GPU NEXT limits question

[Blakk76]

Good day Everyone

I have question regarding right usage of render limits in GPU NEXT renderer.
As far as i can see there's time/samples/noise levels limits, however, only time limit makes total sense, render stops when a certain time limit is reached.
When I set noise level limit and time limit (let`s take for example default settings), then even if noise level is reached it still renders something untill time limit is reached, and IF I set time limit to zero, it still takes some (quite a lot of) time to "render" something.
And Samples Limit does not works logically at all, let`s say, I saw on previous frame, that render after 3 minutes and 800 samples is OK, so I can set samples limit to 800 and ignore other limits, but it does not works like that at all...
so can anyone guide me through that questions?

[yavor.rubenov]

Hello,

All limits for GPU Next stop the render only on a finished pass. Because of that depending on the scene, resolution etc it might seem as if the rendering continues even after the limit. Also limits can be combined and rendering finishes when any of them is reached.

Noise level - when you set a specific noise level and render - there should be a message in the log: "Progressive rendering finished: specified noise level reached". It is only when you see this message that the noise level for all pixels in the image is reached. If you specify only noise level as a limit rendering might take longer than needed because edges need a lot of samples to reach lower noise levels while the rest of the image will be quite clean. That's why it is always a good idea to combine the noise level limit with samples limit - for example the default of 5000 should be enough for most cases.
You can add the SampleRate render element to see which pixels V-Ray GPU is currently sampling - they will be white and will change in time.

Samples limit - with the sample limit you can set a limit for each pixel - it means that for each individual pixel V-Ray GPU won't trace more than the specified amount of camera rays. On the other hand the statistic you see in the progress when rendering is number of passes - this is the number of times V-Ray GPU will go through the whole image to sample pixels. When rendering V-Ray GPU uses dynamic noise threshold - this means that initially it will start with some high value like 0.5, sample all the pixels for a while and then calculate the noise for each pixel. When most of the pixels have less noise than 0.5 V-Ray will halve the value to 0.25 and start again. This way the resulting noise in the whole image is more evenly spread. Because of this dynamic noise threshold some pixels might not be sampled each pass - meaning that the number of passes is not equal to the number of samples.

So if you want to set limits for an animation it is best to set either a time limit or noise threshold limit depending on your goal - do you need the whole animation done in certain time or with certain quality. You can leave the Samples limit at their default value of 5000 for most cases. You might need to change it only if your render is still noisy but stops and there is a message in the log: "Progressive rendering finished: maximum sampling level reached".

Best regards,
Yavor

[Blakk76]

Hello,

All limits for GPU Next stop the render only on a finished pass. Because of that depending on the scene, resolution etc it might seem as if the rendering continues even after the limit. Also limits can be combined and rendering finishes when any of them is reached.

Noise level - when you set a specific noise level and render - there should be a message in the log: "Progressive rendering finished: specified noise level reached". It is only when you see this message that the noise level for all pixels in the image is reached. If you specify only noise level as a limit rendering might take longer than needed because edges need a lot of samples to reach lower noise levels while the rest of the image will be quite clean. That's why it is always a good idea to combine the noise level limit with samples limit - for example the default of 5000 should be enough for most cases.
You can add the SampleRate render element to see which pixels V-Ray GPU is currently sampling - they will be white and will change in time.

Samples limit - with the sample limit you can set a limit for each pixel - it means that for each individual pixel V-Ray GPU won't trace more than the specified amount of camera rays. On the other hand the statistic you see in the progress when rendering is number of passes - this is the number of times V-Ray GPU will go through the whole image to sample pixels. When rendering V-Ray GPU uses dynamic noise threshold - this means that initially it will start with some high value like 0.5, sample all the pixels for a while and then calculate the noise for each pixel. When most of the pixels have less noise than 0.5 V-Ray will halve the value to 0.25 and start again. This way the resulting noise in the whole image is more evenly spread. Because of this dynamic noise threshold some pixels might not be sampled each pass - meaning that the number of passes is not equal to the number of samples.

