Jump to content

Valtiel

Members
  • Posts

    59
  • Joined

Personal Information

  • Country
    Canada

Valtiel's Achievements

Newbie

Newbie (1/14)

10

Reputation

  1. Hey folks. I haven't been around in ages (I ended up doing a degree in urban planning instead of jumping into archviz). My employment's been compromised, to put it simply. I found my way back here with all the extra indoor-time I have and planning my exit strategy once Autodesk denies me an activation on my old Max license. For now, I've got my machine crunching away on Folding @ Home COVID-19 therapy simulations. Good to see the forums alive and everyone doing fairly well and mostly employed.
  2. Looks like it's back up. Good timing too. Given the news about Autodesk's impending....retcon of the meaning of a perpetual license, I'm thinking about my exit strategy from Max. It will be terrible to give up Forestpack and Railclone but knowing there's some archviz community in other packages will help me figure out where to go. Thanks, Marton!
  3. Hey folks. Here’s how I got them to eventually play nice. If you aren’t interested in the preface and supplementary notes please skip down to the first numbered header in bold. Preface I bought Ozone awhile ago and finally got around to trying to make it work. On paper, it appeared to give the best control and results with expressed compatibility with my other plug-ins, however documentation/conversation on these things has been a bit sparse and Ozone’s pedigree in the primarily stand-alone Vue is valid cause for pause. I finally got around to wrestling with the three, trying to get them to work. It has been a bit of a journey and I started making notes for myself on best practices and pitfalls to be aware of when working with them. Given the lack of conversation out there supporting users exploring this workflow, I want to share it with the community as I don’t feel everyone should slog through the same process just to get a sweet sky. Extra notes and disclaimers For those interested, ForestPack Pro appears, so far, to get along well enough with the rest of this group. Can’t see why they wouldn’t but software developed by different people has all sorts of ways of getting tangled under the hood. “Best Practices” is a term I’m applying lightly here. I haven’t trialed all settings and a lot more testing needs to be done to call any of these recommendations “best” by any measure. They should be a good start though. Please share your own experiences, corrections, or expansions; they are most welcome. Sorry but I haven’t produced any visual aides or examples yet and may not get around to it. Hopefully I’ve worded things clearly enough for the text to suffice on its own. Versions: I’ve been using Max Design 2013, VRay 3.4, Ozone 2015, and Forest Pack Pro 5 ------------------------------------------------------------------------------------ 1) Global Ambience does not appear to shadow properly: e-on’s documentation claims it is more sophisticated than standard and will render only slightly slower. In my tests with VRay, however, GA renders noticeably faster than Standard. However, it creates unrealistic “anti-shadows”. Where you would expect to see occlusion shading and soft shadows - where objects rest on the groundplane - it actually gets brighter. This is more pronounced with VRayMtl than Standard. This issue is apparently isolated from the Sunlight peculiarities discussed below. Testing with different ground geometries also suggests it is not due to the VRay Plane issue discussed here. I’m not sure what is causing this but it appears light from the sky will illuminate geometry and create GI bounces of light but will also pass right through opaque materials, effectively multiplying intensity while failing to shadow areas that should be dark. 2) VRay Sun + Ozone for Hybrid Daylight systems If you were to compare exclusive use of the VRay daylight system to the Ozone model, each has a discrete sun/direct and sky/scattered element to them. VRay’s is the sun object and a sky environment map (Assuming you aren’t using an HDRi dome light). Ozone’s is a standard direct light, an atmosphere effect, cloud objects, and some other magic all moderated by the ‘upstream’ plugin style that understandably owes this detachment to its Vue DNA. Early tests got me the fancy skies that probably attracted you to Ozone but the lighting in the scene was off or unusable. Washed out light and AO, and a difficulty getting strong cast shadows are among some of the more common ones. I understand this can be an issue when lighting with HDRI’s on dome light too and a VRay Sun is often helpful in getting your direct light and cast shadows back. The sun created by Ozone is a standard directional light, even with a photometric spectral model, is very weak with a low multiplier. This multiplier can be turned up but in many of my tests, ozone resets it to a low value when you hit the render button. I have otherwise had little luck with Ozone’s internal settings getting the relative intensity to my liking. The result is an absence or weakness in direct illumination and shadows from the apparently bright environmentally mapped sun. I initially found it lighting my scene adequately but with no palpable direct lighting or cast shadows, even with a dramatically low sun and no atmosphere/fog/clouds to scatter the light. I found that using a VRay Sun in conjunction with Ozone’s atmosphere to be a good solution but it took some duct tape and workarounds to make it functional. The short of it is that the best results I have found is to use the VRay sun as your sun object and Ozone as your environment + sky model, matching their positions, colours and intensities in a semi-manual way, detailed below. 