View Full Version : depthmap shadows
08-30-2004, 08:37 AM
Can someone enlighten me here, please. I need to use depthmap shadows (raytrace shadows is not an option) but as long as the spotlight that casts those shadows is in close proximity to the object, everything is fine and the shadows are nice and clean. However, as soon as I move the spotlight to illuminate a larger area, the shadows start to look horribly pixelated. I understand it has to do with dmap resolution, but the problem remains unless I change it to something ridicuulous like 6000 or something at which point Maya crashes.
Any help would be appreciated.
08-30-2004, 09:37 AM
There's the option to solve this with the use of shadowlights (Saw this in a gnomon video), eg. lights which don't light to object but only give it a shadow. So you need one shadowlight per object, all in the same angle to get the correct results. Positive side is that you can position all these shadowlights relatively close to their objects, so your dmap resolution can be quite low to achieve the proper results.
08-30-2004, 10:08 AM
Do you need this light to generate shadow for the whole scene?. Or you can live with only the main focus to have shadow?.
Depth map based shadow relies on a depth map, if you pull your light far away then the bigger area of the scene needs to be included in the map, sometimes this will lead to a tiny (maybe only several pixel) areas where the main object shadows you need are mapped. So when the depth map applied back to the scene it will be pixelated so you will see shadow artifact when it is rendered.
If you only interested on the shadow of the main object, you could turn-off Use Dmap Auto Focus, and set your Dmap Focus manually. I usually turn-on Disk Based Dmaps temporarily, to Overwrite Existing Dmaps, and do a quick render (lowest quality, lowest resolution), and check the resulting Dmaps in RenderData\depth subdirectory (use fcheck and press Z to show depth channel). With default Dmap Focus (90 deg), you may find the Dmap consist of mainly the grounds and the main object shadow is only a very small part of it.
In this case setting a Dmap Focus down to 20 deg, will reveal the main object shadow, but you will lost other part shadow (outside the Dmap) which maybe acceptable. So the idea is to balance the focus to have main object shadow reasonable sufficient while maintaining certain area to still get shadows.
Increasing the resolution to 6000 x 6000, but only 10 x 10 pixel is used is a bit redundant.
This is also applicable to directional light, but the Dmap focus should be read as orthographic camera width.
01-19-2006, 02:00 AM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.
vBulletin v3.0.5, Copyright ©2000-2014, Jelsoft Enterprises Ltd.