Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - Ryuu

Pages: 1 2 [3] 4 5 ... 45
Corona itself does not use any AVX at all. The highest we use is SSE4. We're using the F16C instruction set when available, which can technically be considered a part of AVX and TBH I have no idea if just using this might cause the CPU to switch to the AVX clock or not. But we use these instructions only during rendering, not when computing denoising.

We do however have some dependencies which use AVX:
  • Embree uses AVX and AVX2 when available. We can turn this off if needed.
  • The runtime library of the compiler we're using uses AVX(2) for the implementation of some functionality. This is something that is completely outside of our control.

That being said, like Juraj already noted, no AMD CPU at this point supports AVX-512 and we cannot execute instructions which are not supported by the CPU. So even if we had some AVX-512 code, all that could ever cause on an AMD CPU would be to immediately crash Corona due to executing an unknown/unsupported instruction.

The denoising is probably the only part of Corona which is not limited by memory accesses and therefore really utilizes the CPU as much as possible - from our CPU FAQ:

Why is my CPU temperature higher/fan louder during the denoising compared to the rendering?
The memory access pattern of the rendering code is very random - for example, each ray may hit a completely different object. This makes it very hard for the CPU to predict which parts of the memory will be accessed next, and in the end the CPU ends up waiting for data to be fetched from memory most of the time.
When the image is denoised, it is accessed sequentially one pixel at a time. Also the processing of each pixel takes long enough so that the CPU is able to read the next pixel from memory before it is needed. This is a best case scenario for the CPU - it does not have to wait for memory, and the execution units are (almost) fully utilized.
When the execution units sit idle waiting for memory, the CPU is able to optimize its energy consumption (and therefore heat generation), but when the denoising kicks in, the CPU units are kept busy all the time and so the temperature goes up.

So the most probably culprit is just an unstable overclocked configuration, which runs just fine until the CPU is under heavy load.

Hi, we have some problems with parsing time in the few last daily builds. It's a nice coincidence, because I'm just looking into this code right now :)

Anyway, it would be really great if you could share some of the problematic scenes with us, so that we can look into those and make sure that when we fix this issue that it's fixed for your case as well.

Any info on aerial perspective, and "altitude" in the new sky?

Altitude + turbidity should be ready soon(ish). Unfortunately it's currently a bit delayed because of SIGGRAPH and the fact that our researcher working on this is still just a human being and needs to sleep at least 2 hours a day :) Our current estimate is that we should have it ready in daily builds sometime in March. We definitely plan to have it available in v7.

Aerial perspective is scheduled for later. Probably more like v9 than v8.

I see that the new PBR material was marked as complete on the Trello board on the 19/01 but the latest daily build is 08/01 - will there be an updated daily build any time soon with the completed PBR material?

There are no big changes in PBR since the last daily build, just some very small UI tweaking. Sometimes it just takes us a few days to remember to update the Trello board, unfortunately. However this does not mean that there can be no more changes to the PBR material before the release. We're still gathering & processing feedback to see how can we improve it. About next daily build, with some luck by the end of this week, otherwise sometime next week.

Was wondering if Tyflow instancing / motionblur was still being considered?
It's not in our immediate roadmap for v7. But I can see that we're talking with the tyflow developers about this and from what I can see it may be relatively easy to add, so it may still happen for v7 (no promises though).

Hi, I'm not really sure if we should treat this as a feature request or as a bug report :)

The thing you're asking for is definitely possible, but TBH this feels like asking for a workaround for something that ideally should just work. Would you be willing to share with us some of the problematic scenes you mention? We could take a look at those and see if we could tweak the adaptivity/noise level calculation based on this.

Note that all changes in this build will be released in v6 hotfix, so this is more of a hotfix preview than actual v7 daily build.
  • Fixed crash when Falloff Map with fresnel type and world - z axis direction is applied to refraction
  • Fixed DR being stuck when rendering second frame of an animation
  • Fixed the issue where having Norwegian locale lead to 3ds Max crash at startup.
  • Fixed: Scatter missed update in IR causing "popping" while scrubbing timeline
  • Fixed LEGION_STOP in scene with falloff + RGB multiply maps combo
  • Reverted behaviour of the no tiling mode in CoronaBitmap which was causing unwanted results when used as a mask

This post is reserved to be used for compiling a full "all changes since v6" changelog as v7 daily builds get released.