3) You might have to adjust sun brightness or physical scale to get decent photometric spectral results on reasonable physical camera settings. Photometric spectral is much brighter than standard spectral. When using VRay exposure controls and cameras, your exposures will have to admit less light than with standard spectral atmospheres. Relative to the Ozone sky, your VRay Sun will be brighter in standard than in photometric, affecting the relative intensity of your direct-light and shadows vs. ambient. Adjusting multipliers or physical scale can rebalance this. I shoot for 0.25-0.5 on the sun lets me keep using realistic physical camera exposure settings. 4) VRAySky Environment Map + Ozone work additively. Don’t use the Sky Map: The VRaySky environment map will work beside Ozone but will not respond to Ozone, which can be expected. While they don’t create errors or crashes, they both act as if they are the sole atmosphere/sky system and will stack light contributions. VRay will want to create a Sky Environment Map whenever you make a VRay Sun (which you will still want to). If you know you want both environment/atmospheres adding light that way, gopher it. Otherwise, I tend to deactivate the environment map so I’m only worrying about one environmental contribution to the scene. 5) Make the VRay Sun invisible: This keeps its direct light contribution to the scene but it seems that even a time/location match produces slight differences in the position of each system’s sun. This can produce two coronas in the background and/or reflection. Even if matched positionally on a clear sky then things should be fine but this might be tricky to do. Moreover, if the Ozone sun is behind atmospheric effects like clouds or fog, your VRay corona will present an unscreened corona and compromise the aesthetic. In said cases, the VRay Sun’s corona may look unrealistically bright. You may also want to dull its brightness as it is not apparently sensitive to these atmospheric obstructions. 6) Ozone will (sort-of) track your VRay Sun and match its sun position, but not when used in a 3ds Max daylight system assembly: I believe this is in the e-on documentation but it is worth bearing mention here. If you create a daylight system and set the sun to VRay type, Ozone will not see it. You have to deactivate it (unless you want two sun objects) and create an independent VRay Sun. If you want to govern the position of it using the daylight system rig, you’ll simply have to align and link it to the deactivated system’s sun head. This would be a nice change in future Ozones versions as it is redundant. The (sort-of) is elucidated below. 7) Manually set/animate Ozone’s internal sun to match VRay Sun’s. Know how the VRay Sun object and Sky map adjust their colour/intensity and the effects of atmospheric scattering according to the sun object’s position in the sky? Also, recall how Ozone has a setting to match its sun’s position with the VRAy sun at render time? The Ozone atmosphere and the ambient light in it do not seem to respond to the VRay Sun’s position. Instead, the atmosphere’s contribution appears to refer to the position of the Ozone sun settings in the atmosphere editor even when the Ozone sun is being automatically moved to the VRay sun’s position in the sky at render time. The workaround is pretty simple: set the position of the sun in the atmosphere editor to the same position as the VRay sun so that the Ozone atmosphere is more closely reproducing light at a comparable time of day. If doing a timelapse, your scene will still get brighter and darker throughout due to the VRay Sun’s self-attenuation but the sky’s contribution will remain constant. If your VRay sun is animated via a linked daylight system as detailed above, you will need to keyframe the timelapse again in the Ozone dialog. Make sure you set the same global location (lat/long) too. At this point, the feature where Ozone performs sun-to-sun tracking is made moot; I reckon you could disable it if you wanted. In my first attempt at matching these timelapse animations, for a reason I haven’t been able to figure out yet, Ozone stopped following the set animation even though the preview inside the dialog showed it going through the whole animation properly. 8) Adjust Ozone's FG Contribution: One of the few posts on e-on’s forums that address the washed out light that many users get will advise you to toggle the switch that connects/disconnects Ozone from VRay’s FG solution. I generally want to keep Ozone’s GI contribution on so that its sophisticated phenomena can light the scene so, in addition to using a VRay Sun for direct light, it may help to drop Ozone’s FG contribution multiplier (found near the disconnection toggle) to 0.5 or less. You may still want to do a material override with a VRayDirt material for and AO pass. I still find my scenes, from a variety of e-on’s templates, to be quite blue from said contribution but that colour can be adjusted in your atmosphere settings or in post. 9) VRay Plane and Ozone don’t get along well: If using a VRay infinite plane and a scene where the sun is below the horizon, create a flipped copy that is offset down by a few units. Otherwise the sky will continue to light your object from the bottom when the sun is below the horizon. It seems after doing a render (I’m on Max 2013 Design w- Ozone 2015), Max will fail to identify, select, or even acknowledge in layer browser the existence of a VRay Plane object until the next restart. 