Page 3 of 42 FirstFirst 1234513 ... LastLast
Results 31 to 45 of 619

Thread: 11.5's BenchmarkMarbles.lws - share your machine's render time here

  1. #31
    Studio Animator DonJMyers's Avatar
    Join Date
    Aug 2003
    Location
    Los Angeles, CA
    Posts
    566
    Quote Originally Posted by Celshader View Post
    I did not change the settings. I loaded up the scene as-is from the 11.5 content directory and hit F9. What happens when you render this benchmark on your machine?
    My screw up Jen. I lied... I haven't really run the whole benchmark yet so I thought it would be as pebbly at the end as the beginning. When I rendered a limited area ALL 192 TIMES it came out smooth. Ooops!

    Oh well, it was worth screwing up just to read your detailed explanation of what happens.

  2. #32
    Super Member XswampyX's Avatar
    Join Date
    Aug 2010
    Location
    Kernow
    Posts
    2,057


    2hrs 40 mins. Phenom II X6 @3.20 GHz

    What's with the awful/incorrect colour space settings?

  3. #33
    that looks linear, switch to sRGB in the Preferences.
    (can't remember, but LW10 or LW11.0 seemed to forget the mode)
    LW vidz   DPont donate   LightWiki   RHiggit   IKBooster   My vidz

  4. #34
    Super Member COBRASoft's Avatar
    Join Date
    Dec 2005
    Location
    Oudenburg, BEL
    Posts
    3,178

    New XNB Laptop: 2h 33m

    My new laptop took 2h 33m 3s to finish.

    Intel i7-3610QM @2.3GHz, 16GB ram, Win8 Pro, 2x SSD HD's.

    Any idea why the dithering is set to 'normal' instead of 'off'?

  5. #35
    dithering set to normal should look better in this case, afaik.
    LW vidz   DPont donate   LightWiki   RHiggit   IKBooster   My vidz

  6. #36
    Triglycerous Gluteous Dave Jerrard's Avatar
    Join Date
    Apr 2005
    Location
    Burbank
    Posts
    1,061
    Quote Originally Posted by DonJMyers View Post
    Jen I noticed the version of this scene sent as a sample has shading samples and light samples at 1. This makes renders fast but pebbly. But your render looks great. Did you change those settings and, if so, doesn't that make the benchmark times messed up?
    It will render clean. This scene was originally set up in 9.6, and the render settings are set in 11 to match what 9.6 & 10 would do with the same AA and threshold. In 9.6, I'd set the samples for various nodes to the minimum to save time. You can bump this up, but it will just take a lot longer, and the difference would be negligible. If you increase the samples, EVERY Adaptive Sampling pass is going to have to render those extra samples. On scenes with a lot of detail, that usually mean a lot of wasted time on samples that aren't really needed. I came up with my settings over several years of experimentation & production. My normal production settings are AA of 3 and a threshold of 0.03, which give me a maximum of 48 samples in 9.6. I've rarely had to increase that maximum.

    For those who want to try it in 9.6, you'll have to edit the nodes. LightWave 11 has a nasty habit of butchering node settings (and lights) by stripping out the samples that have been set before, so when you load the scene into 9.6 or 10.x, all the samples are set back to 8. Set them all to 1, except for the Kappas, which I left at 8. You can set those to 1 as well - it won't make an real difference in quality or time. I just recommend leaving them for an honest comparison.

    Quote Originally Posted by Celshader View Post
    Dave Jerrard tends to throw the burden of render cleanup on the back of Adaptive Sampling. That is why he set the Light/Shading Samples to 1, and the Rays Per Evaluation for Monte Carlo to 4. With the Minimum Samples set to 3, the scene actually starts with a base level of 3 Light/Shading Samples -- 1 Light/Shading sample for each of the 3 camera rays. The scene also has a base level of 12 RPE for the Monte Carlo Radiosity -- 4 RPE for each of the 3 camera rays initially fired per pixel.
    By the way, it's that Monte Carlo that's creating the nice caustic effects on the wood. You can't get that any other way in LightWave. The Caustics setting ONLY 'works' with LW's lights, and is useless at best. Interpolated radiosity will just make this blurry & blobby. I generally use 4 RPE for radiosity, either for Backdrop Only or Monte Carlo, and I never use interpolated unless I absolutely have to. 4RPE gives a good smooth noise that AA can clean up pretty well. Only rarely do I find I need to use a higher value.

