Viewers: in countries Watching now:
Create highly realistic 3D architectural drawings with V-Ray, a popular third-party renderer for SketchUp. This course shows how to take a single scene with interior/exterior elements and add lights, move cameras, and enhance objects with translucent and reflective surfaces. Author Brian Bradley explains concepts like irradiance mapping, perspective correction, and fixed rate sampling, while showing how to leverage each of the V-Ray tools and its material and lighting types to achieve specific effects.
Because photon mapping is generally regarded by V-Ray users to be the weakest of V-Ray's GI Lighting tools, and so is hardly ever utilized in that role, we're going to skip over using photons as a GI tool and look instead at the last of V-Ray's GI options available to us, this being the Deterministic Monte Carlo, or DMC Engine, as we will refer to it from this point on. As we've done a couple of times in this chapter, we need to go and enable V-Ray's GI systems to have a look at the DMC engine, so let's come into our Indirect Illumination rollout, let's put a check in the On box and then let's scroll down and set Deterministic Monte-Carlo as our Primary Engine, and of course, we want to disable our Secondary Engine by setting it to None.
Then, scrolling down just a little way, you can see we've our Deterministic Monte Carlo GI rollout. In here, we have really only one parameter with which to work in order to produce our Primary light bounce solution, this being our Subdivs value. The Bounces parameter you see here is only ever utilized when DMC is set as a Secondary Engine, so we can just ignore that for now. At this point it probably is worth noting that very few experienced V-Ray users would generally choose to use the DMC engine for an interior GI solution, unless, that is, there was a very specific need for it.
Although excellent for exteriors, it is by far and away the slowest and hardest of V-Ray's GI engines to clean up on interior renders, and so, as we say, would only be used when there was a very specific need. Now that could be if we have a scene with lots of geometric detail. The DMC Engine is extremely good at pulling that out in our renders. As DMC is a lot slower than both the Irradiance Map and the Light Cache engines in V-Ray, whenever we make parameter changes in this video, we're going to utilize some pre-rendered images that will just demonstrate the changes that would occur with those parameter alterations.
So, as our current settings are the defaults inside of the DMC Engine, let's switch over to Adobe Photoshop and have a look at the render that these settings would give us. As you can see, what we get is not very pretty-looking at all. We've lots of extreme noise in the scene, and in our recesses we have some really unnatural darkness. Now these dark areas exist because, like irradiance mapping, the DMC Engine, when set in the Primary slot, only gives us a single bounce of light, so we have no secondary rays being traced into these recessed areas.
With regard to the noise found in our render, we could improve this in a couple of ways. We could either work with our DMC Subdivs value or because with the DMC engine we are essentially looking to just clean up noise in the scene, we could just as easily use V-Ray's Image Sampling control to deal with this problem. First though, let's jump back into SketchUp and see if we can deal with these unnaturally dark areas. Now we could of course increase our initial Subdivisions value.
Because we would have more single or initial light bounces in the scene, there would be a chance that we would actually trace more initial light paths into these dark recessed areas. To be honest though, in order to make any significant difference, our Subdivs value and our render times would be much, much higher than is practical or even necessary. Here again you see we run into the absolute need for a secondary engine in V-Ray, if, that is, we want to get a timely high-quality solution out of the V-Ray's GI systems.
So let's do just that. Let's enable a secondary engine. The question is, which one would we choose? Well, one option would be to use Deterministic Monte Carlo for both the Primary and Secondary Engines. Of course, don't forget once we initialize the Secondary engine, if we want physically correct behavior from it, we just need to make certain that that Secondary Multiplier is set to a value of 1. Now of course, our Bounces parameter is coming into play, and we're basically saying that we want three extra bounces of light in our scene.
So let's have a check at how that would change things inside of Photoshop. So we would go from a single bounce of light with DMC as the Primary Engine to extra bounces of light using the DMC Engine in both the Primary and Secondary slots. Now as you can clearly see, we've made quite a significant difference to the light distribution in our scene. We're certainly seeing some light trace into those unnaturally darkened areas. In fact, as we compare renders, you can see that particularly around the curtain interaction with the wall over here, and in this darkened recess to the right of the image, we are getting what almost looks like naturalistic light behavior.
Of course, we still have some problems at the far end of the room, and of course we clearly still have an excessive amount of noise in our GI solution. Well, we have mentioned a couple of times already in this chapter that when it comes to interior rendering, the preferred choice for Secondary Engine would typically be the Light Cache system. So let's jump over into SketchUp and see what enabling that particular option as a Secondary engine will do for us. So once again, we need to come to our Engine Type dropdown, and this time we will set Light Cache. And we'll just accept the defaults and go and check our Photoshop render.
