The point of Multiscatter is to preserve more energy and look brighter. Yes, it's supposed to be different

vertex weights from files saved in 3.0 are not compatible with older versions, 2.93 gives you a warning of loss of data.

It might also be nice to specify an Origin based on a point of your choosing

I wouldn't mind extra side separation the deeper in subpanels you go, even if space is wasted, so it's much more legible

Please keep Burley, Random walk in it's current state is very finicky and not reliable for all looks. The color doesn't converge to the correct one at very small radii, it's hard to exclude parts from affecting others except by separating meshes as different objects (other render engines let you exclude parts from each other in the shader), the sharp edges are not always wanted, and so on and so forth.

Linear interpolation is more descriptive, while solving ambiguities that "normal" and "mix" cause

The exact twist doesn't matter, the intention of the curve example isn't saying "it should be this exact twist", but to showcase the other two problems. The uneven twist/deformation that flattens the shape at places, and the whole body twist changing by length, the former being the worst problem of them two.

Bump

The problem lies more in these things not generating and undo item in the undo history, and so they are grouped together with the next undoable action like the stroke.

Yes, this should also help better with performance since separating everything in render layers or multiple renders just to render the SAME geometry is an overkill https://devtalk.blender.org/t/why-is-bvh-built-for-each-renderlayer/14670

Here's my proposal on how to improve the performance as well as brush behavior, which was already bad before 2.9. https://blender.community/c/rightclickselect/hncbbc/ Basically, it's about having the option for the brush falloff to be just a sphere instead of a projected 2d tube, in line with the other brushes. The projected tube paints all accros, and this forces the user to use Occlude, which is what kills the performance.

In D8687#214523, @Luciano Muñoz Sessarego (looch) wrote:

Which is awful.
Can we assign just first one or last one of the curve?
this behaviour in the 3d vieewport and in the graph editor of having objects be the active one when nothing is selected or selecting a lot of them and not having an active one, is incredibly confusing and its a bug imo.

Currently it's only possible to view the scene's active camera.

In T74581#922698, @Russell (russ1642) wrote:

Without digging into the code it seems that adaptive sampling counts up to your render samples setting but won't change pixels if they're within your noise threshhold. It doesn't stop rendering when each pixel in the tile reaches the noise threshhold, it must keep running noise checks or something on every pixel for every sample.
Try this: With the default cube set the noise threshold to 0.1, minimum samples to 200, and the render samples to 10,000,000. It'll render the noisy tile and then spend minutes not changing the image at all but counting up to the 10,000,000 samples for no reason. Even for tiles that are entirely background. It seems this algorithm doesn't actually stop when the noise threshhold is reached.

Edit: This is for GPU rendering. CPU rendering doesn't do this.

If post processed motion blur is used to blur between ground truth motion blur steps, then both problems are caused by the same issue, GTMB not working, so yeah

I agree with the option. When I thought about it, and saw that using mb steps as the AA method gave the same results on the static parts of the image, I too was worried about the moving parts. Would they be too artifacty? But on my results, given the individual contribution of each sample is so faint, the alias of an individual moving step is not noticeable, but I guess it will be more noticeable with very small, very bright parts, which is the only case I can think of, that's why I'd prefer my suggestion as default. This also opens the door to a larger argument about customizing samples for other things that now are tied together to the one and only render samples, like soft shadows or alpha hashed materials.

A big current hinderance for usability is performance as outlined here, https://blender.community/c/rightclickselect/hncbbc/ but the fix could be potentially easy

Blender by default zooms to an invisible pivot with a fixed that gets a while to get used to know how to control, with focusing and such.

Hi. I'd love it if you guys check out my small proposals about UV editor UI, both garnered some attention on rightclickselect

Maybe I'm wrong but I think it's based on OpenSubdiv, which in itself is another standard, so pick your poison. But I agree, I think it should come back as an option, particularly if it gives old performance back

Could another bone, or another bone in another object be used? Then calling it custom would make sense.

Judging by the image of the list collapsed, my first instinct is that it doubles as a menu. In that case I don't think it's too much clicking if you use a click and drag, the advantages and niceness would outweight.

This only solves the cycling, but not what bone is supposed to be selected first for both pose mode and edit mode. If a bone is inside another and your mouse is over the smaller bone inside, the old behavour would be for the smaller bone inside to be selected first. This new behaviour makes it harder in some rigging scenarios where the order of active or selected bones matter.

Wait, so this is a different algorithm than the current unsubdivide function? Then, please, add it as its own command in edit mode and in the decimate modifier where the current Unsubdivide is. It seems to handle more/better cases. The old algorithm is still useful, though, for those mid-unsubdivisions that make diagonal faces.

