Cheatsheet: How to get a Cleaner image with SP3

[Micha]

I think it's a question of the use - pro or hobby. From a pros view - my setup works stable and is improved over the time. General I have no problem to get a render setup done. My problem is to get much renderings done within tight deadlines. And here it doesn't help to get more automatic features, but less speed.

So, the question stay open - doe's the new workflow improve speed at same quality or not? (LC+IM speed) My impression is a lot of the new core power is lost for the automatic and the missing speed shortcuts. I don't talk about animation, I mean high res interiors. (Also I don't get my old animation setups running at the old speed.)

[^Lele^]

I think it's a question of the use - pro or hobby. From a pros view - my setup works stable and is improved over the time. General I have no problem to get a render setup done. My problem is to get much renderings done within tight deadlines. And here it doesn't help to get more automatic features, but less speed.

So, the question stay open - doe's the new workflow improve speed at same quality or not? (LC+IM speed) My impression is a lot of the new core power is lost for the automatic and the missing speed shortcuts. I don't talk about animation, I mean high res interiors. (Also I don't get my old animation setups running at the old speed.)

Micha, you are asking questions in the wrong forum.
In Max, nothing is lost, and all the old methods are available to those which need them, although we *strongly* advise against it.
V4R has its own section, and that's where you should direct questions about its UI and specific implementations.
As explained in these sections, however, post 3.3 will *not* map 1:1 with pre 3.3 scenes, as the sampler is entirely different, and so what the UI controls mean changes as well (defaults should reflect this, and that's where you ought to start from. even in v4r.).
It's simpler, stabler, quicker, and overall better, but it will require some mental agility and willingness to adapt from those users which were set in the pre 3.3 ways.

Either way, the debate ran for pages (and a little over a year), forum search and attentive post reading is your best friend.

[Micha]

I leaved the Rhino section because this it's not so easy to get answer. And I was glad to find my question in this thread, but the answer was missing too.

I will take with me: "It goes without saying that if you're rendering super simple, super even geometries and shaders, there is likely no beating an IRMap in those large, largely uneventful, areas."
And if I understand all the posts, the answer of the question "Is the purpose of the new approach to greatly reduce tweaking only? Or does it always provide more speedy noise-free renders too?" is, if an interior was fast, than it will not be faster per BF-LC. If an interior was tricky, than jumping to BF-LC could help.
Great to hear that the "advanced control" is built into the 3.3+ approaches of 3dsmax.

The Rhino forum is a third class forum only. You can post hundred bugs and questions, but it's not sure to get an answer. The plugin seems to be mainly developed for an other host application and Rhino is at last place now. Now the situation is that advanced options are wiped out and only if you had luck than you get some back to keep the old quality. So a big question for me was, is there the same situation on 3dsmax, how is this handled at 3dsmax. It looks like your render world is fine yet, no big cut downs, you are free to use your old workflow or the new way dependent from your project.

I'm using Vray for Rhino 12 years nearly every day to live from a good running render biz. The current situation is a big problem for me.

[sharktacos]

I'm curious if Lele and the Chaos gang have done any benchmarks on using the Vray denoiser? We've generally found that setting the AA threshold to 0.05 works well (i.e. the denoiser can usually get it to look like it was rendered with 0.01), and makes the render times around 4 times faster.

Curious to hear other's experience.

[^Lele^]

I haven't personally followed up after the denoiser creation, with benchmarks, but what you state seems about right compared to what i was experiencing back then.
I did all my early tests with NT 0.1, they'd look like 0.01 (if i cross-frame denoised.), and the denoiser would even reconstruct texture detail lost in a frame, in that case.