    With a Threshold of 0.01 and a Maximum Samples of 192, the camera is forced to fire up to 192 rays per pixel to clean up the initial mess. It will keep firing rays until the difference in brightness between each pair of pixels falls below 0.01, or until 192 rays have been fired for each pixel -- whichever comes first. Each additional camera ray takes an additional 1 Light Sample, 1 Shading Sample, and 4 Rays Per Evaluation. A pixel that hits the maximum 192 camera samples will also have 192 Light Samples (1 Light Sample x 192 camera rays), 192 Shading Samples (1 Shading Sample x 192 camera rays) and 768 Rays Per Evaluation (4 RPE x 192 camera rays).
    Since this scene is rendered using ONLY radiosity as a light source, the Lighting Samples really makes no difference. As for the Shading Samples, I've used 1 in 9.6 for all nodes, and would only increase it for those nodes that actually needed a bit more. How wonderful it is to know that now if one node needs some extra samples, I have to bump ALL of them up and suffer a major render hit. Same with lights.


    The Ray Cutoff (Render Globals > Render > Ray Cutoff) is just as finicky as the Threshold in this scene -- a value of 0.01. This is the "Threshold" setting for Ray Recursion. The Ray Recursion Limit is set to 12, but the rays will stop recursing if the difference in brightness between the last two bounces was less than 0.01 or until they stop hitting a reflective/refractive surface. If anyone want this render to take even longer, set the Ray Cutoff to 0.0 to force all rays to recurse up to 12 bounces until they stop hitting a reflective/refractive surface.
    That will make any machine cry. I tried that on a single marble. I stopped the render after about an hour since it never even got through the first pas yet. Then again, that was also with recursions set to 64.



    Quote Originally Posted by jburford View Post
    Hmnnn, to be honest am really confused as to why it takes so long on my system, would think that it could power through it better.

    Time 3h 18m 57s
    See that swirly marble to the left? He's the bastard. He was designed to push the machine. Each one of those glass swirls in it has a different refractive index. Then there's the reflections of the marbles in the floor. This was not desighned to be a fast render. On the original scenes I did, this frame took over 11 hours on my quad core. This version is made to be LW 11 native - no third party plugins.


    Quote Originally Posted by Philbert View Post
    OMG. I just did the render, 9h 32m. The whole time I 'm thinking it's strange that the Render Threads was set to 4, not 8 like I would normally set it. But I figured the instructions said "load the scene and hit F9" so that's all I did. Then I came here again and saw everyone else's set to 8. Then I remembered that setting is saved in LW not in the scene. Since I had just installed 11.5 I hadn't changed it from the default.
    Set them to automatic and load the scene again. Odd that your install didn't set this for you.



    Quote Originally Posted by XswampyX View Post


    2hrs 40 mins. Phenom II X6 @3.20 GHz

    What's with the awful/incorrect colour space settings?
    Nothing wrong with it, though your image looks like it's been converted to sRGB. The scene is set to be exactly what I wanted, and is a match for 9.61, which this scene was originally created in. See the image Jen posted to see what it should look like. I even rendered a nice big 9000 x 6000 version in 9.61 that I printed out as a 30x20 poster and it looked great.



    Quote Originally Posted by COBRASoft View Post
    My new laptop took 2h 33m 3s to finish.

    Intel i7-3610QM @2.3GHz, 16GB ram, Win8 Pro, 2x SSD HD's.

    Any idea why the dithering is set to 'normal' instead of 'off'?
    I was saving a 24bit image, not a HDR. After having to deal with various print shops that can't even handle a .psd, I doubted they could handle anything with higher range. If you're saving HDR, then turn that to none.


    He Who Doesn't Mess With Color Space Stuff If He Doesn't Have To.
    - Any typos are the result of a creative, yet rebellious keyboard.

    ASUS KGPE-D16 MB
    2 AMD Opteron 6272 CPUs
    32GB Kingston KVR1333D3D4R9S RAM
    EVGA|01G-P3-1556-KR GTX550Ti video card

  7. #37
    He Who Is Dave Jerrard sez:

    By the way, it's that Monte Carlo that's creating the nice caustic effects on the wood.

    Yup. I remember back when 9.x was out, I built a "camera" model in LW complete with a glass lens, and used MC radiosity to project an external scene onto a white "film".

  8. #38

  9. #39
    Not so newbie member lardbros's Avatar
    Join Date
    Apr 2003
    Location
    England
    Posts
    5,802
    Quote Originally Posted by Dave Jerrard View Post
    How wonderful it is to know that now if one node needs some extra samples, I have to bump ALL of them up and suffer a major render hit. Same with lights.