10) Spectral models do not work with VRay RT Pretty short point. It’s probably the lighting model. I’ve only experimented with Spectral models but haven’t gotten around to the simpler ones. 11) If you’re not gunning for fancy volumetric effects or flying over the clouds, turn your Ozone sky into an HDRI. This has been discussed in the other thread and I didn’t think about it until reading said thread. Set up an empty scene with a spectral sky and a spherical VRAy Camera. Make a fancy Ozone sky and render it to a radiance or EXR file. You can then bypass the plug-in for most purposes and render it using a standard dome light. Honestly I foresee this being my primary use for Ozone, given the clunkiness of the upstream atmosphere editor vs. the simplicity of the dome light, not to mention the render-time differences. You can create an endless number of skies for re-use in other scenes and save a couple hundred dollars on purchased HDRI’s, not to mention the fact that you can animate daylight and changing cloud patterns to HDR sequences. -------------------------------------------------------------------- Hope someone finds that useful, I spent a good chunk of 2 days working on it instead of my Master's degree. I welcome your contributions, critiques, and corrections Cheers, Riley
  4. Agreed, though I guess I can't blame them for putting their resources towards the vendor who reciprocates. Nice choice of speakers by the way. Rocking a pair myself.
  5. Ah, all good. Thanks for being willing to do so either way. So AMD's not simply holding real functionality to pro cards...Good news I suppose. The image is appearing broken for me. Is it just a photo of gpus?
  6. Just a few things to toss in here. GPU Rendering: In addition to what Dmitris has mentioned, I've been reporting my experience with trying to get VRAY RT to work on my Tahiti cards here and on the Chaos Group forums. The short of it is they’re not usable. The long story is they finally work again after being broken for so long but they’re still gross under-performers; Apparently even old HD58xx hardware outperforms the 79xx’s so something is up at the software/driver/API levels. Vlado and Chaos Group say the OpenCL kernel is scratch built rather than a port of the CUDA version and they insist AMD’s drivers are to blame. AMD has yet to respond to my inquiries/complaints. I’m inclined to believe Chaos Group but it is apparent that Nvidia has simply been more active in working with them. I'm not a 3.0 Beta tester but I don't expect it to employ all of this wasted horsepower. I’ve asked a few review sites to look into trying APU models, 290’s, S10000s, and W9000s to see if the issue is just Tahiti drivers but so far nothing’s come of it. AMD, or anyone else, do not seem to be pushing hard enough to get OpenCL on a solid rendering platform. Next Limit and Pixar have not visibly committed to GPU applications and few people use Luxmark or RatGPU outside of benchmarks as far as I know. Thinking about what happened to the Brazil/Caustic team’s efforts just makes me angry. Nvidia is certainly doing their part to stay on top in both a partner-support capacity and straight-up owning one of the most prevalent renderers; I expect Mental Ray/iRay will never be OpenCL, effectively locked to Nvidia hardware. It looks like Octane, Thea, and finalRender 4’s implementation are CUDA-only as well. Just read this article and ctrl+f : ”opencl”. http://www.fxguide.com/featured/the-state-of-rendering-part-2/ The find count is sad pandas. While AMD’s been very keen to show off OpenCL/HAS advantages with physics or stereoscopy in Maya, I think rendering is what we professionals are watching and Nvidia has it cornered. I honestly think catching up in Vray RT is AMD’s best bet in turning it around. Prices: High end AMD cards (79xx, 280X, 290, 290X) are all selling well above MSRP and are globally difficult to find due to cryptocurrency miners snapping up more high-end cards per system than CG pros or gamers put together. NVidia cards, known to be poorer performers in mining, are well in stock and easier to find good deals on. AMD must be both laughing and struggling to keep up with demand. I’m late to the mining party but I’m glad I finally have a good use for my two 7950’s. They’re great for folding proteins with the new core_17 client but again, I don’t expect them do be great gpu rendering cards for awhile; it’s a CUDA game. Dmitris: Regarding my Firepro driver curiosity above, have you tested your W5000 on VRay RT by chance? Any better results than consumer cards? Do Intel iGPUs at least perform proportionately to their feeble potential (to shed light on the CUDA-bias vs. AMD’s driver neglect debate).
