Author Topic: Corona elements on photoshop  (Read 7856 times)

2014-04-30, 18:57:46

benfattino

  • Active Users
  • **
  • Posts: 6
    • View Profile
Hi. I try to work with elements, (Corona A6) but there is something doesn't work.
I use cessential pass and put it on photoshop:
First attemp.
- Direct
- Indirect (screen)
- Reflect (screen)
- Refrection (screen)
- Emission (screen)
- Translucency (screen)

Second attemp.
- Direct
- Indirect (lighten)
- Reflect (screen)
- Refrection (lighten)
- Emission (screen)
- Translucency (screen)

The second part of each image is the beauty pass, but in all cases there are a little difference.
Somebody has some advices to obtain same image? Thanks.

2014-05-01, 21:16:07
Reply #1

Ludvik Koutny

  • VIP
  • Active Users
  • ***
  • Posts: 2557
  • Just another user
    • View Profile
    • My Portfolio
There are two variables:

1, Add mode in Photoshop was never reliable, so i am not sure compositing elements in Photoshop may even be possible. I may be wrong though.

2, To compose render elements 1:1, you absolutely must render the output linearly. That means you need to have highlight compression parameter set at 1.

EDIT: Oh, i just noticed which blending modes you used...  yeah... absolutely no chance. You need to use Linear Dodge (Add) mode for all of the layers. Having highlight compression in Corona settings still applies too. And even then, Photoshop may not necessarily do it right.

2014-05-02, 00:41:23
Reply #2

Captain Obvious

  • Active Users
  • **
  • Posts: 167
    • View Profile
First of all, you're trying to do something that Photoshop simply was not built for, and is very insufficient for. If you want to recombine render elements like this, you'll need at least After Effects, but Fusion or Nuke would be better. Hell, even Blender's compositing tools are better suited for this. Or Autodesk Composite, I guess?

Secondly, screen is a nonlinear add function, and lighten simply picks the brightest value. You'll want linear dodge, which is just a straight-up add.

2014-05-02, 01:31:20
Reply #3

Juraj

  • Active Users
  • **
  • Posts: 4761
    • View Profile
    • studio website
I am not compositing myself, but when I do work in linear space in post, I didn't honestly notice a difference between doing so in Photoshop and AfterEffects. For compositing lineary you need either Photoshop Extended (I think CS4+, although I only ever had CS6 from start) or Photoshop CC, both which allows layers in 32bit mode(only time PS will allow you to load Linear RGB color profile, and will do so automatically if opening linear file, .exr/.hdr). I don't see why any problem would arise after than.
Please follow my new Instagram for latest projects, tips&tricks, short video tutorials and free models
Behance  Probably best updated portfolio of my work
lysfaere.com Please check the new stuff!

2014-05-03, 15:13:40
Reply #4

benfattino

  • Active Users
  • **
  • Posts: 6
    • View Profile
THANKS for answers.
I try to save all images in TIF format 32 bit.
I use the linear dodge (add) for compositing in photoshop and it's work.
All the operation must be do in 32 bit mode and only at the end save at 16 or 8 bit.

S.

2014-05-05, 13:59:56
Reply #5

Captain Obvious

  • Active Users
  • **
  • Posts: 167
    • View Profile
I am not compositing myself, but when I do work in linear space in post, I didn't honestly notice a difference between doing so in Photoshop and AfterEffects. For compositing lineary you need either Photoshop Extended (I think CS4+, although I only ever had CS6 from start) or Photoshop CC, both which allows layers in 32bit mode(only time PS will allow you to load Linear RGB color profile, and will do so automatically if opening linear file, .exr/.hdr). I don't see why any problem would arise after than.
The key difference between After Effects and Photoshop in terms of linear floating point workflow is that Photoshop loses more functionality in 32-bit mode than After Effects does.

Either way, both of them are downright horrible at it compared to a proper compositor. In Nuke, everything is done in linear floating point space internally anyway, regardless of your settings. In Fusion, things typically stay in the same bit depth as the input unless you specifically tell it otherwise, and either way everything works fine in 32-bit mode. You don't lose any functionality by using floating point images in either one of them. In After Effects, you lose some functionality and a lot of performance. In Photoshop, you lose even more functionality and performance.

For file formats, generally I would recommend using OpenEXR and buying the ProEXR plugin for Adobe products (it's available for both PS and AE). It lets you load multi-layer OpenEXRs, and makes it easier to deal with things like transparency which Photoshop natively loads incorrectly.