The 1.5 sun/sky camera model vs. the legacy

[davision]

Vlado,

Is there any way to quantify the difference between the old sun/sky and camera model and the new one? So that we can adjust our old scenes easily?

For instance, can we say that if we decrease the Physical Camera's F-stop by a factor of 1.4 (or whatever) it will bring a similar degree of lightness. Or you can turn down the sun by a certain multiplier? Etc.

I have scenes that were setup very carefully -- and it would be nice if there was some direct way to convert the sun and cameras (especially the cameras) so they render similarly in the new model.

I realize that the interrelationships might be too complex and that the changes might not be explicable simply as multipliers... but I thought I'd ask anyway.

[^Lele^]

it seems a simple test to do, to me.
One i haven't done yet, mind you, as the new models for sun, sky and camera work a lot better than the previous ones.
Render a picture with the legacy models on, colorpick a value (or more).
Do it again with the new models and colorpick again.
Better still, render a spherical render, bring it in hdrshop, and ask for the average color value, for both methods, and you'll know.

EDIT: the new sun/sky model includes also a max photometric scale adherence, so yes, it is a bit more complex than just ticking a box. in fact the old sky with max photometric scale checked results very very dark.
If i render both models with that box unchecked, the difference is abyssal.
To be precise, though, the new model should have it checked, the old one unchecked. the results is like this for a non-exposed spherical render of the sky with the sun at the Zenith (ideal midday):

old model pixel average: 34467, 34743, 30160

new model pixel average (do NOT obey max's photometric scale): 3161611, 3182135, 2761404

new model pixel average (obey max's photometric scale): 6592, 6664, 5779

The values you read there are FLOAT, for R, G, B.

[rcollett]

I too have been too busy to do any tests with the new model. Is the only difference in light values, or is there any rendering speed differences?

[Dynedain]

As I understand, it's a fix to an error in the underlying ISO calculations, so there shouldn't be a change in rendering speed, but there will be a change in the visual exposure results.

[vlado]

Dynedain is right, there is no difference in speed, just in the relative intensity of the sun, sky and the camera exposures.

Best regards,
Vlado