So from our DMC and DMC Render to our DMC and the Light Cache solution, and you can see there is quite a significant difference in how the light is now bouncing around this environment. In fact, because we're getting a better tracing of light, we're also cleaning up the noise in a much more timely fashion. The only drawback to using the Light Cache system for the Secondary engine, if we just keep an eye on some of these contact areas, you can see that with the DMC and DMC engine, we get very nice contact shadows, occlusion shadows, but as soon as we add the Light Cache engine in, we start to lose a little bit of that quality and definition.
But again, as we can add that contact detail back in by enabling Ambient Occlusion inside the system, we're probably going to stick with our Light Cache solution here, as we're clearly getting a much more naturalistic bouncing of light in this environment. Of course, we still need to deal with our noise problem, so let's go back into SketchUp and see what we can do about that. Probably our first port of call would be the DMC Subdivs value, so let's go and set that up to a value of 24. So, we've considerably increased what we're getting in terms of samples in the scene.
And again, let's go and see how that would alter our current render. As we make the switch here, remember we are examining the noise in the scene, so keep an eye on some of the areas where the noise is quite prominent, particularly in the foreground on the floor; obviously, that's quite noticeable when it comes to this particular shot that we've set up. So we would go from this to this, which as you can see, is quite a considerable clean up. Now of course, we're well aware that increased render quality is usually going to cost us in terms of increased render times, so let's just go and take a take a look at the times we've been getting from our renders up until this point.
So let's go all the way back to our initial image. So I'm just going to hold the spacebar and then left-mouse-click and down in the left- hand corner, you can see we have our timestamp. And our original render was 3 minutes and 13 seconds. Our second render with DMC and DMC enabled came in at 5 minutes and 16 seconds, so we've definitely increased quite a bit there. When we switch over to Light Cache, we dropped all the way by down to 3 minutes and 36 seconds. So not only did we improve the quality of our scene, we actually improved the render times.
Now when we add our extra Subdivs in to really clean the scene up, we go back up to just short of 6 minutes, which in terms of the quality that we are getting and given the fact that this is a 1280 x 720 render, is not bad at all. Now of course, we did say that there was another way that we could tackle the noise problem in our scene, using V- Ray's Image Sampler controls. So again, let's jump back into SketchUp and see how we would set that up for ourselves. The first thing I want to do is make certain we reset our Subdivs inside of the DMC rollout.
In fact, I'm going to drop this all the way down to a value of 1. Then of course, we need to locate V-Ray's Image Sampler control. So up near the top, we have our Image Sampler rollout. I'm going to leave Adaptive DMC set as the Engine Type. We've got our Minimum Subdivisions set at 1, which we will leave there, but I'm going to increase the Max Subdivs up to a value of 12. I'm also going to lower this Color Threshold so more of those Maximum Subdivs, more of those samples, can actually be used. So a value of .005 should work very nicely for us in this instance.
And again, back over into Photoshop and we'll see what that has done for us. So now we go from our increased Subdivs render to our Image Sampling. And as you can see, there isn't a huge amount of difference between the two renders; we have still reasonably cleaned up the noise in our scene. If we go back to our initial state, we can see we've lots of noise on the beams here, and when we go our Image Sampling, you see that that has cleaned up reasonably well. Of course, our Subdivs render is much cleaner, so you may think that that clearly is the way to go, and in some instances that would be absolutely true.
Do remember, however, that our image sampling approach would benefit every aspect of the scene. We're not just cleaning up noise inside of the GI system; those increased samples would be used for everything, including materials and geometry edges, so we would get a much cleaner look, although probably with scene materials, our render times would increase by quite a bit. In fact, we haven't looked at our render times for this particular render, so let's just check that, and you can see we're all the way up to just a little bit over 8 minutes with this particular render, so quite an increase in render times.
Now, whilst we've seen that the DMC Engine is probably not going to be our first choice for a GI solution when we're working with interiors, we have hopefully seen just how easy it is to work with, especially when you consider its simplified control set. And don't forget that this particular GI engine is extremely good at pulling out scene detail. We noticed that with our occlusion shadows. Ultimately of course, the choice of which of the three GI engines we have examined you choose to use will depend entirely upon the type of scene you're working with and to some extent, upon your own artistic expectations.
If it is that we want lots of light bouncing around inside of an interior and we still want a reasonably high degree of physical accuracy to our light bounce, well maybe Irradiance Mapping and Light Cache would be the way to go. If however we had lots of detail in the scene, then maybe DMC and Light Cache would be the way to go. Certainly, if we have scenes that have a high memory requirement, we might even want to use DMC as both the Primary and Secondary Engines. You see, DMC only keeps the buckets currently being rendered in memory.
That course can actually mean the difference between our scene rendering or crashing SketchUp. In the end, whichever Global Illumination setup gets us the shot that we want in the time that we have available for it, well, that is the right setup for us.
There are currently no FAQs about SketchUp Rendering Using V-Ray.
Access exercise files from a button right under the course name.
Search within course videos and transcripts, and jump right to the results.
Remove icons showing you already watched videos if you want to start over.
Make the video wide, narrow, full-screen, or pop the player out of the page into its own window.