camera subdivisions for DOF?

[darloman]

Hi,
Was this setting removed? What is the current/best way of controlling DOF subdivisions?
thank you
Pawel

[Jussing]

They've been removed along with motion blur subdivision settings, and you're supposed to achieve the same result by tweaking AA and color threshold values instead.

Check this thread.
And this thread in the wish list.

I've started my own personal crusade to bring back the Vray 2.5 settings, as I think the new workflow is gibberish.

- Jonas

[vlado]

...аs I think the new workflow is gibberish.

<shrug> That's your opinion, which I don't necessarily share.

Best regards,
Vlado

[vlado]

What is the current/best way of controlling DOF subdivisions?

You need to increase the max. AA subdivs.

Best regards,
Vlado

[Neilg]

I've started my own personal crusade to bring back the Vray 2.5 settings, as I think the new workflow is gibberish.

Before vray 3, when you switched on mb or dof it overrode your AA settings and they didnt do anything. now instead of having 3 different sets of the same AA settings depending what you've got turned on or off, they're in one place.

[Jussing]

Before vray 3, when you switched on mb or dof it overrode your AA settings and they didnt do anything. now instead of having 3 different sets of the same AA settings depending what you've got turned on or off, they're in one place.

Why is this good?

You still need different AA settings depending on what you have turned on and off, right? If you only have one AA setting for everything, you need to change the AA's all the time.

[Neilg]

because it's just aa. most people don't render animations with motion blur and then high res still in the same scene.

[Jussing]

because it's just aa. most people don't render animations with motion blur and then high res still in the same scene.

Yes, render/shading/lighting artists do that all the time.

Plus, it's not just AA - based on the feedback I've been getting from Chaos Group support and developers, it's some arbitrary combination of four parameters (image filter, color threshold, minimum samples and maximum samples) - as opposed to the old single parameter "subdivisions", which you just increased until satisfied.

To replace one easy parameter with four parameters that you will need developer's knowledge to fully understand the combined effect of, to get the same result, is, strictly objectively speaking, a step back. There's nothing positive about it unless it does miracles for the render times, which I haven't seen yet, but maybe there's a bug involved...?