg) primary visibility
is on deck... and your shadow & reflection woes (when unwanted) will be nothing
more than a bad memory.
If all goes well, we should soon have the ability to make things invisible to camera
and/or rays - including emission objects...
is on deck... and your shadow & reflection woes (when unwanted) will be nothing
more than a bad memory.
If all goes well, we should soon have the ability to make things invisible to camera
and/or rays - including emission objects...
Win10Pro || GA-X99-SOC-Champion || i7 5820k w/ H60 || 32GB DDR4 || 3x EVGA RTX 2070 Super Hybrid || EVGA Supernova G2 1300W || Tt Core X9 || LightWave Plug (v4 for old gigs) || Blender E-Cycles
enricocerica wrote:I agree assuming you go for compositing, my need is simplier but as a paradox, it requires more controls. I'd like to avoid the usage of bilboards to insert a background image like landscape as it cast shadows and receive light and sometimes it's a real problem.FrankPooleFloating wrote:I am (personally) not so sure bro -- the way I work, I would never tweak the image to
my scenes - but always tweak the scene to the image... that I am rendering for. This feature
was begged for by me and some other blokes who do a ton of compositing... this really is
all about compositing.
And by not having controls, this actually helps ensure that you will never accidentally render
your stuff incorrectly... to me, it just seems that with the current bg image, this is essentially
a non-brainer. I love it!
I agree with Enricocerica, not everyone has an exact image they need to composite onto, some people are creating a scene and choose an image they think will work well. It doesn't need to match with a video, for example, they are just doing what looks good. So, if they have a high res image, they may want to zoom in and just composite over a certain area, and doing that on the fly instead of modifying and praying in Photoshop back and forth would be much better.
I fully understand what you mean though, if you are compositing onto something that absolutely cannot change, a lock button would be needed.
Intel quad core i5 @ 4.0 ghz | 8 gigs of Ram | Geforce GTX 470 - 1.25 gigs of Ram
- FrankPooleFloating
- Posts: 1669
- Joined: Thu Nov 29, 2012 3:48 pm
Fair enough.. let's do it!!.. everyone wins! 

Win10Pro || GA-X99-SOC-Champion || i7 5820k w/ H60 || 32GB DDR4 || 3x EVGA RTX 2070 Super Hybrid || EVGA Supernova G2 1300W || Tt Core X9 || LightWave Plug (v4 for old gigs) || Blender E-Cycles
Enricocerica :
I agree with FrankPooleLoading :
For good integration of a 3D object in the real world, it is very important to use a prepared image of exactly the resolution, or pixel ratio of the render.
If this image is a cut out of a larger image, it has to be perfectly centered on the optical axis of camera used for the shot, otherwise, the "manual camera tracking" will be very difficult or not accurate.
On the first image, I cut the image from a bigger one, without much caution. Not centered on the optical center. The matching has been very difficult. The scale is almost respected, but I haven't been able to match the angle perfectly due to a vertical shift of the optical center.
The second has the optical center respected, but I haven't been able to match correctly the focal length, I think, or find the exact height of the camera.
In the examples below, I made measures with a decameter, on the parking area, and built the red template/guide object in Blender according to the measured values. Despite that, the matching is still not perfect. Matching the background perfectly requires finding the very same camera parameters (Height, distance, tilt, focal length, horizontal position, and an image respecting the original optical axis.
Difficult to get all these values working correctly from a cut out in a random photo, imho...
The most difficult value to adjust without measures and marks taken when shooting the photo is the real scale of the object.
I use a Bridge digital camera, and it doesn't display the focal length. I only know that I am at 28mm (equivalent to 35mm film) at full angle. Not very useful data indeed, so respect of the optical axis of the original shot and precise measurements on the location are very important.
I forgot also to say one thing : the parking area has a slight slope for drainage of rain water ! I used an horizontal plane instead, thinking that it would be negligible... Error ! Nothing is negligible when it is a matter of matchmoving !
FrankPooleLoading :Would be great to have some controls over the background image, like zoom, pan, x&y scale, ... to be able to fine adjust it to the scene
@Kavorka & Enricocerica :And by not having controls, this actually helps ensure that you will never accidentally render
your stuff incorrectly... to me, it just seems that with the current bg image, this is essentially
a non-brainer. I love it!
I agree with FrankPooleLoading :
For good integration of a 3D object in the real world, it is very important to use a prepared image of exactly the resolution, or pixel ratio of the render.
If this image is a cut out of a larger image, it has to be perfectly centered on the optical axis of camera used for the shot, otherwise, the "manual camera tracking" will be very difficult or not accurate.
On the first image, I cut the image from a bigger one, without much caution. Not centered on the optical center. The matching has been very difficult. The scale is almost respected, but I haven't been able to match the angle perfectly due to a vertical shift of the optical center.
The second has the optical center respected, but I haven't been able to match correctly the focal length, I think, or find the exact height of the camera.
In the examples below, I made measures with a decameter, on the parking area, and built the red template/guide object in Blender according to the measured values. Despite that, the matching is still not perfect. Matching the background perfectly requires finding the very same camera parameters (Height, distance, tilt, focal length, horizontal position, and an image respecting the original optical axis.
Difficult to get all these values working correctly from a cut out in a random photo, imho...
The most difficult value to adjust without measures and marks taken when shooting the photo is the real scale of the object.
I use a Bridge digital camera, and it doesn't display the focal length. I only know that I am at 28mm (equivalent to 35mm film) at full angle. Not very useful data indeed, so respect of the optical axis of the original shot and precise measurements on the location are very important.
I forgot also to say one thing : the parking area has a slight slope for drainage of rain water ! I used an horizontal plane instead, thinking that it would be negligible... Error ! Nothing is negligible when it is a matter of matchmoving !