So if you want to set limits for an animation it is best to set either a time limit or noise threshold limit depending on your goal - do you need the whole animation done in certain time or with certain quality. You can leave the Samples limit at their default value of 5000 for most cases. You might need to change it only if your render is still noisy but stops and there is a message in the log: "Progressive rendering finished: maximum sampling level reached".

Best regards,
Yavor

Thank you, this is basically how I understand it, however many times after "Progressive rendering finished: specified noise level reached" message V-ray still renders the image for about 2-3 mins. And only time limit works dead on.

[yavor.rubenov]

That is strange and shouldn't happen - can you send a log file from such render ?

[Blakk76]

So in this example "Final noise threshold" was set on 3:30, but render continued untill 04:57 (5 mins was set as time limit) is there an option to user Noise OR Time limit, not only Time limit as the main thing. Because during animation I have some frames that are reaching noise limit faster than time limit and some frames where it is other way around.

[2018/Dec/4|15:38:56] ========================

[2018/Dec/4|15:38:56] Starting frame 162.

[2018/Dec/4|15:38:56] Preparing camera sampler.
[2018/Dec/4|15:38:56] Preparing scene for frame...
[2018/Dec/4|15:38:56] Compiling geometry...
[2018/Dec/4|15:38:56] Preparing ray server.
[2018/Dec/4|15:38:56] Using embree ray tracing.
[2018/Dec/4|15:38:56] Full rayserver build. Took 0.007 ms.
[2018/Dec/4|15:38:56] buildFindAccelStructures. Took 0.000 ms.
[2018/Dec/4|15:38:56] Scene is empty.
[2018/Dec/4|15:38:56] Preparing direct light manager.
[2018/Dec/4|15:38:56] 4 render channels in sequence data
[2018/Dec/4|15:38:56] Preparing global light manager.
[2018/Dec/4|15:38:56] Using 0.61 MB for adaptive lights buffer
[2018/Dec/4|15:38:56] Irradiance sample size is 84 bytes
[2018/Dec/4|15:38:56] Cannot use fast GI because secondary engine is not brute force or none
[2018/Dec/4|15:38:56] Broadcasting NOTIFY_PRE_RENDERFRAME.
[2018/Dec/4|15:38:56] Broadcasting NOTIFY_BEGIN_RENDERING_ACTUAL_FRAME.
[2018/Dec/4|15:38:56] Waiting for image data...
[2018/Dec/4|15:38:56] [RTRenderSession::renderImage] Calling renderImage()
[2018/Dec/4|15:38:57] Running RTEngine
[2018/Dec/4|15:38:57] Old kernel (-1), new kernel (3)

[2018/Dec/4|15:38:57] Initializing CUDA renderer (kernel2 version)...
[2018/Dec/4|15:38:57] Initializing environment kernel
[2018/Dec/4|15:38:57] Number of CUDA devices: 2
[2018/Dec/4|15:38:57] nVidia driver version: 391.33

