วีดีโอ

Muchas gracias por tus palabras!! Valen muchísimo! Un abrazo ;)

Yay! I'm glad to hear that!

Blender 4.3! :)

I'm not sure! I haven't had that happen before. Sorry that I can't help more!

Thank you!

Eevee!

Muchas gracias! :)

Thank you!

I see your point, but it's not quite what I'm doing. Since I'm using the viewport compositor I can't have separate character and env render layers. What I'm doing is trying to reconstruct that behaviour just using the beauty render. But, as you say, this should be overkill. For any non interactive compositing, people should render different layers and just merge them on top with an alpha over. Good observation! I hope we get support for view layers in the viewport compositor in the future!

@@homspau Ooooh I see. That does make sense, sorry!

@morlankey Thank you for pointing it out!

Replying out of the top of my head, so some of my observations about the spiderverse films might be inaccurate. For characters, I don't recall them doing any motion blur at all. Since they use step set animation (variable rate animation) it wouldn't look good. Instead, they use smears and multiples in certain poses to convey motion. That's more of an animation thing rather than something that should be done in compositing, if you ask me. For backgrounds, I don't precisely remember what they do when the camera moves sharply. I would just add that any compositing based effect for motion blur should be driven by the "vector" renderpass. It will give you the direction and magnitude of the movement in screen space for each pixel. The usual way of applying it is by using a vector blur node, but if you're creative you can use it to drive a displace node (similar to the defocus thing), the size of a kuahara or anything you can come up with. Hope this helps!

Thank you!

I'm so glad to hear that!

Thank you! It truly means a lot coming from you :)

Thank you for the comment! :)

I'm so glad it helps! Thank you!

Hahahaha thank you!

Make sure you're using Eevee in blender 4.3. this workflow doesn't work in earlier versions. It should be fine using the standard compositor instead of the one in the viewport. Hope it helps :)

@ thanks

Thank you! I'm so glad it helps! :)

Quina ilu que serveixi! I més encara llegir català als comentaris ;) Una abraçada!

Thank you!

Hi! I think I covered this on part 1. The viewer node affects only the viewport, backdrop and the "viewer node" "image" in the image editor. For final rendering you should plug your graph to the "composite" node. Hope it helps!

​@@homspau Iya, with a bit more testing I've realised the parameters I dialled into nodes like Kuwahara were so low that they were not having effect on the image / UV editor (via the viewer node). Still puzzling as I could clearly see those effects in the viewport throughout... So I'm curious now whether a threshold was affecting the image somewhere along the pipeline (some kind of noise threashold or denoising). I imagine at somepoint a eurika moment will happen and this will all make sense Anyway increasing the intensity here and there has worked. Thanks again for your work here Pau. I also remember the talk you presented at the BCON at the start of the year.. A very useful tutor indeed!🙏

The reason you see white is because the depth pass actually gives a value to each pixel based on how far an object is from the camera. For example, if an object is 0.5m (or scene units) from the camera, the pixels which depict that object will be mid gray. At a distance of 1m, the RGB values of those pixels will be pure white. But, since our renders have a high dynamic range or, speaking properly, the pixels have a floating point bit depth, these values can go above one. An object thats 2m away will have RGB values of 2 and so on. You need that raw depth data to make accurate math on your depth. For example, for making fake DOF, you could get something working with the mist pass but it would be inaccurate and dependent on the scale of your scene. For creating masks the mist pass is fine, but I encourage you to get familiar with the depth pass too! I hope I'm making myself clear! :)

Thank you for the explanation, it really helped me understand a bit more! Is there any way to ‘clamp’ the range of values from 0-1 so that I can get the range of values for the shot?

@seasoned-pastronaut you can use the map range node! Checking clamp and using the "from min and max" values as you would on the start and end of the mist pass should give you the control you're looking for to create masks

Thank you!

Thank YOU for the comment!

Muchísimas gracias Daniel! Significa muchísimo 🙏 Y muchas felicidades por el tuyo de claypencil. Espectacular!