Author Topic: DBs 1.5: Bloom and Glare threshold and CoronaOutput  (Read 1773 times)

2016-07-25, 20:42:56

Frood

  • Active Users
  • **
  • Posts: 1922
    • View Profile
    • Rakete GmbH
It seems that the "Original RGB "value is used as threshold. Shouldn't it be the "Mapped RGB" value instead?

The point is that CoronaOutput doesn't work as expected (at least from me :) any more:

A ("HUD") plane with light material applied and some bitmap with unchecked "Affected by tone mapping" as texmap gets part of the glare/bloom calculation when using lower exposure values because its "Original RGB" value gets higher, when lowering exposure and this exact value is used to control glare/bloom. So sooner or later it reaches the glare/bloom threshold and gets processed.

As far as i know, CoronaOutput calculates the texmap as the opposite of the post settings to maintain the visible output and thus creates higher original rgb values the lower exposure gets to compensate it. However when used together with glare/bloom and "Origninal RGB" value as threshold, the result is questionable for me.

Is this solvable? I´d like to continue using this method to stamp bitmap info (logo/project status/shot information etc.) directly into the cVFB to minimize post.

Good Luck

Never underestimate the power of a well placed level one spell.

2016-07-26, 10:38:16
Reply #1

Ondra

  • Administrator
  • Active Users
  • *****
  • Posts: 9048
  • Turning coffee to features since 2009
    • View Profile
this should be already fixed in the latest daily, have you tried it?
Rendering is magic.How to get minidumps for crashed/frozen 3ds Max | Sorry for short replies, brief responses = more time to develop Corona ;)

2016-07-26, 14:59:33
Reply #2

Frood

  • Active Users
  • **
  • Posts: 1922
    • View Profile
    • Rakete GmbH
No, I hadn't (it´s DB160715 in the first post) but now I have, and it's acting even stranger:

Attached images sequence:

DB_160721_A.jpg:
This is how it looks without any use of bloom/glare. The result is the same (well almost) as Corona 1.4. Note the mapped values of the preview stamp are below or almost equal 1.0 - this is what is to be expected, all ok.

DB_160721_B.jpg:
Now I do a re-render (cleared cVFB to be sure) with only two changed pararmeters:a threshold of 1.2 to be far away enough from 1.0 what should be the maximum CoronaOutput would deliver, and glare power of 500. Now look at original and mapped RGB values and ... the glare :)

DB_160721_C.jpg:
So when I let the above render, it looks like this when noise limit is reached (changes to this when finished)

DB_160721_D.jpg:
And this is what I finally get when I move the mouse over the cVFB with the rendered image


Good Luck




Never underestimate the power of a well placed level one spell.

2016-07-29, 13:09:23
Reply #3

Frood

  • Active Users
  • **
  • Posts: 1922
    • View Profile
    • Rakete GmbH
Ok, I feel stupid because obviously it seems to be fine for others?

- It works exactly how I would expect if 1. no highlight compress at all is used and 2. exposure is 0 during rendering.

Then the threshold corresponds to the mapped rgb value and everything above gets processed. You can even leave your bloom + glare settings active when starting a new rendering - all ok, this is easy to handle and comprehensible.

- If highlight compression is used you have to set bloom + glare power to 0 or the threshold to a high value before starting to render because if not (having e.g. still 1.3 there) it starts to freak out from the beginning.

But then you still need to add the highlight compression value to the mapped rgb value as threshold to get everything above that rgb reading processed and nothing below.

So in almost all cases bloom and glare power settings have to be 0 when starting to render, and the threshold does not directly correspond to the cVFB readings? is this how it should work? Or is it just the current state of bloom and glare (linear output only) what puzzles me?

Thank you

(DB160728)
Never underestimate the power of a well placed level one spell.