[2018/Dec/4|15:38:57] Found cache driver data in C:\Users\KOZLIA~1\AppData\Local\Temp/vray_gpu_driver_status.xml
[2018/Dec/4|15:38:57] Min shading rate set to : 6
[2018/Dec/4|15:38:57] Transferring SD tree to CUDA devices...
[2018/Dec/4|15:38:57] SD tree transferred to device 0
[2018/Dec/4|15:38:57] SD tree transferred to device 1
[2018/Dec/4|15:38:57] Transferring Emb SD tree to CUDA devices...
[2018/Dec/4|15:38:57] Embree SD tree transferred to device 0
[2018/Dec/4|15:38:57] Embree SD tree transferred to device 1
[2018/Dec/4|15:38:57] Transferring MB tree to CUDA devices...
[2018/Dec/4|15:38:57] numNodeFaces=0
[2018/Dec/4|15:38:57] numNodes=0
[2018/Dec/4|15:38:57] numFaces=0
[2018/Dec/4|15:38:57] Successfully transferred MB tree to CUDA devices...
[2018/Dec/4|15:38:57] Transferring Emb SD tree to CUDA devices...
[2018/Dec/4|15:38:57] Embree SD tree transferred to device 0
[2018/Dec/4|15:38:57] Embree SD tree transferred to device 1
[2018/Dec/4|15:38:57] Transferring geometric data
[2018/Dec/4|15:38:57] RTRayserver::getOCLGeomData. Took 0.003 ms.
[2018/Dec/4|15:38:57] Transferring vertex normals
[2018/Dec/4|15:38:57] 0 vertex normals transferred to device 0
[2018/Dec/4|15:38:57] Transferring MOVING vertex normals
[2018/Dec/4|15:38:57] 0 MOVING vertex normals transferred to device 0
[2018/Dec/4|15:38:57] Transferring nodes and meshes
[2018/Dec/4|15:38:57] Nodes and meshes transferred to device 0
[2018/Dec/4|15:38:57] Transferring MOVING nodes, meshes and perface node indices
[2018/Dec/4|15:38:57] Successfully transferred MOVING nodes and meshes to device 0
[2018/Dec/4|15:38:57] Transferring vertex normals
[2018/Dec/4|15:38:57] 0 vertex normals transferred to device 1
[2018/Dec/4|15:38:57] Transferring MOVING vertex normals
[2018/Dec/4|15:38:57] 0 MOVING vertex normals transferred to device 1
[2018/Dec/4|15:38:57] Transferring nodes and meshes
[2018/Dec/4|15:38:57] Nodes and meshes transferred to device 1
[2018/Dec/4|15:38:57] Transferring MOVING nodes, meshes and perface node indices
[2018/Dec/4|15:38:57] Successfully transferred MOVING nodes and meshes to device 1
[2018/Dec/4|15:38:57] Transferring map channels
[2018/Dec/4|15:38:57] Map channels transferred
[2018/Dec/4|15:38:57] Transferring moving map channels
[2018/Dec/4|15:38:57] Moving Map channels transferred
[2018/Dec/4|15:38:57] Transferring moving triangles
[2018/Dec/4|15:38:57] Moving triangles count = 0
[2018/Dec/4|15:38:57] Transfering geometry data to devices. Took 2.163 ms.
[2018/Dec/4|15:38:57] Transfering geometry trees to devices. Took 3.282 ms.
[2018/Dec/4|15:38:57] Setting up 11 thread(s)