    Would it work better if there was an override on each material node? So you could simply plug a multiplier into the nodes you wanted to? That way it would use a global setting for all materials, except the ones with the override plugged in?

    Did you find you got faster render-times with the old system of having individual samples for each and every node?

    Love this sample scene by the way... will post my results later on!
    LairdSquared | 3D Design & Animation

    Desk Work:
    HP Z840, Dual Xeon E5-2690 v2, 32GB RAM, Quadro K5000 4GB
    Desk Home:
    HP Z620, Dual Xeon E5-2680, 80GB RAM, Geforce 1050 Ti 4GB

  10. #40
    My work custom--built PC

    LW 11.5

    Render time: 1h 7m 55s
    CPU: 3.5 GHz i7 3930k six core with HT
    RAM: 32 GB
    Overclocking: to 4.2GHz
    OS: Windows 7 Pro 64-bit

  11. #41
    Super Member Cougar12dk's Avatar
    Join Date
    Dec 2003
    Location
    Denmark
    Posts
    1,078
    Blog Entries
    1
    I just gave it a whirl

    2h 0m 3s
    Core i7 3770K, current overclock @ 3.88Ghz
    16GB DDR3
    Attached Images Attached Images
    Core i7-3770K 3.5Ghz, @ 4.181GHz
    16GB DDR3 PC3-12800 - 1600Mhz
    Gigabyte GeForce GTX 1060 WindForce OC 6GB GDDR5 PCIe-3.0 x16
    Windows 10 Professional 64bit

  12. #42
    Here's mine -

    1:37:38 - Not a bad time I guess

    Click image for larger version. 

Name:	BenchmarkMarbles_Render_Time.JPG 
Views:	198 
Size:	95.3 KB 
ID:	111132
    Last edited by The Wizzard; 02-05-2013 at 09:19 AM. Reason: Added render time
    Windows 7 x64
    Intel i7-4770K @ 4.6Ghz
    Nvidia GeForce GTX 980 ti
    Too many SSD's and Monitors
    32Gb DDR3


    _______________
    Coffee, M&Ms, Red Bull - On Tap
    http://www.youtube.com/3ative

  13. #43
    May the sauce be with you starbase1's Avatar
    Join Date
    Feb 2003
    Location
    London, UK
    Posts
    5,583
    I got 1h 31m 57s with a Chillblast PC I bought just before Christmas, the best I could afford.


    4.5GHz Core i5 3570K, 8GB, Quadro 2000 workstation

    • Fractal Design Define R4 Low Noise Case - Black Pearl
    • Intel Core i7 3770K Processor Overclocked to up to 4.6GHz
    • Corsair Hydro Series H40 Liquid Cooler
    • Generic thermal paste
    • Additional and Quiet Fans Upgrade Pack
    • AeroCool 1000 Touch Screen Fan Controller
    • Asus P8Z77-V Motherboard
    • 32GB Corsair PC3-12800 1600MHz DDR3 Memory (4 x 8GB sticks)
    • NVIDIA Quadro 2000 1GB Professional Video Card
    • 120GB Corsair Force 3 Solid State Drive
    • 3000GB 7200rpm Hard Disk - 6Gbps
    • 3000GB 7200rpm Hard Disk - 6Gbps
    • Samsung 24x DVD-RW Drive
    • Corsair CX 750W 80 PLUS Bronze Certified PSU
    • Asus Xonar DS 7.1 PCI Sound Card
    • 17 in 1 3.5" Internal Card Reader
    • 2 Port USB 3.0 Card
    • Windows 7 Professional 64 bit
    • Standard Chillblast Cable Management



    Nick
    Faith is the opposite of Intelligence.

    Touched by his noodley appendage!
    My Lightwave Graphics & Downloads!!!

  14. #44
    hi, i see you wrote two types

    i5 3570K
    i7 3770K

    was i7 3770K the fast one?...
    LW vidz   DPont donate   LightWiki   RHiggit   IKBooster   My vidz

  15. #45
    May the sauce be with you starbase1's Avatar
    Join Date
    Feb 2003
    Location
    London, UK
    Posts
    5,583
    Yep, its the 17 one...
    Faith is the opposite of Intelligence.

    Touched by his noodley appendage!
    My Lightwave Graphics & Downloads!!!

Page 3 of 42 FirstFirst 1234513 ... LastLast

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •