Hi volume rendering experts, I was playing around with a density volume render of a tiny cube on my laptop, everything went smoothly, however, I notice a change in "overall" brightness between images when I did a rotation. I'm not sure how to explain this, but it seems that from images of different angles, sometimes the average brightness is consistent, but sometimes it suddenly drops and the image would look about half as bright, and after a couple frames it would light back up. I'm guessing this is after each image is ray casted, it gets normalized to a certain value, and that value changes depending on the angle? Is there a way to pin the brightness between images so if I make a movie it won't dim and light up as I rotate? Also I was wondering if I'm the only person to notice this? (the only other post I found about volume render brightness is about AMR cell contributions) From G.S.
Hi Geoffrey, The brightness of the total image is basically normalized to the brightest pixel. Occasionally, when you rotate the domain, a single pixel will be incredibly bright which will cause the rest of the image to be dim by comparison. You can fix this in with the snapshot call by adding the keyword, clip_ratio=some_float. This limits the maximum brightness to no more than a fixed number of standard deviations from the mean. I recommend you start with a clip ratio of 5 or 6 and play around with it. Briton On Mon, Feb 27, 2012 at 7:57 PM, Geoffrey So <gsiisg@gmail.com> wrote:
Hi volume rendering experts,
I was playing around with a density volume render of a tiny cube on my laptop, everything went smoothly, however, I notice a change in "overall" brightness between images when I did a rotation. I'm not sure how to explain this, but it seems that from images of different angles, sometimes the average brightness is consistent, but sometimes it suddenly drops and the image would look about half as bright, and after a couple frames it would light back up.
I'm guessing this is after each image is ray casted, it gets normalized to a certain value, and that value changes depending on the angle? Is there a way to pin the brightness between images so if I make a movie it won't dim and light up as I rotate?
Also I was wondering if I'm the only person to notice this? (the only other post I found about volume render brightness is about AMR cell contributions)
From G.S.
_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
Hi Geoffrey, Another option, which is outlined in the Advanced Rendering talk from the workshop, is to use an enhance script to renormalize the image. This doesn't require setting the clip variable, and is done as a post-processing step. The video of the talk, as well as slides, can be found here (see Advanced Rendering): http://yt-project.org/workshop2012/#presentations It references an enhance.py script, which can be found through the hub, here: http://hub.yt-project.org/AnalysisAndVisualization/enhance-your-renderings/ I'd suggest taking a look, as it will also walk you through other advanced rendering techniques. Best, Sam On Tue, Feb 28, 2012 at 5:50 AM, Britton Smith <brittonsmith@gmail.com>wrote:
Hi Geoffrey,
The brightness of the total image is basically normalized to the brightest pixel. Occasionally, when you rotate the domain, a single pixel will be incredibly bright which will cause the rest of the image to be dim by comparison. You can fix this in with the snapshot call by adding the keyword, clip_ratio=some_float. This limits the maximum brightness to no more than a fixed number of standard deviations from the mean. I recommend you start with a clip ratio of 5 or 6 and play around with it.
Briton
On Mon, Feb 27, 2012 at 7:57 PM, Geoffrey So <gsiisg@gmail.com> wrote:
Hi volume rendering experts,
I was playing around with a density volume render of a tiny cube on my laptop, everything went smoothly, however, I notice a change in "overall" brightness between images when I did a rotation. I'm not sure how to explain this, but it seems that from images of different angles, sometimes the average brightness is consistent, but sometimes it suddenly drops and the image would look about half as bright, and after a couple frames it would light back up.
I'm guessing this is after each image is ray casted, it gets normalized to a certain value, and that value changes depending on the angle? Is there a way to pin the brightness between images so if I make a movie it won't dim and light up as I rotate?
Also I was wondering if I'm the only person to notice this? (the only other post I found about volume render brightness is about AMR cell contributions)
From G.S.
_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
Thanks Britton and Sam, I've went through the videos on volume rendering and tried what Cameron showed with clip_ratio, it lessened the changing brightness from angle to angle somewhat, but it's Sam's enhance.py that does the trick, at least to my eyes. I would highly recommend sticking the links of these videos on the documentation or example page. I know they say a picture is worth a thousand words, so videos such as these are worth say 30 frames per second * length of videos = ... Loved the videos From G.S. On Tue, Feb 28, 2012 at 6:39 AM, Sam Skillman <samskillman@gmail.com> wrote:
Hi Geoffrey,
Another option, which is outlined in the Advanced Rendering talk from the workshop, is to use an enhance script to renormalize the image. This doesn't require setting the clip variable, and is done as a post-processing step. The video of the talk, as well as slides, can be found here (see Advanced Rendering): http://yt-project.org/workshop2012/#presentations
It references an enhance.py script, which can be found through the hub, here: http://hub.yt-project.org/AnalysisAndVisualization/enhance-your-renderings/
I'd suggest taking a look, as it will also walk you through other advanced rendering techniques.
Best, Sam
On Tue, Feb 28, 2012 at 5:50 AM, Britton Smith <brittonsmith@gmail.com>wrote:
Hi Geoffrey,
The brightness of the total image is basically normalized to the brightest pixel. Occasionally, when you rotate the domain, a single pixel will be incredibly bright which will cause the rest of the image to be dim by comparison. You can fix this in with the snapshot call by adding the keyword, clip_ratio=some_float. This limits the maximum brightness to no more than a fixed number of standard deviations from the mean. I recommend you start with a clip ratio of 5 or 6 and play around with it.
Briton
On Mon, Feb 27, 2012 at 7:57 PM, Geoffrey So <gsiisg@gmail.com> wrote:
Hi volume rendering experts,
I was playing around with a density volume render of a tiny cube on my laptop, everything went smoothly, however, I notice a change in "overall" brightness between images when I did a rotation. I'm not sure how to explain this, but it seems that from images of different angles, sometimes the average brightness is consistent, but sometimes it suddenly drops and the image would look about half as bright, and after a couple frames it would light back up.
I'm guessing this is after each image is ray casted, it gets normalized to a certain value, and that value changes depending on the angle? Is there a way to pin the brightness between images so if I make a movie it won't dim and light up as I rotate?
Also I was wondering if I'm the only person to notice this? (the only other post I found about volume render brightness is about AMR cell contributions)
From G.S.
_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
participants (3)
-
Britton Smith -
Geoffrey So -
Sam Skillman