[2018/Dec/4|15:38:57] Threads completed

[2018/Dec/4|15:38:57] Building trees for 2 meshes with all threads per mesh took 79ms.
[2018/Dec/4|15:38:57] Building trees for 87 meshes with one thread per mesh took 64ms.
[2018/Dec/4|15:38:57] BVH statistics:
[2018/Dec/4|15:38:57] Primitives count: 166 (Mem usage:0.04 MB)
[2018/Dec/4|15:38:57] Meshes count: 89 (Mem usage:0.00 MB)
[2018/Dec/4|15:38:57] Tree depth: 13
[2018/Dec/4|15:38:57] Number of tree nodes: 331
[2018/Dec/4|15:38:57] Tree takes 0.58 MB. Build time 1.14 ms.
[2018/Dec/4|15:38:57] GPU tex loaded V:\HDRI\studio setup blur 002.hdr (idx=0) mip 0, w 1024/1024, h 512/512, channels 1, 1024KB
[2018/Dec/4|15:38:57] Transferring per mesh data for geom type 0
[2018/Dec/4|15:38:57] Transferring per mesh data for geom type 1
[2018/Dec/4|15:38:57] Transferring per mesh data for geom type 2
[2018/Dec/4|15:38:57] GPU tex loaded V:\WORK\ARCH_BY_DAY\app_new\New folder\335270.55d103ee3d157\led_01.jpg (idx= mip 0, w 100/100, h 100/100, channels 3, 58KB
[2018/Dec/4|15:38:57] GPU tex loaded V:\WORK\ARCH_BY_DAY\app_new\New folder\335270.55d103ee3d157\led__white_01.jpg (idx=9) mip 0, w 100/100, h 100/100, channels 3, 58KB
[2018/Dec/4|15:38:57] GPU tex loaded C:\work\promo\nab\nabor\hvoynaya-fanera.jpg (idx=1) mip 0, w 450/450, h 450/450, channels 3, 593KB
[2018/Dec/4|15:38:57] GPU tex loaded C:\work\promo\nab\nabor\Woods & Plastics.Finish Carpentry.Wood.Redwood.bump.jpg (idx=0) mip 0, w 1024/1024, h 1024/1024, channels 1, 1024KB
[2018/Dec/4|15:38:57] GPU tex loaded C:\work\promo\map\cloth2.tif (idx=3) mip 0, w 1500/1500, h 1500/1500, channels 3, 6591KB
[2018/Dec/4|15:38:57] GPU tex loaded C:\work\promo\map\cloth.tif (idx=2) mip 0, w 1500/1500, h 1500/1500, channels 3, 6591KB
[2018/Dec/4|15:38:57] GPU tex loaded V:\WORK\ARCH_BY_DAY\app_new\New folder\335270.55d103ee3d157\metal_body_01.png (idx=6) mip 0, w 2048/2048, h 2048/2048, channels 3, 12288KB
[2018/Dec/4|15:38:57] GPU tex loaded V:\WORK\ARCH_BY_DAY\app_new\New folder\335270.55d103ee3d157\metal_refrect_01.png (idx=5) mip 0, w 2048/2048, h 2048/2048, channels 3, 12288KB
[2018/Dec/4|15:38:57] GPU tex loaded C:\work\promo\screencapture-currenttime-tv-doc-2018-11-28-16_09_38.png (idx=7) mip 0, w 1920/1920, h 4325/4325, channels 3, 24328KB
[2018/Dec/4|15:38:58] GPU tex loaded V:\WORK\ARCH_BY_DAY\app_new\New folder\335270.55d103ee3d157\keyboard_character_mas k_01.png (idx=4) mip 0, w 4096/4096, h 4069/4069, channels 3, 48828KB
[2018/Dec/4|15:38:58] Number of lights: 0
[2018/Dec/4|15:38:58] 0 point lights transferred to device 0
[2018/Dec/4|15:38:58] 0 point lights transferred to device 1
[2018/Dec/4|15:38:58] Number of lights: 0
[2018/Dec/4|15:38:58] Total number of lights added by updateLights(): 0
[2018/Dec/4|15:38:58] Total memory allocated with oclbufman: 190504048
[2018/Dec/4|15:38:58] Photon size is 56 bytes.

[2018/Dec/4|15:38:58] Light cache sample size is 136 bytes.
[2018/Dec/4|15:38:58] Light cache built state is 0
[2018/Dec/4|15:38:58] Building light cache.
[2018/Dec/4|15:38:58] Tracing 1000000 image samples for light cache in 64 passes.

[2018/Dec/4|15:38:58] Setting up 11 thread(s)

[2018/Dec/4|15:39:00] Threads completed

[2018/Dec/4|15:39:00] Merging light cache passes.
[2018/Dec/4|15:39:00] Setting up 8 thread(s)

