Hi everyone, here I am to compare two incredible render engines like VRayforC4D and Maxwell Render on an interior scene created in Cinema4D. I often work with both of them but obviously not on the same project. I decided to model this scene and recreate a typical Ikea set to compare render speed, image cleaning, settings and materials used for the same set, lights and camera settings. This test doesn’t want to figure out which is the best render engine simply because it would not be possible, each of them has its own characteristics and they are two excellent products. It would be like comparing Canon and Nikon! In this test I used 1.9 versions of VRayforC4D and 3.2.5 version of Maxwell Render, both of CPU. The images are rendered on clients with two quad Xeon 2.13 Ghz.

The first thing to consider is that we are comparing two different concept of render engines. VrayforC4D is a raytrace engine and Maxwell Render is unbiased. This leads to the first consideration: Maxwell Render has fewer parameters to be set and the light result is physically correct; this is a great advantage if you don’t want to waste time to find the right compromise between image cleaning and render times. In VRayforC4D the time to calculate the the final render is less then using Maxwell Render. It will be up to our experience to evaluate whether to invest time in searching for the optimal parameters render in Vrayforc4d or avoid it and decide to calculate the final render in Maxwell Render. Without the aid of a render farm VRayforC4D is probably the best solution, especially for high resolution images. All the renders have been calculated at 3500x2188px and on all the images you will find the render times of the two engines. The great advantage of Maxwell Render, though, is that you can stop the render to check that everything is ok and then restart it without losing what it has already renderized. This has big advantages, for example we can send the preview to the customer in a short time, or start with the post production on an image not complete to correct levels, backgrounds etc ... and then replace the final image once the render is complete. If we are working on an animation we can begin to edit the video with draft quality frames and replace them at the end of the rendering.... not bad !!! Maxwell Render also uses the Fire for a quick preview of the scene that helps in setting the light. The new version of VRayforC4D 3.4 uses in addition to the GPU some settings for incredible fast previews.

I used a preset of  Stefan (Stefan precise 04A_Universal Preset [IR-LC]) for VRayforC4D render, while for Maxwell Render, as mentioned above, are not necessary particular parameters. What I have noticed comparing the renders is the appearance of some of spots of light in some areas in  the images produced by VRayforC4D (for example on the lamp or on the window areas) and some stains in the most inaccessible areas of the rendering (for example under the bed or on that areas distant from sources of light and very shady). They are little things that only an expert eye grasps, but we are here for that :-). The images produced by Maxwell Render are very clean, the shadows are particularly defined in the areas of the duvet and generally across the image and  there aren’t problems with stains or blemishes. Rather whether the value of SL of the rendering is too low the picture will present a higher granularity. In this test I ‘ve get clean images with a SL of 17 even if a SL of 16 would have been acceptable. You will notice that only in images with few lights on the grain increases, but personally  I think this is great because it remembers the films with high ISO values. In extreme cases, for example with translucent materials, you can set a SL value that is general for the scene, and one that is specific for the most complex materials. Done this the software, once reached the overall value, will continue to calculate only the materials or objects with an upper SL, with a considerable saving of time. For example, if there are windows or curtains in the scene we could only calculate these objects with greater accuracy. I haven’t used this method in these tests.

Another aspect to consider is the influence of the VRayforC4D Color Mapping in the rendering. The scene was illuminated with area lights on the windows, without sky or sunlight. The color mapping is definitely a big plus, a very useful tool in situations of strong contrasts of light. Basically it balances black and white areas of the image. This is a great help, but often the result is that the image is "fogged", with no contrasts and flat color. Especially the whites tend to turn gray and warm colors lose vividness. This involves a massive work in postproduction. I’ve also deleted some wood knots in Photoshop because they were very strong and red. In these pictures I used the Exponential Color Mapping. To avoid this you can work in Linear Multiply but with considerable difficulty in the management of lights, as I said before. In Maxwell Render this parameter doesn’t exist, which makes the management of lights harder, more like that of the photographer (you work as in Linear Multiply the VRayforC4D), but the color rendering and the correspondence with the texture is perfect. Furthermore Maxwell Render has the Multilight that allows us to go in post production to correct any overexposure or burn, but I’ll talk about this later. In the images we can see the difference among renders without post production.

The final output of the images is 16-bit tiff for VRayforC4D and MXI for Maxwell Render. I’ve also activated the alpha channel for both and the material ID channel for Maxwell Render. The possibility of exporting different image channels to facilitate post-production. Both software have this possibility, but as I used the exponential color mapping in VRayforC4D the multipass image would have been lightened. Furthermore, the management of the channel selection in VRayforC4D is not immediate in contrast to Maxwell Render where everything is very simple: check the desired channels from a list, choose the type of file to save and you have done. Usually, if I need some alpha maps of some object for post-production in VRayforC4D, I make a scene without lights and materials so the rendering is fast, then I add the compositing tag to the object I have to cut. The result are a completely black render and an alpha map of my object. It’s a very convenient method.

All the shots have been created using the same focal length, the same aperture and the same shutter speed. We note how, especially for longer focal lengths, the interpretations of the standard cameras of both software are different, both as angle of view and depth of field. Honestly, I would not have expected this:-) The VRayforC4D images are much more blurred and the subject is closer, as if the focal length was longer (I checked several times and are equal, I swear:-). I have set the scene and the DOF in VRayforC4D and in my opinion the DOF are more beautiful, those in Maxwell Render have a weak blurring, but this has nothing to do with the software.

As I mentioned before, Maxwell Render has a powerful tool called Multilight that gives us the ability to changing intensity, color of the stage lights, ISO and shutter speed of the camera in post production. A simple refresh, and you have done! With a single render you can pull out several images with different illuminations, that’s a powerful tool! You can also create videos with animations of lights as the one you will find below. In addition to this possibility, the Multilight is useful in post-production. As stated before there is no Color Mapping in Maxwell Render so it's very easy to have areas or objects "burned" or excessively underexposed. In my scene, for example, the white curtains are subject to burns; with Multilight we can lower the lights that affect the appearance of the curtains by using the Next Limit tool, save the image darker but with the curtains to the right exposure (not burned) and then in Photoshop we simply select them with Material ID Channel and replace them, in order to have a more proper exposure. So, from a render we can create many renders for each object that we need to correct and that’s very convenient!

In conclusion, both render engines are spectacular. Personally  I would not abandon either of them for the other. Basically if I have to deliver work with tight deadlines I use VRayforC4D, it suits to many situations and it’s very flexible. But if I have more time or I have to create still life of jewelry where the management of lights and materials must be very accurate, I use Maxwell Render. Even if it’s slower in most situations I noticed that if the number of lights of a set is very high (like in large set such as supermarkets and hotels), Maxwell Render becomes faster than VRayforC4D and AR, also in the process of animation , with a superior cleaning of image. Maxwell Render require the use of render farm or a dedicated computer if you work in CPU, unless you have to create a low number of images that can be managed by a single computer. As mentioned before the latest versions of both software use GPU computing, like many other render engines that are coming on the market. Personally I prefer not to try every new product of different software house, but I choose to use software of solid companies that give me the confidence to carry on my work. Although I've got the opportunity to try Octane Render and it is spectacular! Currently the GPU version of Maxwell Render is still immature and the one of VrayforC4D I still have to test it to verify stability and compatibility with previous versions, but as I always need to create a large number of images and I have a render farm, probably the CPU rendering (or mixed rendering) it’s still the best solution for me because it gives me the possibility to calculate a large number of renders on different works while continuing to modeling on my machine. I hope you have found this test useful. See you soon! All scenes and models are available at these links: