Author Topic: CAUSTICS AND PHOTONS  (Read 5631 times)

2019-07-11, 20:27:36

Alexandre Besson

  • Active Users
  • **
  • Posts: 132
  • http://www.alexandre-besson.com
    • View Profile
    • Alexandre Besson portfolio
Hi,


impossible to obtain correct caustics,
what's wrong with them ?

2019-07-11, 20:31:11
Reply #1

TomG

  • Administrator
  • Active Users
  • *****
  • Posts: 5462
    • View Profile
We'd need to know something about your scene set up (ideally, the scene itself, but at very least screen grabs of the settings). e.g. which version of Corona, is this the new Fast caustics solver, did you change anything in the developmental/experimental section, and so on.
Tom Grimes | chaos-corona.com
Product Manager | contact us

2019-07-11, 20:40:12
Reply #2

Alexandre Besson

  • Active Users
  • **
  • Posts: 132
  • http://www.alexandre-besson.com
    • View Profile
    • Alexandre Besson portfolio
Hi Tom,

I send you my scene (we transfer) in PM.

2019-07-11, 20:47:25
Reply #3

Alexandre Besson

  • Active Users
  • **
  • Posts: 132
  • http://www.alexandre-besson.com
    • View Profile
    • Alexandre Besson portfolio
Issue comes from corona 4 hotfix 1, caustics are great with corona 4.
« Last Edit: 2019-07-11, 21:18:39 by Alexandre Besson »

2019-07-12, 14:50:49
Reply #4

maru

  • Corona Team
  • Active Users
  • ****
  • Posts: 12754
  • Marcin
    • View Profile
Are you using region rendering here? Can you try if it's the same when you don't use any kind of region/crop rendering, just full frame?
Marcin Miodek | chaos-corona.com
3D Support Team Lead - Corona | contact us

2019-07-12, 15:43:04
Reply #5

Alexandre Besson

  • Active Users
  • **
  • Posts: 132
  • http://www.alexandre-besson.com
    • View Profile
    • Alexandre Besson portfolio
same result without region

2019-07-12, 15:45:39
Reply #6

TomG

  • Administrator
  • Active Users
  • *****
  • Posts: 5462
    • View Profile
Just a quick note, did you submit the scene via the private uploader too? As I noted in my private message, that's the only way for support / developers to have access to it (and trackable, connected to this thread). Thanks!
Tom Grimes | chaos-corona.com
Product Manager | contact us

2019-07-12, 15:48:38
Reply #7

Alexandre Besson

  • Active Users
  • **
  • Posts: 132
  • http://www.alexandre-besson.com
    • View Profile
    • Alexandre Besson portfolio
Tom, yes I did it.

2019-07-12, 16:06:37
Reply #8

TomG

  • Administrator
  • Active Users
  • *****
  • Posts: 5462
    • View Profile
Awesome, thanks Alexandre! :) That way you aren't dependent on just one person being able to have time to have a look - I appreciate it!
Tom Grimes | chaos-corona.com
Product Manager | contact us

2019-07-12, 17:15:14
Reply #9

maru

  • Corona Team
  • Active Users
  • ****
  • Posts: 12754
  • Marcin
    • View Profile
Here are my findings:
1) User box instead of plane for the pool water. It *should not* cause any issues, but better safe than sorry.
2) Unplug the RaySwitch map from the pool water reflection/refraction color. For some reason it seems that just plugging it cuts rays/s by half. So just use pure white or some specific color, if possible. It will be best not to use it at all.
3) Bump map on the water surface seems to be too strong, so it scatters caustics in all directions. Lowering the bump intensity to something like 0,1 helps greatly. You can also try using displacement instead of bump (autobump will take care of the rest and it is enabled by default).
4) Optionally: use 3ds Max native noise maps instead of Siger Noise. There is nothing wrong with using them, but they are definitely more detailed and more resource-heavy than standard maps.
Another thing is that the scene is quite large compared to the area which is visible and which should show caustics. It's a known issue that caustics will render slower/look worse in large scenes, and we already have a feature request logged for rendering caustics only in specified area (e.g. on specific materials only, or within some specific radius).

Let me know if those hints help or not.
Marcin Miodek | chaos-corona.com
3D Support Team Lead - Corona | contact us

2019-07-16, 16:19:49
Reply #10

Alexandre Besson

  • Active Users
  • **
  • Posts: 132
  • http://www.alexandre-besson.com
    • View Profile
    • Alexandre Besson portfolio
Something is wrong with the hotfix 1.

2019-07-16, 18:45:26
Reply #11

Alexandre Besson

  • Active Users
  • **
  • Posts: 132
  • http://www.alexandre-besson.com
    • View Profile
    • Alexandre Besson portfolio
In my case, only refraction's caustics are usable in production.

2019-07-16, 21:17:01
Reply #12

pokoy

  • Active Users
  • **
  • Posts: 1861
    • View Profile
Hmm, this is something the devs should be able to judge, but if a standard noise is being used, I think it's animated by default. This could lead to a smoother caustic pass since hf1 doesn't take motion blur into account. Again, only a guess, I have no idea if this is the case with the comparison between v4 and v4 hf1.

2019-07-17, 13:10:25
Reply #13

maru

  • Corona Team
  • Active Users
  • ****
  • Posts: 12754
  • Marcin
    • View Profile
We will look into this.
Please make sure you are not doing comparisons with region rendering, only full frame.
I think currently there is some issue with regions + caustics, but can't confirm it yet.
Marcin Miodek | chaos-corona.com
3D Support Team Lead - Corona | contact us

2019-07-18, 11:05:37
Reply #14

maru

  • Corona Team
  • Active Users
  • ****
  • Posts: 12754
  • Marcin
    • View Profile
Hmm, I am getting pretty much reverse results to yours. Are you sure you did not mistake the file names?
Is it ok to share renders of that scene here?
Marcin Miodek | chaos-corona.com
3D Support Team Lead - Corona | contact us