Vray Brute Force Vs Light Cache

16 views
Skip to first unread message

Bartley Trowbridge

unread,
Aug 5, 2024, 1:07:44 AM8/5/24
to beertjucillong
Thissection controls and fine-tunes various aspects of the Light Cache. The Light cache rollout appears only when the Light cache method is selected as a Primary engine or Secondary engine on the Global Illumination rollout.

Note that Path guiding produces different results, when Max ray intensity or Sub-pixel mapping are enabled compared to when switched off. Different results in the raw render elements are expected, but they produce the correct result when multiplied by the respective filter render element. In some scenes containing diffuse surfaces with low poly count, the results may differ compared to rendering with this option switched off.


The Sample size parameter controls the size of the individual light cache samples. Smaller values produce a more detailed lighting solution, but are noisier and take more RAM. Larger values produce less detail, but take less RAM and may be faster to calculate.


The first set of images shows how the Retrace parameter can be used to reduce light leaks due to the light cache interpolation. The scene is an interior with parts of the exterior visible. For the GI settings, Brute Force is used for primary bounces (it is selected as the Primary engine in the Global Illumination Rollout) and Light cache is the Secondary engine.


The bright light cache samples from the exterior blend with the darker samples from the interior, causing light leaks when the light cache is calculated. The Retrace option (with the default value of 2.0) successfully resolves the problem at the expense of increased calculation time for the light cache.


In this second example, when Retrace is turned on, its value determines whether brute force is used instead of light cache near corners or objects that are close together. When the Retrace is greater than 0, only brute force is used which helps give a far more accurate result, although this may slow down rendering. However, if the Retrace value is 0.0, then only the light cache is used and thus retracing is disabled. This is faster but less accurate and more prone to noise and artifacts from light leakage.


If a high value is used for the Retrace, light leakage and artifacts will disappear and brute force will be used for a larger area. As a result, the render may appear more noisy. Therefore, the default Retrace value of 2.0 is sufficient in most cases.


For interior renders or other complicated lighting situations, increasing the light cache Subdivs parameter makes the light cache smoother and more accurate; values around 2000 or 3000 typically work well.


Do not apply perfectly white or near-white materials to the majority of the objects in the scene, as this causes excessive render times. This is because the amount of reflected light in the scene will decrease very gradually and the light cache will have to trace longer paths. Also avoid materials that have one of their RGB components set to maximum (255) or above.


There is no difference between light caches computed for primary bounces (direct visualization) and for secondary bounces. You can safely use light caches computed in one of these modes for the other.


Similar to the photon map, you can get "light leaks" with the light cache around very thin surfaces with substantially different illumination on both sides. Sometimes it might be possible to reduce this effect by assigning different GI Surface ID's to the objects on both sides of the thin surface (see the Object settings window); the effect can also be reduced by decreasing the Sample size and/or the filtering.


An alternative way of loading a light cache is by dragging and dropping a .vrlmap file directly in your viewport. Just be aware that doing so will load it for usage immediately, turning GI on, changing its Secondary engine to Light cache and its Mode to From file.


Beyond this, you may see a better image quality from Brute force with things like a close up detail of hair and fur or things of that hyper detailed nature. The general reality of it though is that Brute Force is a noisy solution unless you are going to crank up the settings, in which case you are heading out on vacation after you hit render.


I will add that I have begun using Brute force in some testing because I believe that if the noise isn't too bad, it does create a softer render. Times are longer, but there has to be a balance somewhere.


My VERY GENERIC understanding is that Brute Fore is the Unbiased version of Vray. I doesn't interpolate between samples and make blurred down-sampled maps of the geometry before rendering which average the light samples of an area and remove some of the complexity. Light renders more accurately and more subtle- hence the softness.


In the end, I think this is the debate people have when they start talking about Maxwell as a better render engine than VRay. I don't agree, but Maxwell, among other things, is basically Brute force with some other twists in lighting.


I can bet you that if Peter used anything other than brute force and used your typical vray settings, he would have gotten the same image in 1/4th the render time. It's a great image, but nothing about it screams use brute force. It was an absolute waste of time and contributed nothing to the final image. It's just my 2 cents, but brute force is idiotic for stills.


Brute force I see as a luxury, for when you have plenty of time to get renders out. Other than that, it never gets used by me - and I disagree that you need to use it when you have moving objects in your scene, the irradiance map is still very much usable. In fact, aside from exceptionally fine lighting details the irradiance map will do it all - depending on how you set it up.


Brute force is great, because there's so little setup time involved and many see it as a get out clause instead of learning how to fine tune the irradiance map. If you just invest some time and little research/test renders into how the irradiance map works, you can get some excellent results out of it very quickly.


The only times I really consider using brute force is when I have a scene with either a complex lighting/indirect lighting solution or lots of fine details that get lost due to irradiance map interpolation, for example ornate mouldings/cornicing.


IMO the main thing is know your tool deep enough to get the most of it, brute force is the precise way to calculate GI everything else is an approximation, average or interpolated way to speed up the process, there is not good or bad here, just different ways, now Glossy reflections and refraction react in a different way when you use brute force, also it works better than light cache at least as second bounce when you work with displacement or intense landscaping type of mesh, as mentioned before when you have fine details in your model, brute force will give you the right interpretation of it.


with that being said, you need to adjust V Ray in a proper way to use brute force, in that making off I am not surprised that he used Brute force as first bounce, what called my attention was that in his antialising values(very important when you use brute force) he set it up with min subdibs as 1 and max subdiv as 50 with a threshold controlled by the general VRay DMC sampler threshold, this is not efficient at all, he could put 500 as max subdivs but if the DMC sampler threshold is to high it will work until 8 subdivs maybe then it wont sample any more, if you set up you General DMC smapler so low that you get a clean brute force solution you also will push all the other noise values so high that you have double or triple your render time for not reason.


You should see Christopher Nichols video and he explain very well his workflow( he says that it is used in his animations project because quality and best solution for moving objects) of course when we talk about Movie industry render times are different than our standards, but I managed to do animations with brute force as second bounce and good still with brute force as first bounce. top artist as Bertrand Benoit, Alex Roman and others use brute force for some of their projects, to get the maximum quality, but again they are very savvy on V-RAy and the the key.


Ive been using vray for still renderings for a 7 years now but Ive only recently begun using in for animations. I have had no problems until recently when my usual animation setup produced horrible results, the problem seems to be that i have animating objects and animating lights.. The typcial flythrough settings i use with light cache and irradiance map just dont do it...


Its a 3000 frame animation and its starts out with all of the lights on but with 0 intensity (vray light materials too), the lights are set up to turn on (get brighter) as the camera moves through the space to give a gradual feeling from moody to well lit. So, lights are constantly turning on through the whole animation. Doors, blinds and curtains open up as well, these are my only moving objects but they should also play with the lighting greatly..


It almost seems like the light cache didn't calculate any of the lighting because the animation starts with them all off? I dunno that's the only thing i could figure. Either way, its a sad result after calculating the LC and IR after 10 hours lol.


Is this kind of animation not possible using the LC? Is brute force or some other setting the answer? Any help on this will be greatly appreciated, it took some time and thought to set this up the way i wanted, it would be a shame to end up rendering it as a standard fly through with the lights always on and only the camera animating. Thanks allot!!


Animated lights, objects, and camera in one file should be done through light cache as a secondary and single frame with Brute Force set as the primary. You could probably go 1500 on the light cache and save a few of those hours of render time. 15000 is outrageous.

3a8082e126
Reply all
Reply to author
Forward
0 new messages