Another vote for thin translucency being MUCH more important to me and honestly I thought it was under active development since Principled v2 BSDF is–what happened?
This is why I continually push about visibility/linkage/transparency between feature requests and roadmaps. What was the data used to make these planning decisions when Thin Translucency has been known as a missing feature for years? Where do working artists go to indicate what features are most useful to them? Do the devs care?
I know this isn’t strictly the case, but it feels sometimes like blender devs give only the barest nod to user feedback when planning–and I know it’s not just me with that perception. Eevee-Next and its blowback was notable because of its prominence but the pattern has been there for ages. None of the artists I know want anything to do with blender development because of the perception that their input or even contributions will be discarded.
Oh, i am not so sure about that. Matching real lens distortions in final renders for animations that are based on camera tracked footage is like the bread and butter of a substantial portion of the VFX industry. For me it is the upcoming feature i look forward to the most.
Well, for you maybe. Sure, there will always be some specialists who do very unusual thing on daily basis. I was talking about users on average, not a specific user.
If users on average needed OSL scriptable camera deformation (I do not mean few additional parameters in a camera object, I really mean actual OSL scripts) more than user friendly way of expressing thin translucent materials, then it would not be case that pretty much every mainstream renderer out there got thin translucency in the main renderer uber-material much sooner than OSL camera lens scripting.
In fact V-Ray, probably the most widely used renderer, and likely the renderer used for most VFX shots in history, still doesn’t have OSL for camera lens scripting to this day, just for materials and textures, while being one of the oldest ones around.
V-ray’s physical camera has many conveniences that cover this specific need. You can even control the barrel distortion with an STMap at render time so there is no need for the trickery we do in Blender with overscans and the lens distortion nodes in compositing.
I think that by focusing on the OSL scriptable deformation to make your argument is kind of unfair as this is only a part of the whole OSL lens project. If it was just that, i would agree that Translucency would be higher in the to-do list.
One other think on my mind is that the developers that work on the OSL lens implementation is not necessarily from the same pool that would solve the Translucency in Principled v2. So one is not against the other.
I did not focus on OSL scriptable deformation. @etti did, stating it as at least one of the reason that a feature originally scoped for Principled v2 is delayed.
Yes, V-Ray physical camera has many convenience features to cover that specific need. That does not matter. What matters is that even these convenience features were added significantly later than user friendly means of expressing thin translucent surfaces in the combined material.
I think you are missing a point. This is not about whether one or other feature has more priority for some specific people. This is about whether one or other feature has more priority for users, overall, on average. There, it’s just undeniable because I am having a hard time to thinking of one mainstream renderer which tackled advanced lens distortions sooner than having a complete basic feature set in their uber-material.
And it was not only this alone but the frustrating fact that thin translucency was promised as part of Principled v2, then dropped for no apparent reason, and apparently won’t be resumed anytime soon.
Significant majority of the PRs listed in the Principled V2 task are by Lukas, with a few of them by Brecht - that’d make it 2, not dozens. In the OSL Cameras tasks, all commits are by Lukas also.
Non the less, why are you replying to my post though? I wasn’t the one who implied the reason Principled V2 is in a limbo and essential functionality thin translucency is still not in it is because of OSL Cameras…
If you review the entire conversation above, you will realize that all I did is ask what happened that the thin translucency mode is still not in after all the years while at the same time marginal improvements to much more niche things like alternative BSDFs or nested dielectrics were still being worked on. That was it. I was just echoing the sentiment of the post by Lolwel21 above mine, adding a bit of context to show how clunky the thin translucency setup is at the moment because of this missing feature.
The whole thin translucency vs OSL Cameras conversation was not started by me. I did not make the claim that the development of one is at the expense of other. In fact that claim doesn’t make sense to me.
I agree that the converstation on that subject is overexhausted and maybe counter productive. I removed my previous message and will not add to this anything more.