Comparing it to other aspects of blender isn't fair. Compare it to every other software that does baking and realize that they don't do it to fit a structure of a software, they bake multiple objects and treat them as jobs because it's what's practical. It's what works. Look at any videogame and the level of detail in their texture work. These kinds of workflow demand it. It's best to make it right to begin with or the baking system will be replaced a third time

really cool. for what it's worth you've convinced me. it wouldn't be the first time a feature is "replicated" when in reality they have unique strenghts (displacement mod vs displacement shader, wireframe mod vs wireframe shader, ao shader vs ao passes, hell, even cycles vs eevee) i can already imagine making things with this that aren't strictly cartoon lines.

here's what i found on the wiki about the possible future of simulations.

Any plans for stereo 360? My understanding is that it is not as simple as having two, it involves translation to get correct results, which Cycles handles perfectly. But real time engines like unreal have to make a lot of vertical splices and move the camera on each rotation. Even if not realtime, it would be nice for Eevee

Any progress on this? It would be great to have, I neede this in production a month ago.

Any plans for multiproperties editing for every data type on the Sidebars of all the editors and modes? Nodes, video strips, track points, mask splines and points, Greasepencil layers, Fcurves and keyframes, NLA tracks and strips, plus the modifiers that some of them have, etc

As a way of storing data for animated models, Multires modifier has advantages over displacement textures. With displacement textures, you first have to UV map, you have to bake (which there is no support for baking vector displacement maps currently) you have to make sure the texture resolution is proper, sometimes you get artifacts if the mesh vertices changes or don't align, or other kinds of artifacts due to the model deforming and CatmullClark subdivision, then you have to worry about using the modifier vs using the shader node in case you need to see the detail in the viewport, etc etc. Some of these are fixable by the user by learning new practices but some inevitably make things slower.

I can confirm as well, I went through the original Intel documentation and was surprised at how good results they have on their page, and the claim that Denoising Albedo is the best input for the denoiser to have. The current denoiser albedo pass is simply adding the Glossy Color pass to all the others Color passes, but the Glossy Color pass doesn't contain percentage of reflectivity for things with frenel. The intel documentation doesn't specify a correct solution though, it's up to you what passes you choose and what combination, for example, if you want to add reflections to the mix or not, and how to do it, taking just the color or the actual first bounce. My current setup is to add Diffuse Color, SSS color and Translucent color, leaving Glossy color out.

In other software and Blender addons that have tried to tackle this, they use a design called Jobs. In these jobs you specify the passes and objects (could be a collection) that should be baked, to which UV map, as well as cage relationship. Then you may bake one or all jobs. The less you have to set up per object and per material, the better. Then it's up to the user if they need a special case of individual bakes per material or per object. You don't have to design a UI where you select one UV map for each object either, instead, in the Job you select just one UV map index or name, and it's up to the user to make sure the objects don't overlap in UV space, they can always create more jobs.

Not related to this task, but having search on the Preferences would be great too

Sweet! I hope these proposals are considered
https://blender.community/c/rightclickselect/cQdbbc/
https://blender.community/c/rightclickselect/bQdbbc/

Is there a task for to improve the UV editor itself? Right now it still has remnamts of the image editor and it could be improved. I made a couple of proposals about it here and here

My understanding is the same. It's used to blur face corner colors into all the adjacent corners, but it doesn't actually blur like in Weight paint or with the blur brush.

Nice! Unrelated but, shouldn't displacement affect both the specular and clearcoat layer? Technically displacement is a stand in for real geometry displacement so it would affect both, which is what Cycles does btw, unlike Bump to Principled where each reflection has it's own normals for good reasons.

It's not as before. Pressing A once always selects, and pressing Alt+A deselects. You can also double tap A to deselect but it has to be fast. If you want to use the old behaviour from 2.79, go to Preferences, keymap, and click on Select All toggles

It's not per image file but per image datablock, you can have two image datablocks that have the same image file as pointer, but different color spaces. Just click the little number besides the datablock and this will duplicate it.

Then maybe we don't want those tools at all, or speaking for myself, I don't. They sound so specific that they're more fit to be part of an addon than anything else.

Any updates on this?

Considering that I model a lot and haven't encountered the problem again, even when I try to replicate my own instructions, I say this has been solved

Color management isn't applied until you save the image and it's not applied on EXR. You could've saved it as EXR, and applied the color management to that image later, to not have to re-render again

I can't replicate it on blender-2.80-e3c586e262dd-win64. Eevee and Cycles values look the same for me, and color management works as well.

Contact shadows are screen space. They have the typical limitations of the other screen space things like AO and reflections. The thickness of the wall behind doesn't exist as far as screen space is concerned. I barely use them because of that and other limitations with thin object.

I can replicate it in the sense that they happened to me in some frames of an animation but I couldn't pinpoint what angles point of view caused it. All that I can confirm is that they happened with AO+Bent normal + Subsurface. It was force to fix this using the Compositor to clamp the high values, erase the alpha and use the Inpaint filter.