  • Fixed the “unsupported parameter used” warning sometimes name of different map than the one which triggered the warning (when the experimental “use standalone maps” option is used)
  • Fixed loading of .acv files in curve editor
  • Fix of crash during changing time on timeline due to double initialization of some materials
  • Fixed Standalone incorrectly interpreting paths containing spaces
  • Fixed NaNs in certain cases
  • Material library
    • Fixed the “incorrect normals gamma” warning in material Concrete Plates Dirty
    • Fixed missing bump map in plaster materials

  • Fixed wrongly greyed out “next” button in licensing dialog.
  • Fixed 4K cache not distributing precompute samples evenly across the rendered image.
  • Scene parsing time is now accessible through MAXScript (CoronaRenderer.CoronaFp.getStatistic(14))
  • The warning about incorrect normals is no longer shown when the map is disabled
  • Fix of random crash with the ColorCorrect curves editor in linear interpolation mode
  • Fixed Corona Scatter to trigger recomputation on scale map change.
  • Renamed inputs of some Corona maps that are shown in the slate material editor, to make them overall more consistent (e.g. texmap vs map vs color)
  • Fixed different material editor preview of Corona bitmap and Max bitmap when the map channel is other than 1 (now the previews are the same)
  • Material preview functionality added to the new version of Corona Standalone (use --material-preview-help on command line to get more details)

  • 4K cache was marked as experimental and UHD cache was reverted as default secondary solver
  • Fixed 'Now' update button in Corona Lister listing scatters.
  • Fixed Corona Scatter wrongly scaling instances displayed in the viewport after clearing the scale map slot.
  • Improved user experience in Corona Scatter scenes using high-poly distribute-on objects. It is faster now.
  • Fixed maxscript function getStatistic returning slightly different value for rays/s from what we display in VFB stats.
  • Fixed masks being generated incorrectly for objects with opacity less than 0.5 when mask propagation is disabled
  • Added missing support of SiNi SiClone sweeps
  • Fixed occasional crash during parsing when scene contains mesher compound object
  • Fixed distance map not working in displacement modifier

the MatrialLibrary downlodable in Dropbox is updated with the new materials?
I download it but seems to be the old one.


It is now. I forgot to update it yesterday, sorry for that.

  • Fixed crash when exporting 3dsmax bitmaps to standalone
  • Fixed crash in some cases where motion blur is used with changing mesh topology
  • Standalone export:
    • Fixed crash when exporting more than 40M triangles
    • Old displacement is forced on export since standalone does not support the new grid displacement yet
  • Improved error detection when loading cgeo meshes - both for corona proxies and loading exported geometry in standalone
  • Memory statistics are now available through maxscript
  • Fixed differences in rendering on different CPUs (mostly Intel vs AMD)
  • Added clamp tiling mode to Corona Bitmap
  • Added new materials to the Corona material library
    • 13 new concrete materials
    • 31 new ceramic tile materials

No global toggle for real world scale has been planned for v6. I guess this may have been mentioned as "something we'd like to do in the future" along with some other updates which were planned for v6, leading to some confusion.

There is another update for the material coming before the final v6 release, containing new ceramic tiles & concrete materials. This update will be released with the next RC.

Thanks for the feedback. I'll forward it to our material library guys.

It would also help debugging a bit more if you could also provide few more details about how this happened. Was the texture loaded as a Corona Bitmap or as a 3ds Max Bitmap? Were you running IR or any other rendering at that time? Thanks

  • Added possibility to add secondary GI precomputation time into VFB render stamp - with %pts
  • When the environment is visible through a reflection or refraction during mask propagation, the result is now a black color instead of the mask of the reflecting/refracting object.
  • Fixed crash when saving scene with Corona Scatter switched to point cloud display mode.
  • Fixed freeze when loading some Corona Scatter scenes.
  • Fixed material library layout when displaying materials with long names
  • Fixed crash when modifying material in the material editor while IR is running when “Enable material editor” option in the IR section of system settings is disabled.
  • Fixed issues appearing in light material (exclusion and black enviro after hidden in mask works for light material now)

Hi, yep, we have update to the new Ornatrix API in the pipeline. Unfortunately we did not manage to fully finish it for RC1, but it should be in RC2.

Pages: 1 2 [3] 4 5 ... 45