  7. I may be comparing apples to oranges here, and I do like the variety of structures too, but as a part-time freelancer I'm hoping they do the "Good-guy Pixologic" thing with VRay 3's price point. ZBrush certainly feels like it has the lion's share of the digital sculpting sector, perhaps even moreso than VRay in the rendering sector. Between my initial academic license and subsequent commercial upgrade I don't think I've spent more than a grand on it, yet I've been pleasantly surprised that every update since buying into version 3 has been free. I certainly wouldn't demand/expect anything free from anyone and would gladly pay, say... 500-800 for a ZBrush or VRay upgrade, but Pixologic's practice suggests to me that a company on top of their market can keep software affordable/accessible and still be profitable on a volume of sales. I think I'd stay on my version 2.0 license if having those nodes in 3.0 cost 2k. All due respect to the R&D expense that goes into these things but I really hope prices don't go the way they did with realflow.
  8. Aye, I think you'll almost always do your production renders in Adv. I expect that those familiar with the supported features of RT are using it to rapidly test the general look of material, lighting, and compositional changes, speeding up the fine-tuning stages of a scene/animation. One would be able to set up and build their scenes such that they can switch on/off the unsupported features when they come to need this fast draft-turnover. Multiple Titans, afaik, are the ideal cards on an infinite budget for this application. In short, I'd say that technically RT in the right flow can save you human hours on the production of a scene but I believe most would stick with farms and Adv for production render.
  9. Hi Allan, Probably best to go with the GTX. As much as I like AMD's product otherwise, they haven't really made any concerted efforts to resolve this and, from what I now hear from others, VRAY RT was originally written with a CUDA codebase that seems to affect even the OpenCL client. While the HD7000 series is a beast when it comes to compute, I'm beginning to think we'll never see it run RT. Perhaps RT 3.0 will sort this out. Riley
  10. Hi Arqing. You're right, CUDA works only for nVIDIA hardware and AMD cards trounce nVidia at OpenCL processing (excepting maybe Titan). I'm trying to get my 7950's working on the RT OpenCL client which, like you said, should run well. The problem is that they hardly run at all; it's not really the cards' raw performance that's the issue but a program-to-metal miscommunication that's causing poor times in the cases when it runs at all. There's another thread going where I've spoken to vlado on the issue. I've only gotten any RT OpenCL scene to work on the VRAY and Catalyst release combination listed above before AMD's next driver release fudged any OpenCL rendering at all. It hasn't worked since. I'm sure it'll run great when the program and drivers are speaking properly to each other, but I suspect it's not in AMD's list of priorities; they're too busy chasing fps in Crysis 3. It's disappointing having all of this unused horsepower.
  11. yow, that's somethin'. They're really hitting it out of the park with these promotions. Even got me to pull the trigger and pick up a second 7950 with the Never Settle: Reloaded bundle. Bioshock: Infinite is delicious btw. I might check into this one later this year if I come across the cash.
  12. Here's an update I posted on chaos group's forums a few mins ago. Edit: Just tried with Max 2013 and the Cat13.12 Beta 6 drivers. Same issue as 2011. -------------------------------------------------- Small update on my experiences here. While the performance is still poor in general on RT 2.30, I can say it scales well with multiple GPUs. I grabbed a second 7950 and got a 7m, 40s time on both uncrossfired cards, better than linear from a 16m time on one card. Unfortunately I can't get the OpenCL client on RT 2.40.03 to work at all again. The RT server seems to be stuck sorting the scene out (In the "waiting for image data" stage of the activeshade window). I know this typically takes a few minutes the first time around for a scene in a working setup but I tried letting it think overnight until the VRAY standalone process crashed. The CPU client works fine, as expected. As an aside, it seems AMD crippled folding performance with Cat 13.1 as well. GPU utilization is down to ~40% on the 7000 series for most people, though some have managed to get it back to high 90%'s after exhaustive driver cleaning and custom SDK/beta driver combinations. I know more of their market is gamers but it's still a shame they're throwing away potential compute functionality to chase fps in Crysis 3. Maybe I should've held onto my dollars for a single titan card >_
  13. Hmm. Maybe it's a Max Design 2011 thing. I'll be picking up 2013 soon though so I'll give it another whirl then. Thanks for going through that, shadowzone! Cheers!
  14. That's alright. The benchmark's not intended to output a clean final image. The results are strange indeed, though. You are using a retail 7950, yes? Which version of Max? because my results are yielding three times that turnover time.
  15. http://www.chaosgroup.com/forums/vbulletin/showthread.php?52400-test-scene The posts are arranged in reverse chronological order compared to ones like this, so the newest posts are on the first page there. Vlado's posted the scene on either the earliest or second-earliest pages. It's set up already for you to hit activeshade when you open it and will halt/finish once 512 rays/pixel have been cast. The time it takes to hit that sampling rate is your benchmark score, the quicker the better.
×
×
  • Create New...