[2018/Dec/4|15:39:00] Threads completed

[2018/Dec/4|15:39:00] Post-merging pass 0 with 505 samples
[2018/Dec/4|15:39:00] Post-merging pass 1 with 532 samples
[2018/Dec/4|15:39:00] Post-merging pass 2 with 506 samples
[2018/Dec/4|15:39:00] Post-merging pass 3 with 464 samples
[2018/Dec/4|15:39:00] Post-merging pass 4 with 554 samples
[2018/Dec/4|15:39:00] Post-merging pass 5 with 511 samples
[2018/Dec/4|15:39:00] Post-merging pass 6 with 560 samples
[2018/Dec/4|15:39:00] Post-merging pass 7 with 488 samples
[2018/Dec/4|15:39:00] Post-merging pass 8 with 563 samples
[2018/Dec/4|15:39:00] Post-merging pass 9 with 589 samples
[2018/Dec/4|15:39:00] Post-merging pass 10 with 582 samples
[2018/Dec/4|15:39:00] Post-merging pass 11 with 619 samples
[2018/Dec/4|15:39:00] Post-merging pass 12 with 590 samples
[2018/Dec/4|15:39:00] Post-merging pass 13 with 644 samples
[2018/Dec/4|15:39:00] Post-merging pass 14 with 676 samples
[2018/Dec/4|15:39:00] Post-merging pass 15 with 642 samples
[2018/Dec/4|15:39:00] Post-merging pass 16 with 631 samples
[2018/Dec/4|15:39:00] Post-merging pass 17 with 617 samples
[2018/Dec/4|15:39:00] Post-merging pass 18 with 585 samples
[2018/Dec/4|15:39:00] Post-merging pass 19 with 680 samples
[2018/Dec/4|15:39:00] Post-merging pass 20 with 644 samples
[2018/Dec/4|15:39:00] Post-merging pass 21 with 647 samples
[2018/Dec/4|15:39:00] Post-merging pass 22 with 659 samples
[2018/Dec/4|15:39:00] Post-merging pass 23 with 674 samples
[2018/Dec/4|15:39:00] Post-merging pass 24 with 714 samples
[2018/Dec/4|15:39:00] Post-merging pass 25 with 674 samples
[2018/Dec/4|15:39:00] Post-merging pass 26 with 685 samples
[2018/Dec/4|15:39:00] Post-merging pass 27 with 673 samples
[2018/Dec/4|15:39:00] Post-merging pass 28 with 657 samples
[2018/Dec/4|15:39:00] Post-merging pass 29 with 725 samples
[2018/Dec/4|15:39:00] Post-merging pass 30 with 683 samples
[2018/Dec/4|15:39:00] Post-merging pass 31 with 734 samples
[2018/Dec/4|15:39:00] Post-merging pass 32 with 752 samples
[2018/Dec/4|15:39:00] Post-merging pass 33 with 731 samples
[2018/Dec/4|15:39:00] Post-merging pass 34 with 673 samples
[2018/Dec/4|15:39:00] Post-merging pass 35 with 710 samples
[2018/Dec/4|15:39:00] Post-merging pass 36 with 740 samples
[2018/Dec/4|15:39:00] Post-merging pass 37 with 770 samples
[2018/Dec/4|15:39:00] Post-merging pass 38 with 722 samples
[2018/Dec/4|15:39:00] Post-merging pass 39 with 773 samples
[2018/Dec/4|15:39:00] Post-merging pass 40 with 751 samples
[2018/Dec/4|15:39:00] Post-merging pass 41 with 749 samples
[2018/Dec/4|15:39:00] Post-merging pass 42 with 790 samples
[2018/Dec/4|15:39:00] Post-merging pass 43 with 790 samples
[2018/Dec/4|15:39:00] Post-merging pass 44 with 814 samples
[2018/Dec/4|15:39:00] Post-merging pass 45 with 751 samples
[2018/Dec/4|15:39:00] Post-merging pass 46 with 749 samples
[2018/Dec/4|15:39:00] Post-merging pass 47 with 765 samples
[2018/Dec/4|15:39:00] Post-merging pass 48 with 755 samples
[2018/Dec/4|15:39:00] Post-merging pass 49 with 752 samples
[2018/Dec/4|15:39:00] Post-merging pass 50 with 769 samples
[2018/Dec/4|15:39:00] Post-merging pass 51 with 743 samples
[2018/Dec/4|15:39:00] Post-merging pass 52 with 762 samples
[2018/Dec/4|15:39:00] Post-merging pass 53 with 810 samples
[2018/Dec/4|15:39:00] Post-merging pass 54 with 787 samples
[2018/Dec/4|15:39:00] Post-merging pass 55 with 816 samples
