QuoteTwo new render settings on the Advanced tab which control the region in which ray traced polygons are fully subdivided. Outside of this region the ray traced polygons are only coarsely subdivided. Previously this region was the frustum seen by the rendered image or crop region, but now you can also choose to have no detail or to have detail everywhere within the camera frustum regardless of crop settings. "Ray detail region padding" can be used to enlarge (or shrink) the region. A value of 0 means no padding. A value of 1 adds a border to each side of the frustum which is equivalent to the width or height of the image or crop region, which makes the frustum 3 times as wide and 3 times as tall.
wish to speak english better so i could understand something out of this explaination... :'(
I understand English and I barely get it. It seems like it should be easier to understand than this, but then I remember that Planetside is in England and they speak
real English...so, maybe we'll get lucky and someone will throw some slang our way.
;D
Quote from: Seth on April 03, 2009, 12:22:31 PM
QuoteTwo new render settings on the Advanced tab which control the region in which ray traced polygons are fully subdivided. Outside of this region the ray traced polygons are only coarsely subdivided. Previously this region was the frustum seen by the rendered image or crop region, but now you can also choose to have no detail or to have detail everywhere within the camera frustum regardless of crop settings. "Ray detail region padding" can be used to enlarge (or shrink) the region. A value of 0 means no padding. A value of 1 adds a border to each side of the frustum which is equivalent to the width or height of the image or crop region, which makes the frustum 3 times as wide and 3 times as tall.
wish to speak english better so i could understand something out of this explaination... :'(
Essentially this now allows you to set how detailed the calculations are outside of your crop or image area. This helps for problems in animating where shadows cast from things outside the image area were not being calculated as accurately as shadows cast by objects inside the viewing area. I believe this is the solution for Richard's problem with his animation here:
http://forums.planetside.co.uk/index.php?topic=5417.0 (http://forums.planetside.co.uk/index.php?topic=5417.0)
This is my understanding anyway.
heres quick render comparisons:
first: normal render, no crop
second: render cropped halfway, with detail in crop only (the usual way it used to work)
third: render with detail set for camera while cropped halfway like before.
Good example Neuspadrin.
I believe this function also allows for more accurate reflection of water if things outside the frustum are being reflected in that water.
There's no difference between the first and the third render (above), is there? If not, why use it?
Quote from: calico on April 03, 2009, 01:58:46 PM
There's no difference between the first and the third render (above), is there? If not, why use it?
thats the whole point. the third render was cropped, the first was not.
so people who want a nice huge render but their computer cant handle it, or they want to slowly build it and have stopping points, they can crop the image into areas while using the detail from camera instead of crop such that the shadows match up.
the left side is one crop, the right side is the other crop, look at the bottom line that goes red to green to see when the crop happens. if you notice in the second one, following that line is where shadows distort and act funny and doesn't add up together as nice + shadows on the right look funny and non realistic.
Thanks for explaining, but the second one makes sense. I still don't get the use of the first vs. third.
the first one is the image as rendered without a crop just to show what it for sure should look like and get a comparison. the second was to show how the previous versions were forced to always render (where the shadows wouldn't match up if you cropped), and the third is to show how now with the feature enabled, it looks just like the original again doing the same crops.
it was just a couple renders to show some comparison so people could see what the feature did.
Thanks again.
So...(I'm tired, but not really so dense), if someone uses crop - WHY would they use crop? That is my real question. Does using the crop setting somehow cut render times?
well its a matter of then they can crop it and do it in groups. like dom. He only had access to a computer for rendering for a short while, and tried cropping his large image to then piece together, but he couldn't due to the shadows n such not matching across.
Ahhh. I see. I thought for some reason this was being used for animation in some way.
Thanks for explaining, neuspadrin.
I had this issue once where I was rendering, and wanted to check a setting in the main window. I then got a runtime error and managed to make a screen shot before the program shut down. I could finish the render by doing crops so I didn't have to render the whole thing again.
Another time I had a lot of large populations, and I didn't have enough RAM for the whole image, bit I did manage to render in 3 parts, cropped.
So that's why someone would use Crop...
I believe it also helps animation to stop shadow popping using padding and such i believe.
This makes sense.
Quote from: neuspadrin on April 03, 2009, 03:32:59 PM
I believe it also helps animation to stop shadow popping using padding and such i believe.
I am just testing it now but I believe that setting detail to camera is also intended to fix the problem where the global illumination didn't match if a small crop render was made to repair a displacement cut off or to allow an object to me moved slightly without re-rendering the whole scene. Often the colours of the cropped region could be significantly different.
Quote from: Mr_Lamppost on April 03, 2009, 06:14:35 PM
I am just testing it now but I believe that setting detail to camera is also intended to fix the problem where the global illumination didn't match if a small crop render was made to repair a displacement cut off or to allow an object to me moved slightly without re-rendering the whole scene. Often the colours of the cropped region could be significantly different.
That's correct, this new function should aid in this.
This is going to be really useful. I was making a VR earlier of a few planets and when I stitched the images, the front panel and bottom panel had different shadows, resulting in an obvious light-line between the two images. Hope this fixes that problem.
I tried to render two areas again, however I"m still getting a problem. The attached picture showed the obvious stitch. I turned this ray detail region setting to Camera. What else do I need to do?
How did you configure the setting?
Quote from: Tangled-Universe on April 05, 2009, 04:44:02 PM
How did you configure the setting?
Under the Render node I went to Advanced > Ray detail region > Detail in Camera. I assume there's something else I need to do...
And you increased the value to >0 I suppose? Hope you don't mind I ask this stupid question :)
By the way, I think you should use detail in crop region, since you're rendering crops. Then increase the padding to 1 and you should be fine.
Also, even with this function seams can still occur.
Alright I"ll give that a go. I'm actually making a QTVR so 6 full renders, just changing the camera angle. I haven't had any seem problems yet... just with this scene. Thanks though.
@Aagam - I saw something similar with the exact same stitch in the exact same place. May be a coincidence. Please let us know what you find out. When I followed what TU mentioned here about the setting of Detail in Crop Region and a little padding, I didn't see that again.
That's good to hear Calico, because I wasn't entirely sure either.
This setting may not fix all problems, but hopefully it will help in many of these kinds of situations.
- Oshyan
Still having troubles. I don't know if I"m doing something wrong, but when I stitched the VR together, this is what I got:
http://imagearts.ryerson.ca/agam/QTVR/testvr.mov
As you can see, the stitches on the planet are mostly visible where the front cube fuses with the bottom, and the front side to the right. I don't know why the lighting changes so drastically. The only setting I touched in these renders was the camera and that's it :(
Quote from: Aagam on April 09, 2009, 05:26:57 PM
Still having troubles. I don't know if I"m doing something wrong, but when I stitched the VR together, this is what I got:
http://imagearts.ryerson.ca/agam/QTVR/testvr.mov
As you can see, the stitches on the planet are mostly visible where the front cube fuses with the bottom, and the front side to the right. I don't know why the lighting changes so drastically. The only setting I touched in these renders was the camera and that's it :(
The camera-change could well be THE reason why there are differences, since this can influence the GI-calculations quite some. You make a new POV and the lighting differs from the previous POV. GI will then be calculated accordingly and this could result in differences. Even with ray detail padding this may not be solved. Unless you can increase and use that value beyond 1 to make it cover 2 frustums.
Yea that makes sense. It's just weird cause I've made other VRs with no trouble... I only get this problem if the camera is past the atmosphere of the planet. By the way, awesome gallery T-U.