Last edited by ROUBAL on Fri Apr 05, 2013 6:31 pm, edited 1 time in total.
French Blender user - CPU : intel Quad QX9650 at 3GHz - 8GB of RAM - Windows 7 Pro 64 bits. Display GPU : GeForce GTX 480 (2 Samsung 2443BW-1920x1600 monitors). External GPUs : two EVGA GTX 580 3GB in a Cubix GPU-Xpander Pro 2. NVidia Driver : 368.22.
Great release!
found a bug; probably one of my most used features; in texture environment (havent tried it elsewhere) when hitting the refresh button, it does nothing, you have to click open and reselect the file. I use this very frequently with HDR light studio (which would be amazing to see integrated) and have to refresh often to check the results.
otherwise, love where this is going; also looked at the Brigade post, and looked at the vids out for it; Simply put, WOW. I hope to see this as an engine option in octane, that would replace my quick maya MR renders for basic projects.
Thanks!
found a bug; probably one of my most used features; in texture environment (havent tried it elsewhere) when hitting the refresh button, it does nothing, you have to click open and reselect the file. I use this very frequently with HDR light studio (which would be amazing to see integrated) and have to refresh often to check the results.
otherwise, love where this is going; also looked at the Brigade post, and looked at the vids out for it; Simply put, WOW. I hope to see this as an engine option in octane, that would replace my quick maya MR renders for basic projects.
Thanks!
Windows 8.1 | i7 950 | GTX 780ti | 24gb ddr3
This version is a bit sensible to GPU overclocking... one of my 580GTX get failed (always until I set it to default) no priority.
I noticed also that when you pause the render, change kernel, then resume, the render continue with previous kernel, not like 1.11.
but this version is faster, but not enough to compensate the overclocking of my 580GTX.
I noticed also that when you pause the render, change kernel, then resume, the render continue with previous kernel, not like 1.11.
but this version is faster, but not enough to compensate the overclocking of my 580GTX.

I7-3930K 64Go RAM Win8.1pro , main 3 titans + 780Ti
Xeon 2696V3 64Go RAM Win8.1/win10/win7, 2x 1080Ti + 3x 980Ti + 2x Titan Black
Xeon 2696V3 64Go RAM Win8.1/win10/win7, 2x 1080Ti + 3x 980Ti + 2x Titan Black