[2018/Dec/4|15:39:00] Post-merging pass 56 with 798 samples
[2018/Dec/4|15:39:00] Post-merging pass 57 with 810 samples
[2018/Dec/4|15:39:00] Post-merging pass 58 with 809 samples
[2018/Dec/4|15:39:00] Post-merging pass 59 with 780 samples
[2018/Dec/4|15:39:00] Post-merging pass 60 with 787 samples
[2018/Dec/4|15:39:00] Post-merging pass 61 with 837 samples
[2018/Dec/4|15:39:00] Post-merging pass 62 with 808 samples
[2018/Dec/4|15:39:00] Light cache contains 14735 samples.
[2018/Dec/4|15:39:00] Light cache samples collected; compiling lookup tree.
[2018/Dec/4|15:39:00] Light cache base memory usage is 7.3 MB.
[2018/Dec/4|15:39:00] Prefiltering light cache.
[2018/Dec/4|15:39:00] Average rays per light cache sample: 69.08 (min 1, max 1144)
[2018/Dec/4|15:39:00] Using premultiplied light cache.
[2018/Dec/4|15:39:00] Using adaptive lights ver2
[2018/Dec/4|15:39:00] LightCache does not include valid adaptive lights data. Adaptive lights will be calculated at runtime
[2018/Dec/4|15:39:00] Lightmap successfully loaded.
[2018/Dec/4|15:39:00] GPU light cache: 14735 samples stored.
[2018/Dec/4|15:39:00] GPU light cache: 2617 nodes stored.
[2018/Dec/4|15:39:00] GPU light cache: 2618 leaves stored.
[2018/Dec/4|15:39:00] Mtls 119.11MB, Geo 70.80MB, LightCache 1.04MB, Reserved 0.26MB (data index 0)
[2018/Dec/4|15:39:00] Setting up 17 thread(s)
[2018/Dec/4|15:39:00] Building optimized CUDA kernel (3) for comp cap 5.2 for device 1...
[2018/Dec/4|15:39:00] PTX file generated with CUDA Toolkit v7.5 for CUDA compute capability 2.0. CUDA runtime version 7050.
[2018/Dec/4|15:39:00] Optimized CUDA kernel assembled successfully
[2018/Dec/4|15:39:00] Total memory for device 1 : 8192 MB, free 6147 MB
[2018/Dec/4|15:39:01] Progressive sampling base buffer memory usage: 190.44 MB
[2018/Dec/4|15:39:28] Noise threshold lowered to 0.142857 with 1.2 percent active pixels.
[2018/Dec/4|15:39:30] Noise threshold lowered to 0.090909 with 2.5 percent active pixels.
[2018/Dec/4|15:39:34] Noise threshold lowered to 0.052632 with 8.4 percent active pixels.
[2018/Dec/4|15:39:45] Noise threshold lowered to 0.037037 with 40.3 percent active pixels.
[2018/Dec/4|15:40:01] Noise threshold lowered to 0.028571 with 51.2 percent active pixels.
[2018/Dec/4|15:40:17] Noise threshold lowered to 0.023256 with 55.6 percent active pixels.
[2018/Dec/4|15:40:32] Noise threshold lowered to 0.019608 with 48.5 percent active pixels.
[2018/Dec/4|15:40:48] Noise threshold lowered to 0.016949 with 52.2 percent active pixels.
[2018/Dec/4|15:41:06] Noise threshold lowered to 0.014925 with 66.8 percent active pixels.
[2018/Dec/4|15:41:22] Noise threshold lowered to 0.013333 with 58.5 percent active pixels.
[2018/Dec/4|15:41:39] Noise threshold lowered to 0.012048 with 56.0 percent active pixels.
[2018/Dec/4|15:41:57] Noise threshold lowered to 0.010989 with 60.3 percent active pixels.
[2018/Dec/4|15:42:14] Noise threshold lowered to 0.010101 with 61.3 percent active pixels.
[2018/Dec/4|15:42:32] Final noise threshold set: 0.010000
[2018/Dec/4|15:43:58] Threads completed
[2018/Dec/4|15:43:58] Closing RTEngine
[2018/Dec/4|15:43:58] Max. Mpaths per second: 5.072204
[2018/Dec/4|15:43:58] Sampling level reached: 1936 samples per pixel
[2018/Dec/4|15:43:58] V-Ray GPU kernel render time: 297.38s
[2018/Dec/4|15:43:58] Total GPU samples 1331992544
[2018/Dec/4|15:43:58] Device 0 statistics:
[2018/Dec/4|15:43:58] GPU time: 91.608634 %
[2018/Dec/4|15:43:58] GPU ray trace time 99.932257 %
[2018/Dec/4|15:43:58] Transfer to GPU time 0.064580 %
[2018/Dec/4|15:43:58] Transfer from GPU time 0.003163 %
[2018/Dec/4|15:43:58] Device 1 statistics:
[2018/Dec/4|15:43:58] GPU time: 92.675970 %
[2018/Dec/4|15:43:58] GPU ray trace time 99.107025 %
[2018/Dec/4|15:43:58] Transfer to GPU time 0.082495 %
[2018/Dec/4|15:43:58] Transfer from GPU time 0.810480 %
[2018/Dec/4|15:43:58] [RTRenderSession::frameEnd] Calling endFrame()
[2018/Dec/4|15:43:58] Number of raycasts: 3283942 (1.58 per pixel)
[2018/Dec/4|15:43:58] Camera rays: 1000087 (0.48 per pixel)
[2018/Dec/4|15:43:58] Shadow rays: 0 (0.00 per pixel)
[2018/Dec/4|15:43:58] GI rays: 2012434 (0.97 per pixel)
[2018/Dec/4|15:43:58] Reflection rays: 271421 (0.13 per pixel)
[2018/Dec/4|15:43:58] Refraction rays: 0 (0.00 per pixel)
[2018/Dec/4|15:43:58] Unshaded rays: 0 (0.00 per pixel)
[2018/Dec/4|15:43:58] Clearing global light manager.
[2018/Dec/4|15:43:58] Clearing direct light manager.
[2018/Dec/4|15:43:58] Number of light evaluations: 0 (0.00 per pixel)
[2018/Dec/4|15:43:58] Clearing ray server.
[2018/Dec/4|15:43:58] Maximum memory usage for resman: 0.00 MB
[2018/Dec/4|15:43:58] Clearing geometry.
[2018/Dec/4|15:43:58] Number of intersectable primitives: 166
[2018/Dec/4|15:43:58] SD triangles: 0
[2018/Dec/4|15:43:58] MB triangles: 0
[2018/Dec/4|15:43:58] Static primitives: 166
[2018/Dec/4|15:43:58] Moving primitives: 0
[2018/Dec/4|15:43:58] Infinite primitives: 0
[2018/Dec/4|15:43:58] Static hair segments: 0
[2018/Dec/4|15:43:58] Moving hair segments: 0
[2018/Dec/4|15:43:58] Clearing camera image sampler.
[2018/Dec/4|15:43:58] Clearing camera sampler.
[2018/Dec/4|15:43:58] Clearing DMC sampler.
[2018/Dec/4|15:43:58] Clearing path sampler.
[2018/Dec/4|15:43:58] Clearing color mapper.
[2018/Dec/4|15:43:58] Broadcasting NOTIFY_POST_RENDERFRAME.
[2018/Dec/4|15:43:58] [RTRenderSession::renderEnd] Calling endSequence()
[2018/Dec/4|15:43:58] 0 error(s), 0 warning(s)
[2018/Dec/4|15:43:58] ================================
[2018/Dec/4|15:43:58] Broadcasting NOTIFY_POST_RENDER.
[2018/Dec/4|15:47:04] Saving to file.
[2018/Dec/4|15:47:04] Saving distributed renderer.
[2018/Dec/4|15:47:04] Saving filter parameters.
[2018/Dec/4|15:47:04] Saving done.

After that I`ve tried setting time limit to 0, but still the same effect, threshold is reached in 3:30 and render stops at 5:23 then it stars doing "number of ray per pixel changed" for another 2 mintes...with number of passes going beyond 1500000

[yavor.rubenov]

Ahaa from the log I see that you are not using Update 1 but an older version. In older versions the message you see about the noise threshold - "Final noise threshold set: 0.010000" was a bit unclear - it doesn't mean the noise is reached. It means that we've now set the noise threshold to 0.01 but there are still many pixels above that noise level - this is part of the dynamic noise threshold thing I explained. After you see this message V-Ray will sample with noise threshold 0.01, before it it was using 0.010101 and before it 0.010989..

is there an option to user Noise OR Time limit

- that's how things currently work.

with number of passes going beyond 1500000

- that's the reason that always settings Max Sample limit is recommended.


There are some more details here:
https://forums.chaosgroup.com/forum/...logic-question